Natural killer cell formulation and method for its production
By expanding NK cells in a lactate-rich medium and transiently enriching FasL on their surface, the method addresses lactate-induced suppression, improving NK cell cytotoxicity and therapeutic outcomes against tumors.
Patent Information
- Application Number
- PCT/EP2026/064234
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-16
- Filing Date
- 2026-05-15
- Publication Date
- 2026-11-19
AI Technical Summary
Elevated lactate concentrations in the tumor microenvironment impair the therapeutic efficacy of NK cell-based therapies by suppressing NK cell cytotoxicity, creating an immunosuppressive environment that hampers effective antitumor responses, particularly in solid tumors.
A method for producing NK cell formulations involves expanding NK cells in a culture medium with elevated lactate concentrations and transiently enriching Fas ligand (FasL) on the surface of the NK cells to enhance their resistance to lactate-induced suppression, using either exogenous decoration or transient overexpression of FasL.
The NK cell formulations exhibit increased cytotoxicity and resistance to lactate concentrations typical of tumor microenvironments, thereby enhancing therapeutic efficacy against cancer.
Smart Images

Figure EP2026064234_19112026_PF_FP_ABST
Abstract
Description
[0001] Natural killer cell formulation and method for its production Field of the invention
[0002] The field of the present invention relates to producing natural killer (NK) cell formulations for cancer therapy.
[0003] Background
[0004] The therapeutic potential of immune cell-based therapies has received considerable attention in recent years, driven by promising outcomes observed particularly in the field of oncology. Among immune effector cells, NK cells represent a class of lymphocytes that inherently exhibit potent cytotoxicity against malignant and infected cells. Their capacity to recognize and destroy tumour cells, while typically sparing normal healthy tissues, positions NK cells as particularly attractive candidates for adoptive cell therapy applications. Consequently, NK cell-based therapeutics have emerged as powerful complementary or alternative treatment modalities (see Vivier et al, 2024 ), especially in cases where traditional chemotherapy, radiotherapy, or conventional immunotherapy are inadequate.
[0005] NK cell formulations can address multiple oncological indications (Myers et al, 2021 ). Among them are haematological malignancies such as acute myeloid leukaemia (AML), acute lymphoblastic leukaemia (ALL), lymphoma, multiple myeloma, and solid tumours including melanoma, renal cell carcinoma, ovarian cancer, and lung cancer. WO 2022 / 104109 Al discloses formulations of genetically engineered NK cells, also for cancer therapy. WO 2022 / 260968 Al discloses transgene expression in NK cells, and providing NK cells for cancer immunotherapy.
[0006] CN 113462645 A discloses a composition comprising NK cells and an activator of plant-derived extracellular vesicles (EVs). Itis described that incubating NK cells with these EVs increases their cytotoxicity.
[0007] WO 2022 / 080894 Al describes a feeder-free method for culturing NK cells using a specific medium containing IL-2, an anti-NKp46 antibody, and plasma.
[0008] US 2024 / 0307536 Al discloses "super-activated killer" cells produced by expanding peripheral blood mononuclear cells for at least nine days in the constant presence of both IL-2 and a CD3 activator.
[0009] WO 2023 / 072813 Al discloses a semi-automated method for the large-scale expansion of NK cells, focusing on optimising the expansion by controlling culture conditions.
[0010] Berg et al. relates to a method for the large-scale expansion of clinical-grade human NK cells. The method involves coculturing bead-purified NK cells with irradiated Epstein-Barr virus-transformed lymphoblastoid cell lines (EBV-LCLs) as feeder cells. Granzin et al. describes a fully automated, large-scale process for the expansion and activation of clinical-grade NK cells, also utilising irradiated EBV-LCLs as feeder cells.
[0011] Elevated lactate concentrations in some tumour microenvironments (TMEs) are a hallmark of altered cancer cell metabolism, particularly aerobic glycolysis, commonly termed the Warburg effect. Although the Warburg effect -characterized by increased glucose uptake and lactate production, even in the presence of oxygen - is broadly prevalent among malignant cells, the extent and implications of lactate accumulation may differ among various tumour types. Specific cancers are consistently reported to have notably high lactate concentrations in their TME. In particular, while physiological lactate concentrations in blood and healthy tissue are typically around 1.5-3 mM, many cancers harborsignificantly higher lactate concentrations in their TME, e. g. around 10-30 mM (de la Cruz-López et al, 2019).
[0012] Importantly, lactate is not only a secondary product of cellular metabolic waste of tumour cells, but also a key molecule involved in carcinogenesis as well as in tumour immune evasion (de la Cruz-López et al, 2019). Elevated lactate levels inhibit effector immune cells such as NK cells, while supporting immunosuppressive cells, such as regulatory T cells and myeloid-derived suppressor cells, thus fostering an immunosuppressive environment (Gu et al, 2025).
[0013] This immunosuppressive nature of the TME, characterized in particular by elevated lactate levels, poses substantial challenges to the therapeutic efficacy of NK cell-based therapies, particularly in the context of solid tumours.
[0014] Elevated lactate concentrations within the TME substantially impair NK cell function through various interrelated mechanisms. In brief, lactate suppresses the cytotoxic activity of NK cells, presumably by impairing their metabolism, thereby severely reducing their ability to sustain effective antitumour responses. Accordingly, efficacy of NK cell therapy against solid tumours in particular is still poor (see also Terren et al, 2019).
[0015] Approaches to overcoming these immunosuppressive barriers are predominantly centred around reprogramming or conditioning NK cells prior to therapeutic administration. For instance, genetic modification to enhance resistance to metabolic stress, cytokine pre-activation regimens such as interleukin-2 ( IL-2 ) stimulation, and combinatorial approaches e. g., involving co-administration of immune checkpoint were suggested. Some of these approaches are reviewed in Jia et al, 2023. Further along these lines, van Vliet et al, 2021, discusses overcoming the hurdle of tumour immune escape in the context of metastatic melanoma.
[0016] However, there remains an urgent unmet medical need to develop novel, more effective approaches capable of addressing the hostile metabolic and immunological milieu of tumours, thereby substantially enhancing NK cell therapeutic outcomes.
[0017] Accordingly, it is an obj ect of the invention to provide NK cell formulations for cancer therapy (and methods for their production) which address this need and increase resistance of NK cells to the conditions in the TME, first and foremost by increasing their cytotoxicity under high lactate concentrations typical of the TME of many cancers.
[0018] Summary of the invention
[0019] The present invention provides an in-vitro method for producing NK cell formulations for cancer therapy, comprising at least the following steps:
[0020] - expanding at least one NK cell in a first culturing step,
[0021] - transiently enriching (preferably native) Fas ligand (FasL) on the surface of the expanded NK cells, and
[0022] - providing the expanded and transiently FasL-enriched NK cells in an NK cell formulation for cancer therapy.
[0023] Preferably, in the first culturing step, the lactate concentration exceeds (at one point, typically at the end of the culturing) a threshold of 5 mM lactate, preferably 7.5 mM lactate, more preferably 10 mM lactate or even 20 mM lactate (due to accumulation of lactate in the culture medium).
[0024] The present invention also provides an NK cell formulation for cancer therapy obtainable by this method.
[0025] In another aspect, the present invention provides an NK cell formulation for cancer therapy, wherein (preferably native) FasL is transiently enriched on the surface of NK cells.
[0026] In the course of the present invention, it was surprisingly discovered that medium supplementation with soluble FasL protein during a cytotoxicity assay strongly increased the resistance of NK cells to higher lactate concentrations (this was even more surprising due to the fact that the NK cells used in that assay had lost a significant portion of their baseline FasL expression due to having been pre-incubated in medium with 15 mM lactate for three days). However, systemic administration of FasL to patients is not feasible as off-target effects are expected (which would lead to severe side effects). Accordingly, presentation of FasL directly on the surface of NK cells is a much more viable approach. It is also expected that the effect of lactate resistance will be stronger when FasL is presented in this manner, in particular membrane-bound ( full-length) FasL which will also lead to further downstream effects in the NK cells. The enrichment of FasL being only transient confers another advantage, because it also serves to reduce the risk of delayed off-target effects (which may occur when more time after administration of the NK cells has passed).
[0027] As explained in further detail below, transient enrichment of FasL on the surface of NK cells may be achieved by decorating the surface of NK cells with (exogenous) FasL or transient overexpression of FasL within the cells.
[0028] Detailed description of the invention
[0029] The following detailed description and preferred embodiments apply to all aspects of the invention and can be combined with each other without restriction, except where explicitly indicated.
[0030] In addition to perf orin / granzyme release, activated NK cells display death ligands (FasL and TRAIL) on their surface. FasL (also known as CD95L; human FasL is represented by UniProt entry P48023) is a type II transmembrane protein of the tumor necrosis factor (TNF) superfamily that binds the Fas (CD95) receptor to trigger apoptosis. It is encoded by the FASLG gene. The FasL-Fas interaction activates caspase-8 and the extrinsic death pathway in Fas-expressing cells, and serves in immune regulation. FasL can be proteolytically cleaved to a soluble form, but membrane-bound FasL is generally more effective at inducing target cell apoptosis. Under healthy conditions, immune effector cells (e. g., activated T cells and NK cells) can express FasL to eliminate unwanted cells. A recent overview on FasL functions in cancer and autoimmune diseases is given in Rossin et al, 2019.
[0031] Under certain conditions, NK-cell FasL contributes to tumour cell killing (cf. Guia et al, 2025). However, strikingly, the FasL pathway is also hijacked by tumours as a means of immune evasion. Many cancers express FasL themselves and secrete soluble FasL and FasL-bearing vesicles to induce apoptosis in infiltrating lymphocytes and NK cells (see e. g., Andreola et al, 2002, Azambuja et al, 2020). Furthermore, while some tumour cells may co-express the Fas receptor, downstream apoptotic signalling is often impaired, allowing tumour cells to resist Fas-mediated apoptosis. Taken together, it was therefore surprising that increasing FasL concentration in the medium led to lactate-resistant cytotoxicity in NK cells, as found in the course of the present invention (see examples),instead of leading to reduced NK cell cytotoxicity (or, at best, having no effect at all).
[0032] NK cell formulations of the present invention can be generated starting from autologous or allogeneic sources (the latter being extensively discussed e. g., in Berrien-Elliott et al., 2023). Autologous NK cells, usually derived from the patient's own blood or bone marrow, inherently avoid risks associated with immunological rejection. Allogeneic ("off-the-shelf") NK cell products offer distinct advantages, including immediate availability, improved consistency, and the potential for large-scale industrial manufacturing. Such allogeneic NK cell products may employ carefully selected donor-derived cells, umbilical cord blood cells, or induced pluripotent stem cell ( iPSC) -derived NK cells.
[0033] Clinical efficacy of NK cell therapy depends on delivering an adequate and therapeutically effective cell dose. Typical NK cell doses used in the clinic vary considerably depending on the indication, patient condition, and formulation characteristics, but may range between 106and 108viable cells per kilogram of patient body weight. Strategies exist to non-genetically further increase the cytotoxicity of the cells (see Villalba et al, 2020). These strategies may also be employed in the context of the present invention, in addition to the inventive transient FasL-enrichment.
[0034] In the context of the method of the present invention, in particular to reach these therapeutically effective doses, at least one NK cell (which may be autologous or allogeneic, as disclosed above) is expanded in vitro in a first culture. In vitro expansion may rely on feeder cells, typically irradiated or genetically modified stromal cells, immune cells, or immortalized cell lines, to provide cytokines and costimulatory signals essential for NK cell proliferation andactivation. Alternatively, feeder-free methods may also be used. Such feeder-free systems typically utilize media e. g., supplemented with cytokines (e. g., IL-2, IL-15, IL-21 ), but usually still with serum as a (chemically undefined) component. A further suitable approach to NK cell expansion involves the use of artificial feeder constructs, which is also understood herein as "feeder-free". These artificial feeder constructs may be, for instance, microbeads or nanoparticles coated with specific stimulating substances, such as antibodies against NK cell activating receptors (e. g., anti-CD16, anti-NKG2D) or recombinant ligands for costimulatory molecules, thereby mimicking the stimulatory signals provided by live feeder cells. Specific NK cell expansion methods suitable in the context of the invention are for instance disclosed in Spanholtz et al, 2011, Oyer et al, 2015, Bröker et al., 2019, and Wang et al, 2024. An overview of NK cell manufacturing methods which may be adapted for the present invention is given e. g., in Motallebnejad, et al., Klingenmann, 2015, and Fang et al., 2022.
[0035] The expanding step starts with at least one NK cell, preferably a population of NK cells isolated from a suitable source, such as peripheral blood mononuclear cells (PBMCs), cord blood, induced pluripotent stem cells (iPSCs) - either of which may be obtained from a donor or from the cancer patient - or from an established NK cell line.
[0036] When a (mixed) population of cells is used as the starting material of the expanding step, it is preferred that this initial population exhibits a high degree of purity with respect to NK cells. Specifically, it is preferred that the VCC of NK cells in the starting cell population of the expanding step ("the at least one NK cell") is at least 70%, preferably at least 80%, more preferably at least 90%, even more preferably at least 95% or even at least 97%, especiallyat least 98% or even at least 99% of the total VCC of the population. Such purity can be achieved through various well-established NK cell enrichment techniques, for example, magnetic-activated cell sorting (MACS) or fluorescence-activated cell sorting (FACS) employing antibodies against specific NK cell markers (e. g., CD56 positive selection, CD3 negative depletion). Starting the expansion with a highly purified NK cell population is advantageous as it minimises off-target effects and improves the overall efficacy and / or consistency of the final NK cell product.
[0037] Alternatively, the expanding step may be started with one or more cells from a NK cell line (in other words, "the at least one NK cell" may be one or more cells from an NK cell line). As such cell lines are typically pure, they offer a distinct advantage for producing "off-the-shelf" therapeutic formulations. NK cell lines, such as NK-92, KHYG-1, NKL, or derivatives thereof, are populations of NK cells that have been immortalised or are capable of continuous proliferation in vitro. Utilising an NK cell line as the starting material circumvents the need for donor-derived primary cells for each batch, thereby eliminating donor-to-donor variability and the logistical complexities associated with collecting and processing primary human material. This approach allows for a highly standardised and scalable manufacturing processes, as the cell line provides a consistent and inexhaustible source of NK cells with well-characterised properties. Furthermore, NK cell lines can be more readily genetically engineered to express specific receptors or, as in the context of the present invention, to facilitate transient overexpression of FasL (e. g., by already having recombinant FasL under the control of an inducible promotor).
[0038] The expansion occurs in a first culturing step designed to significantly increase the number of NK cells (preferably atleast 104-fold, more preferably at least 105-fold, in particular at least 106-fold or even at least 107-fold).
[0039] Culture media supplemented with cytokines like IL-2, IL-15, and / or IL-21, and / or other stimulants and potentially feeder cells, to support robust proliferation may be used.
[0040] Put differently, expanding at least one NK cell may encompass initiating culture from a single NK cell, a clonal population thereof (e. g., produced in a pre-culture or starter culture), or an NK cell population comprising heterogeneous subsets or, more generally from any biological sample containing NK cells (e. g., obtained from individuals such as patients (i. e., autologous) or donors (i. e., allogeneic). In practice, starting cultures typically contain an NK cell population enriched from sources such as peripheral blood mononuclear cells (PBMCs), umbilical cord blood, or induced pluripotent stem cells (iPSCs), or cells from NK cell lines for off-the-shelf solutions.
[0041] As transient FasL enrichment has turned out to be particularly suitable for NK cells pre-incubated with lactate (see Example, in particular Fig. 7 ), it is particularly suitable to address reduced cytotoxicity in view of lactate having accumulated in the culture medium. Therefore, according to a preferable embodiment, the lactate concentration exceeds (at one point, typically at the end of the culturing) during the expansion phase, i. e. in the first culturing, the threshold of 5 mM lactate, preferably 7.5 mM lactate, more preferably 10 mM or even 20 mM lactate. Along the same lines, the inventive FasL enrichment is therefore particularly advantageous when the first culturing is batch or fed-batch culturing (see below).
[0042] The first culturing step (the "expansion phase") may be carried out e. g., as a perfusion process or as a batch process. It may also be carried out as a fed-batch process.
[0043] A "perfusion process" involves the continuous or semi-continuous addition of fresh nutrient medium to the bioreactor and the simultaneous removal of spent medium containing metabolic waste products, while at least a large portion (or all) of the cells are retained within the culture system, for instance, by means of a filtration device, a centrifuge, or an acoustic separator.
[0044] By contrast, the term "batch process" refers to a culture system wherein all necessary components, including the initial cell inoculum and the complete volume of growth medium containing all nutrients, are combined at the initiation of the culture, and the process is allowed to proceed without further addition of fresh medium or removal of spent medium during the cultivation period. Consequently, nutrients are consumed and metabolic by-products accumulate over time.
[0045] The term "fed-batch process" denotes a culture system wherein nutrients supporting NK cell proliferation and function, e. g., glucose, amino acids, vitamins, minerals, and cytokines such as IL-2, IL-15, and optionally IL-21, and / or other stimulants are added periodically or continuously (e. g., dynamically in response to monitored parameters) throughout cultivation to support cell growth and metabolism, optionally after an initial delay. However, unlike perfusion or continuous cultivation systems, metabolic waste products such as lactate or ammonia, are not continuously removed but generally allowed to accumulate progressively in the culture medium.
[0046] At this point, NK cells have typically exhibited significant expansion, but their cytotoxic capacity may be partially compromised, e. g., due to accumulation of waste products in the medium, or high cell density. Accordingly, the method preferably further comprises a second culturing step also termed the "recovery phase", designed to restore or recover atleast a portion of the compromised NK cell cytotoxic capacity. Prior to initiating this recovery phase, NK cells may be transferred to fresh culture medium (or the existing culture medium may be significantly supplemented with fresh culture medium). The starting lactate level should be significantly lower than the highest lactate concentration during the expanding step. Ideally, it should approach zero or near-physiological baseline concentrations (e. g., typically below 2 mM lactate). Yet in practice, it may be higher than that as achieving the replacement of existing culture medium beyond a certain point (e. g., beyond 95 vol%) is challenging.
[0047] The recovery step (if present, e. g., preferably when the first step is a batch or fed batch process) occurs after completing the initial expansion phase, i. e., it is not concurrent with the first culturing step. The recovery phase may feature reduced, minimal or even no NK cell expansion compared to the first culturing step. Conditions during the recovery phase may include optimized media formulations, cytokine concentrations, temperature, and pH settings to facilitate maximal restoration of cytotoxic function.
[0048] It is preferred that the recovery step occurs directly after the first culturing step, i. e. with as few processing steps as possible (e. g., only with exchanging part of the medium). In other words, it is preferred that the second culturing step occurs within 12 hours, preferably within 6 hours, more preferably within 3 hours, even more preferably within 2 hours or even within 1 hour, especially within 30 minutes or even within 5 minutes, after the first culturing step has concluded. Alternatively, or in addition thereto, it is preferred that the temperature is kept above 5°C, especially above 10°C, in particular above 15°C, between the first and the second culturing step, in particular that no cooling orfreezing step takes place. The second culturing step may be a continuous process, a batch process or a fed-batch process.
[0049] Preferably after at least expansion has been concluded (or concluded to a large extent, e. g., to at least 80% or at least 90%), the step of transiently enriching FasL on the surface of these expanded NK cells is performed. Transient enrichment ensures that the enhanced phenotype is temporary, mitigating long-term safety concerns associated with constitutive FasL overexpression. This transient enrichment may occur e. g., in the recovery step described above or in a separate step. If the transient enrichment involves FasL overexpression, it is preferably conducted during this step; if it involves decorating with exogenous FasL, it is preferably conducted after this step.
[0050] After optional further processing steps such as harvesting (e. g., by standard techniques such as centrifugation or filtration) and washing, the NK cells (which are both expanded in number and transiently enriched with FasL on their surface, preferably membrane-bound FasL, in particular membrane-bound full-length FasL) are provided as a formulation for cancer therapy (i. e., in a pharmaceutically acceptable formulation suitable for therapeutic administration, for instance, as an intravenous infusion). This typically involves resuspending the NK cells in a suitable carrier solution, potentially including stabilisers or cryoprotectants if the formulation is intended for cryo-storage.
[0051] In embodiments, the NK cell formulation may be used fresh (in particular in case of autologous therapy). Alternatively, in particular in case of an off-the-shelf solution, it may be frozen and e. g., as stored in frozen aliquots. After thawing (e. g., at the site of use), the NK cell formulation may require further final processing, such as simple dilution, washing ( for example, to remove residual DMSO), or brief
[0052] recovery incubation steps. In other words, the NK cell formulation of the present invention is not necessarily the final formulation which is administered to the patient.
[0053] In a preferred embodiment, the step of transiently enriching FasL on the surface of the expanded NK cells comprises decorating the surface of these cells with exogenous (i. e., externally produced - not produced by the cells themselves) FasL, typically after the expanding has been (substantially) concluded. This approach involves applying or conjugating FasL (which may e. g., be purchased in purified form) to the NK cell surface. Such decoration can be achieved e. g., through various non-genetic techniques, for instance, by biotinylating the FasL protein and binding it to streptavidin expressed on or conjugated to the NK cell surface, or by using lipid-based anchoring methods to embed FasL into the cell membrane, or via antibody-mediated tethering. This offers the advantage of modifying the NK cell phenotype without introducing foreign genetic material into the NK cells, thereby streamlining the manufacturing process and alleviating regulatory concerns associated with genetically modified cells.
[0054] In contexts unrelated to cancer, such FasL decoration approaches were disclosed and may be adapted for the present invention. For instance, Yolcu et al, 2002, used FasL-decorated cells (based on streptavidin fused with the extracellular portion of FasL) to prevent islet graft rejection. Another FasL decoration approach is disclosed in US 7, 927, 602 B2 and US 8, 076, 096 B2. An approach termed ProtEx for decoration with FasL and other exogenous proteins is disclosed in Singh et al, 2005.
[0055] Alternatively or additionally, in another preferred embodiment, the transient enrichment of FasL is achieved through transient overexpression of FasL within the expanded NK cells. This may involve (or be preceded by) introducinggenetic material encoding FasL into the NK cells in a manner that leads to temporary, rather than permanent, overexpression (according to a particular preference, the chromosomes of the NK cells remain unchanged, in particular, the vector used does not integrate into a chromosome of the NK cells). Preferably, this transient overexpression is initiated only after the expansion step has been substantially completed. Inducing overexpression post-expansion has the advantage that the expansion phase, which is necessary for achieving sufficient cell numbers, is not perturbed by the metabolic demands associated with FasL overexpression or the delivery system used. Furthermore, timing the overexpression closer to the point of administration may maximise the surface presentation of FasL on the final therapeutic formulation. In this context, the term "overexpress" means that FasL is detectably present at the surface of NK cells at levels above an appropriate control or reference population. Preferably, the total expression level (normalized to VCC) is at least 50%, preferably at least 100%, even more preferably at least 200% higher than the total expression level (normalized to VCC) of the control.
[0056] If transient overexpression is employed, a transient expression system may be introduced into the expanded NK cells (in particular only after the expansion step has been substantially concluded) to achieve the desired FasL overexpression, preferably wherein the transient expression system is introduced into the expanded NK cells by electroporation, virus, virus-like particle or lipid particles (such as lipid nanoparticles or liposomes). Such systems are designed to deliver the genetic instructions for producing FasL (e. g., its corresponding DNA or RNA sequence) into the cells without integrating into the host cell genome and / or by using methods where expression diminishes over time (e. g., byrelying on inducible expression). This ensures that the overexpression of FasL is temporary, aligning with the transient nature of the enrichment sought by the invention. This approach provides a controllable means to induce endogenous production and surface presentation of additional FasL by the NK cells themselves (as opposed to relying on exogenously produced FasL when using decoration). While the method of protein production (and targeting to the plasma membrane) is endogenous in this embodiment, the FasL sequence itself is not necessarily endogenous. Although using the endogenous (wild-type) FasL sequence is preferred as discussed in the following paragraphs (e. g., human FasL if the patient is human), it is also contemplated that mutant forms of FasL may be used. In this connection, e. g., US 6,451,759 B1, discloses a non-cleavable form of FasL, or WO 2024 / 246427 Al discloses FasL mutants (i. e., non-wildtype sequences). Such a form may be overexpressed in NK cells in the context of the present invention. Other suitable FasL mutants (i. e., nonwildtype sequences) are disclosed e. g., in WO 2022 / 187663 Al.
[0057] In a further preferred embodiment, the FasL protein sequence is the native FasL sequence. This means that the sequence corresponds to a naturally occurring FasL protein, without artificial mutations intended to alter its primary structure or biological function. In particular, it is preferred that both the NK cells being produced or formulated according to the invention and the FasL protein sequence are of human origin, to accurately reflect the clinical scenario and ensure relevance for human therapy. First and foremost, native human FasL is present as transmembrane protein (mFasL), and this membrane-bound form is preferred in the context of the present invention. A soluble form of FasL (sFasL) can be generated through physiological proteolytic cleavage of mFasL, for example by metalloproteinases. The canonical amino acid
[0058] sequence for full-length FasL is known and is disclosed, for example, in the UniProt database under accession number P48023 for human FasL). It is thus preferred that FasL comprises or consists of 281 amino acids, i. e. amino acids 1-281 of UniProt P48023. This sequence encompasses an N-terminal cytoplasmic domain, a transmembrane domain, and a C-terminal extracellular domain which mediates receptor binding. The membrane-bound form includes all these domains. Soluble FasL typically comprises the C-terminal extracellular portion, with cleavage sites reported that can result in soluble forms of varying lengths, for example, a fragment corresponding to amino acids 103-281 or 138-281 of the full-length human precursor.
[0059] Put differently, in the entire context of the present invention, it is preferred that FasL comprises (or consists of ) the sequence of UniProt P48023 or a extracellular fragment thereof such as amino acids 103-281 or 138-281 of Uniprot UniProt P48023.
[0060] In another preferred embodiment, in case of FasL overexpression, the cytoplasmic domain of the FasL is native, in particular having the sequence identified by UniProt P48023, positions 1-80.
[0061] According to a further preferred embodiment, the transient expression system is selected from the group consisting of non-chromosomal DNA vectors, such as plasmids, and exogenous RNA constructs, such as exogenous messenger RNA (mRNA). Such DNA or RNA typically codes for the ( full-length) FasL protein sequence. Plasmids may be introduced via methods like electroporation or viral transduction (using non-integrating and typically non-replicating viruses), leading to temporary gene expression. Exogenous RNA, in particular exogenous mRNA, typically delivered via lipid nanoparticles or electroporation (but may e. g., also be delivered by viruses), offers a particularly attractive option as it is directly translated inthe cytoplasm without needing to enter the nucleus and inherently possesses a limited half-life, ensuring transient expression and eliminating any risk of genomic integration (as small as it may be with non-chromosomal DNA vectors). The use of such systems facilitates reliable and controlled transient overexpression of FasL.
[0062] In another preferred embodiment, the transient expression system is introduced into the NK cells only after the expanding step. This temporal separation of expansion and genetic modification allows for independent optimisation of both processes; the expansion phase focuses solely on maximising cell yield and viability, while the subsequent introduction of the expression system allows for efficient modification of the expanded NK cell population shortly before formulation. This strongly increases the likelihood that the NK cells possess the desired FasL-enriched phenotype at the time of administration.
[0063] Under some circumstances, fratricide during expansion may diminish NK cell viability or efficacy (see e. g., Valeri et al, 2022 ). The temporal separations discussed above may already act as safeguards. However, in further embodiments, it may be advantageous to modulate the levels or interaction of FasL and Fas receptor during the expansion phase, such that the risk of fratricide is reduced. Therefore, in a preferred embodiment, the expression of FasL and / or Fas receptor is transiently downregulated during the expanding step. Such downregulation, achievable for example using RNA interference technologies like siRNA or shRNA delivered transiently.
[0064] Alternatively, or in addition thereto, the interaction between FasL and Fas receptor may be inhibited during the expansion phase. This inhibition can be accomplished, for instance, by adding blocking antibodies or decoy ligands targeting either FasL or Fas receptors to the culture medium during the firstculturing step. Care should be taken that the downregulation or blocking is sufficiently transient, such that it does not interfere with therapeutic efficacy (in particular lactate resistance).
[0065] NK cells are known to comprise distinct subpopulations, often characterised by the expression level of surface markers such as CD56 (e. g., CD56brightand CD56dimsubsets), which can differ in their functional properties. In a preferred embodiment, the method involves selectively enriching the CD56brightNK cell subpopulation during the expanding step. This acts as another safeguard to avoid (Fas / FasL-facilitated) fratricide, as CD56brightNK cells generally exhibit lower surface expression of Fas (CD95) and Fas ligand (FasL) compared to CD56dimNK cells, thereby reducing the likelihood of activation-induced cell death among the expanding NK cell population. Although fratricide was not observed as a significant problem during the culture experiments performed in the course of the present invention, further reduction of fratricide likelihood may be desirable, especially in culturing methods reaching high cell density.
[0066] Furthermore, the CD56dimsubset is typically more cytotoxic and exhibits greater effector function, making it more desirable for therapeutic applications. The enrichment of this subset is a preferred option (preferably before the transiently enriching step, in particular after expansion or at the end of expansion), in particular in addition to enriching CD56brightNK cells therebefore as explained in the paragraph above, and enhances both the viability and the functional potency of the resulting NK cell preparation.
[0067] CD56bright / CD56dimNK subpopulations are also discussed in Schwane et al, 2020. It is understood that these terms as used herein do not refer to absolute, predefined levels of CD56 surface presentation or fluorescence intensity that are
[0068] universally fixed across all samples, experimental conditions, or analytical instruments. Rather, these designations are relative and are defined operationally within the context of a specific NK cell population being analysed, typically using flow cytometry. By way of example, when NK cells are stained with a fluorescently labelled antibody specific for CD56, and subsequently analysed by flow cytometry, a distribution of fluorescence intensities corresponding to CD56 expression is observed. The delineation between these "dim" and "bright" populations may be established by setting analytical gates on a flow cytometry histogram or dot plot based on the observed bimodal or differential expression pattern within that particular sample. Thus, CD56dimcells are those that express relatively less CD56 compared to CD56brightcells from the same sample or experimental cohort, and vice versa.
[0069] To facilitate storage, distribution, and clinical logistics, the method preferably further comprises freezing the NK cell formulation. This may involve cryopreserving the expanded and transiently FasL-enriched NK cells by suspending them in a suitable cryoprotective medium containing agents like dimethyl sulfoxide (DMSO) and / or human serum albumin, followed by controlled-rate freezing and storage at ultra-low temperatures (e. g., in liquid nitrogen). Advantageously, the formulation comprises cryoprotectants, such as DMSO. In embodiments, the freezing step advantageously enables the production of off-the-shelf NK cell therapeutic products that can be stored as ready-to-use aliquots.
[0070] In an alternative or additional preferred embodiment the method comprises storing the NK cell formulation at a temperature between 0°C and 10°C. More preferably, the storage temperature is maintained between 2 °C and 8 °C, even more preferably between 3°C and 6°C. Such refrigerated storage may be maintained for a period of at least 1 hour, preferably forat least 12 hours, more preferably for at least 24 hours, even more preferably for at least 48 hours, yet more preferably for at least 72 hours, or at least 96 hours, or even for at least 5 days. Typically, however, such refrigerated storage would not exceed 14 days, and preferably would not exceed 10 days, to avoid detrimental effects on cell viability and functionality as well as possible. This short-term storage facilitates immediate or near-immediate use in a clinical setting. This mode of storage is particularly beneficial when the NK cells are derived from the patients themselves (autologous setting) and, especially, when the cells do not undergo any freeze-thaw cycle throughout the entire manufacturing and formulation process. Avoiding cryopreservation altogether for autologous therapies can be advantageous as it avoids the cellular stress and potential damage associated with freezing and thawing, which may sometimes impair cell viability, recovery, and subsequent in vivo effector functions. This approach is particularly suited for scenarios involving on-site or near-patient manufacturing. In yet another preferred embodiment, NK cells in the first culturing step (expanding step) are expanded to a VCC of at least 107, preferably at least 108, especially at least 109. This substantial expansion facilitates even large-scale production suitable for both autologous and allogeneic ("off-the-shelf" ) applications.
[0071] According to a further preference, in the NK cell formulation (as obtained by the inventive method), the total VCC is at least 107, preferably at least 108, especially at least 109. Additionally, or alternatively, the cytotoxic capacity (total cytotoxic capacity or, more preferably, cytotoxic capacity of the NK cells) is at least 106, preferably at least 107, especially at least 108. High VCC at formulation beneficially provides sufficient cell numbers upon administration to thepatient. Concurrently, ensuring high cytotoxic capacity at the time of formulation guarantees potent NK cell activity may directly translate into enhanced therapeutic effectiveness upon administration.
[0072] In another preferred embodiment, the VCC of NK cells in the formulation constitutes at least 70%, preferably at least 80%, more preferably at least 90%, even more preferably at least 95% or even at least 97%, especially at least 98% or even at least 99% of the total viable cell count of the formulation. Such high purity minimises the presence of potentially unwanted contaminating cells (e. g., residual feeder cells, other lymphocyte populations) and ensures that the therapeutic effect is primarily mediated by the intended NK cell population. Purity may be assessed e. g., using flow cytometry, in particular based on characteristic NK cell markers such as CD56 and CDS’.
[0073] As mentioned above, FasL is transiently enriched on the surface of the NK cells (as provided in the NK cell formulation for cancer therapy). The transient nature of the enrichment implies that the levels of surface FasL are elevated compared to baseline, but will decrease over time post-administration or upon in vitro culture of these NK cells under standard conditions (e. g., after one day, after two days, after three days, after one week or after two weeks). In particular, "transiently enriched" (as achieved e. g., by decoration or transient overexpression) may be understood as surface FasL decreasing by at least 50% upon 90 days (preferably 60 days, in particular 30 days, especially 14 days or even 7 days) in vitro culture of these NK cells under standard conditions.
[0074] This transient enrichment distinguishes these NK cells both from stably modified NK cells (as e. g., disclosed in US 2022 / 0288117 Al ) and unmodified NK cells of the same origin.
[0075] As used herein, "standard conditions" preferably refer to in vitro culture in medium (in particular RPMI-1640 medium) supplemented with 10% fetal bovine serum (FBS), 1% penicillinstreptomycin, and 100 IU / mL IL-2, maintained at 37 °C in a humidified atmosphere containing 5% CO2, in particular under perfusion conditions.
[0076] In a preferred embodiment of the NK cell formulation, the surface of the NK cells is decorated with (exogenous) FasL, asproduced using non-genetic surface modification techniques (discussed above). Such formulations contain NK cells carrying externally applied FasL molecules bound to their surface.
[0077] Alternatively or additionally, in another preferred embodiment of the NK cell formulation, the NK cells transiently overexpress FasL.
[0078] In formulations where transient overexpression is employed, the NK cells may, in embodiments, comprise a transient expression system for FasL overexpression. In other words, the system used to deliver the genetic material (e. g., plasmid DNA, mRNA molecules), remnants thereof, may still be present within the NK cells in the final formulation.
[0079] The formulation should contain a clinically relevant number of NK cells. Thus, in a preferred embodiment, the cell formulation comprises an NK cell dose of at least 107, preferably at least 108, especially at least 109viable NK cells. This ensures that a sufficient number of therapeutic cells are administered to the patient to potentially achieve a clinical response.
[0080] In a further preferred embodiment, the cell formulation has a cytotoxic capacity (in total or, more preferably, NK cell cytotoxic capacity) of at least 106, preferably at least 107, especially at least 108.
[0081] In another preferred embodiment, the NK cells and / or the FasL used are human.
[0082] The NK cell formulation is preferably provided as a (typically aqueous) cell suspension (in particular for intravenous administration). Excipients suitable for the NK cell formulation are known to the person skilled in the art, upon having read the present specification, for example water (especially water for inj ection), saline, Ringer ' s solution, dextrose solution, buffers, Hank solution, vesicle forming compounds (e. g. lipids), fixed oils, ethyl oleate, 5% dextrose in saline, substances that enhance isotonicity and chemical stability, buffers and preservatives. Other suitable excipients include any compound that is not harmful for the patient). Examples are well tolerable proteins, polysaccharides, polylactic acids, polyglycolic acid, polymeric amino acids and amino acid copolymers. This formulation can (as a drug) be administered via appropriate procedures known to the skilled person (upon having read the present specification) to a patient or individual in need thereof (i. e. a patient or individual having or having the risk of developing the diseases or conditions mentioned herein). The preferred route of administration of said NK cell formulation is parenteral administration, in particular through intraperitoneal, subcutaneous, intramuscular and / or intravenous administration. For parenteral administration, the NK cell formulation of the present invention is preferably provided in inj ectable dosage unit form, e. g. as a solution (typically as an aqueous solution), suspension or emulsion, formulated in conjunction with the above-defined pharmaceutically acceptable excipients. The dosage and method of administration, however, depends on the individual patient or individual to be treated. The NK cell formulation can be administered in any suitable dosage, e. g., as disclose above,or known from other biological dosage regimens or specifically evaluated and optimised for a given individual. The administration may occur e. g., a single time or multiple times (such as two or three times) e. g., within a period of a month, 3 months, 6 months, 9 months or 12 months. As the preferred mode of administration of the inventive NK cell formulation is parenteral administration, the NK cell formulation according to the present invention is preferably liquid.
[0083] The patient to be treated (i. e., the patient having the cancer, in particular solid malignancy as disclosed herein) is preferably human.
[0084] In the course of the present invention, the protein FAM13B surprisingly turned out to be a marker for recovering cytotoxicity after the expansion phase. According to another preferred embodiment, the NK cell formulation (in particular as obtainable by the inventive method) thus comprises NK cells expressing the protein FAM13B (especially human FAM13B, UniProt: Q9NYF5), in addition to being transiently enriched in FasL on the surface. In this context, the term "express" means that FAM13B protein is detectably present within the NK cells at levels above an appropriate negative control or reference population. Preferably, the total expression level (normalized to VCC) is at least 50%, preferably at least 100%, even more preferably at least 200% higher than the total expression level (normalized to VCC) of the control, i. e. relatively strong overexpression occurs.
[0085] In a further preferred embodiment, to enhance the resilience of the NK cells against tumour-mediated FasL counterattacks, expression of Fas receptor is (transiently) downregulated in the NK cells. The transient downregulation of FasR aims to render the NK cells less susceptible to this FasL-mediated apoptosis. This reduction in FasR expression may be achieved using various established molecular biology techniques, forinstance, by introducing short interfering RNA (siRNA) or short hairpin RNA (shRNA) constructs specifically designed to target the mRNA encoding FasR, leading to its degradation or translational inhibition. The step of transiently downregulating FasR may, for instance, be performed, for example, concurrently with the transient enrichment of FasL, or shortly before providing the NK cells in the final formulation.
[0086] In yet another preferred embodiment, to protect the NK cells from apoptosis induced by death ligands expressed by tumour cells, an anti-apoptotic protein, such as Fas-associated death domain-like interleukin-l-beta-converting enzyme-inhibitory (FLIP) protein, is (transiently) overexpressed within the NK cells. FLIP is discussed e. g., in Screpanti et al, 2001; human FLIP, which is preferred, has UniProt accession code 015519. By increasing the intracellular levels of anti-apoptotic proteins such as FLIP, the apoptotic signalling cascade downstream of death receptor activation can be effectively blocked or significantly attenuated. The transient overexpression can be accomplished by introducing a suitable transient expression system into the NK cells, such as those previously described for FasL overexpression. In this context, the term "express" means that the anti-apoptotic protein such as FLIP is detectably present within the NK cells at levels above an appropriate negative control or reference population. Preferably, the total expression level (normalized to VCC) is at least 50%, preferably at least 100%, even more preferably at least 200% higher than the total expression level (normalized to VCC) of the control, i. e. relatively strong overexpression occurs.
[0087] In another preferred embodiment, the NK cells provided in the inventive formulation (or utilised in the inventive method), are wildtype or native with regard to the Programmed Death 1- 1 - (PD-1 ) gene (Ensembl Gene ID ENSG00000276977 in humans;
[0088] UniProt entry Q15116). In particular, this means that the NK cells have not undergone any intentional genetic modification specifically targeting the PD-1 locus, such as gene knockout (e. g., using CRISPR / Cas9) designed to ablate or permanently diminish the expression or function of the endogenous PD-1 protein. In other words, the NK cells employed in this embodiment possess their native PD-1 gene and preferably also its associated regulatory elements. Any expression of PD-1 on the surface of these NK cells will therefore be governed by natural cellular regulatory mechanisms and the microenvironment, rather than being artificially eliminated or permanently suppressed through genetic engineering.
[0089] In another preferred embodiment, the NK cells provided in the inventive formulation (or utilised in the inventive method), are wildtype or native with regard to the FasL protein sequence. In particular, this means that the NK cells do not harbour a FasL mutant sequence, let alone express any FasL protein mutant (but express the endogenous FasL sequence, e. g., human FasL in case of human NK cells).
[0090] In another preferred embodiment, the cytoplasmic domain of the (transiently overexpressed) FasL in the NK cells is native, in particular having the sequence identified by UniProt P48023, positions 1-80.
[0091] In another, particularly preferred embodiment, the NK cells provided in the inventive formulation or produced by the inventive method do not comprise a recombinant or transgenic antigen-binding protein, such as a chimeric antigen receptor (CAR). The NK cells are preferably native in this regard, with their antigen recognition mechanisms relying solely on their endogenous activating and inhibitory receptors, such as NKG2D, NKp44, NKp46, CD16 (FcyRIII ), and the killer immunoglobulin-like receptors (KIRs), which are naturally expressed in NK cells. The present inventive approach preferably leverages the innate antigen recognition capabilities of native NK cells. This native recognition functionality is further augmented by the transient FasL enrichment, which specifically addresses the immunosuppressive challenges of the TME without compromising natural recognition. This also reduces safety risk, as it avoids the complexities associated with permanent genetic alterations (or, in embodiments, even further transient alterations).
[0092] In another preferred embodiment, the NK cell formulation comprises an NK cell dose of at least 107, preferably at least 108, especially at least 109VCC. Alternatively, or in addition thereto, the NK cell formulation exhibits a cytotoxic capacity of at least 106, preferably at least 107, especially at least 108. High NK cell doses and high cytotoxic capacity generally correlate with improved clinical response in patients.
[0093] The inventive method is preferably conducted under good manufacturing practice (GMP) conditions and / or the inventive formulation has preferably been produced under GMP conditions.
[0094] The invention further relates to the therapeutic application of the disclosed NK cell formulations. In particular, the NK cell formulation may be for use in cancer therapy (in a patient, in particular human patient).
[0095] Given the challenges posed by the microenvironment of solid tumours, specifically high lactate levels, the NK cell formulation is advantageously provided for use in the treatment of a solid malignancy (of a patient, in particular human patient). The transient enrichment of FasL aims specifically at overcoming such challenges, making the formulation particularly well-suited for targeting solid
[0096] tumours where conventional NK cell therapies have often shown limited efficacy.
[0097] In a preferred embodiment, the solid malignancy to be treated (in the patient, in particular the human patient) is selected from the group consisting of breast cancer (in particular triple-negative breast cancer), colorectal cancer, glioma (in particular glioblastoma such as glioblastoma multiforme), head and neck squamous cell carcinoma, lung cancer (in particular non-small cell lung cancer), melanoma, ovarian cancer, cervical cancer, pancreatic cancer, prostate cancer, and sarcoma, in particular glioblastoma, pancreatic cancer such as pancreatic ductal adenocarcinoma and (triple-negative breast) cancer, especially glioblastoma. These represent dif f icult-to-treat solid tumour types which frequently have elevated lactate levels in the TME (cf. de la Cruz-Lopez et al, 2019, in particular Table 1 and Gu et al, 2025; see also Böhme et al, 2016), where the enhanced NK cells of the invention can provide significant benefit.
[0098] The inventive therapeutic use is particularly relevant for advanced stages of cancer in which lactate levels are generally higher. Thus, in a further preferred embodiment, the malignancy being treated is metastatic. In another preferred embodiment, the malignancy is unresectable.
[0099] Further along these lines, in another preferred embodiment, the solid malignancy targeted for treatment is of the highest grade (e. g., Grade III or IV, depending on the specific cancer type and grading system). These high-grade tumours are often associated with high lactate levels (cf. de la Cruz-López et al, 2019).
[0100] By way of example, breast cancers are graded using the Nottingham Grading System, where Grade 3 represents the highest grade (poorly differentiated); triple-negative breastcancer frequently presents as Grade 3. Colorectal cancer grading describes differentiation, with Grade 3 (poorly differentiated) being understood as the highest grade. Gliomas are graded using the WHO classification, with Grade IV being the highest; glioblastoma is, by definition, a Grade IV glioma. Head and neck squamous cell carcinomas are graded based on differentiation, with Grade 3 (poorly differentiated) as the highest. Similarly, non-small cell lung cancer is graded based on differentiation, with Grade 3 (poorly differentiated) being understood as the highest grade. Ovarian and cervical cancers are graded 1 to 3 based on differentiation and architectural features, with Grade 3 being the highest. Pancreatic cancer, particularly pancreatic ductal adenocarcinoma, is graded based on glandular differentiation, with Grade 3 (poorly differentiated) being understood as the highest. Prostate cancer utilises the Gleason system, reported via Grade Groups 1 to 5, where Grade Group 5 (including Gleason scores 9-10) is the highest grade. Sarcomas are graded using the FNCLCC system (Grade 1, 2, or 3), with Grade 3 indicating the highest grade.
[0101] Furthermore, the NK cell formulation may advantageously be used in patients whose disease has become resistant to standard therapies. Accordingly, in a preferred embodiment, the malignancy is refractory to prior treatment, such as therapy with immune checkpoint inhibitors (in particular anti-PD1, anti-PD-Ll, or anti-CTLA-4 antibodies) and / or refractory to a targeted therapy (e. g., kinase inhibitors). Providing an effective treatment option for these refractory patients addresses a critical unmet medical need.
[0102] In a particularly preferred embodiment of the therapeutic use, the solid malignancy is characterised by a very high lactate concentration within its microenvironment. Specifically, the lactate concentration is at least 5 µmol / g tissue, preferablyat least 6 µmol / g, more preferably at least 7 µmol / g, even more preferably at least 8 µmol / g, yet even more preferably at least 9 µmol / g or even at least 10 µmol / g. Alternatively or additionally, the lactate concentration may be measured in interstitial fluid or other relevant samples and is at least 5 mM, preferably at least 6 mM, more preferably at least 7 mM, even more preferably at least 8 mM, yet even more preferably at least 9 mM or even at least 10 mM. Targeting tumours exhibiting such elevated lactate levels is highly advantageous because the transiently FasL-enriched NK cells are specifically designed to resist lactate-induced dysfunction.
[0103] For instance, lactate in resected or biopsied tumours may be quantified by direct biochemical assays in which the fresh-frozen specimen is homogenized, de-proteinized and subjected to an enzymatic lactate oxidase or lactate dehydrogenase reaction, with detection by spectrophotometry or chromatography. An early overview is given e. g., in Walenta et al, 2004. This provides absolute concentrations in micromoles per gram of tissue but necessarily averages heterogeneous regions. Nuclear magnetic resonance techniques complement this ex vivo approach by analysing intact tumour tissue. Another technique is intratumoral microdialysis, which places a semipermeable catheter into the tumour in vivo, continuously sampling interstitial fluid so that enzymatic assays can monitor lactate dynamics in real time in the immediate extracellular milieu (see Schroeder et al., 2013).
[0104] Herein, entries in the UniProt databases are identified by their accession codes (referred to herein, e. g. as " UniProt accession code" or briefly as " UniProt" followed by the accession code), usually a code of six alphanumeric letters. If not specified otherwise, the accession codes used herein refer to entries in the Protein Knowledgebase (UniProtKB) of UniProt. If not stated otherwise, the UniProt database statefor all entries referenced herein is as of 5 February 2025 (UniProt / UniProtKB Release 2025_01 ).
[0105] As used herein, "cytotoxicity" refers to the percentage reduction in the number of viable target cells following coculture with NK cells compared to control cultures lacking NK cells. Preferably, cytotoxicity is quantified by assessing the loss of viability of K562 cells after a co-culture with NK cells. According to a particular preference, "cytotoxicity" is determined by performing the method disclosed in the example and using equation 6 therein for calculation.
[0106] As used herein, "cytotoxic capacity" refers to a measure of the total cytolytic potential of an NK cell population. It is calculated as the product of the cytotoxicity percentage (as defined above) and the total number of viable NK cells in the culture. This parameter provides an absolute measure of the overall killing potential of the NK cell population. According to a particular preference, "cytotoxic capacity" is determined by performing the method disclosed in the example and using equation 7 therein for calculation.
[0107] In the context of the present invention, the term "native" when referring to proteins or protein sequences such as FasL, is defined as meaning an autologous sequence with respect to the NK cells being used in the formulation or method.
[0108] Specifically, "native" indicates that the amino acid sequence is identical to that naturally encoded by the genome of the organism from which the NK cells are derived, without any engineered mutations, substitutions, deletions, or insertions that would alter its primary amino acid sequence from what would be endogenously expressed. For example, if human NK cells are employed in the therapeutic formulation, then native FasL would correspond to the human FasL sequence; similarly, if murine NK cells were used (e. g., in preclinical studies),native FasL would refer to the murine FasL sequence. Native FasL may exist in either soluble or membrane-bound forms, both of which arise from the same gene through post-translational processing rather than from distinct genetic sequences.
[0109] However, the membrane-bound (in particular full-length) form is particularly preferred in the context of the invention, as it represents the most potent inducer of Fas receptor-mediated signalling. The full-length form includes the complete protein as initially translated, comprising the intracellular domain, transmembrane region, and extracellular domain, prior to any proteolytic cleavage that might generate soluble variants.
[0110] For the purposes of the present invention, GMP preferably refers to the European Union' s guidelines as defined in Directive 2003 / 94 / EC. Alternatively, GMP may preferably be understood according to the United States Food and Drug Administration (FDA) regulations codified in Title 21 of the Code of Federal Regulations (CFR), Part 211, which govern good manufacturing practice for finished pharmaceuticals.
[0111] Alternatively, or in addition thereto, GMP may refer to the European Medicines Agency (EMA) guideline on human cell-based medicinal products (EMEA / CHMP / 410869 / 2006 dated 21 May 2008 ) or the FDA' s Current Good Tissue Practice (CGTP) and Additional Requirements for Manufacturers of Human Cells, Tissues, and Cellular and Tissue-Based Products (HCT / Ps) dated December 2011.
[0112] The present invention further relates to the following embodiments:
[0113] Embodiment 1. An in-vitro method for producing natural killer (NK) cell formulations for cancer therapy, comprising at least the following steps:
[0114] - expanding at least one NK cell in a first culturing step,- transiently enriching Fas ligand (FasL) on the surface of the expanded NK cells, and
[0115] - providing the expanded and transiently FasL-enriched NK cells in an NK cell formulation for cancer therapy.
[0116] Embodiment 2. The method of embodiment 1, wherein said transiently enriching FasL comprises decorating the surface of the expanded NK cells with exogenous FasL, preferably wherein the FasL protein sequence is the native FasL sequence, in particular wherein both the NK cells and FasL protein sequence are human.
[0117] Embodiment 3. The method of embodiment 1 or 2, wherein said transiently enriching FasL comprises transient overexpression of FasL in the expanded NK cells, preferably wherein the transient overexpression is started only after said expanding step, and / or preferably wherein the FasL protein sequence is the native FasL sequence, in particular wherein both the NK cells and FasL protein sequence are human.
[0118] Embodiment 4. The method of embodiment 3, wherein a transient expression system is introduced to achieve transient FasL overexpression in the expanded NK cells, preferably wherein the transient expression system is introduced into the NK cells by electroporation, virus, virus-like particle or lipid particles.
[0119] Embodiment 5. The method of embodiment 4, wherein the transient expression system is selected from the group consisting of non-chromosomal DNA vectors such as plasmids, and exogenous RNA constructs such as exogenous messenger RNA (mRNA), in particular exogenous mRNA.
[0120] Embodiment 6. The method of embodiment 4 or 5, wherein the transient expression system is introduced after the expanding step.
[0121] Embodiment 7. The method of any one of embodiments 1 to 6, wherein FasL and / or Fas receptor expression is transiently downregulated during the expanding step.
[0122] Embodiment 8. The method of any one of embodiments 1 to 7, wherein the FasL / Fas receptor interaction is inhibited during the expanding step.
[0123] Embodiment 9. The method of any one of embodiments 1 to 8, wherein the CD56brightNK cell subpopulation is selectively enriched before and / or during the expanding step, preferably wherein the CD56dimNK cell subpopulation is selectively enriched thereafter (in particular before the transiently enriching step); or wherein the CD56dimNK cell subpopulation is selectively enriched after the expanding step, in particular before the transiently enriching step.
[0124] Embodiment 10. The method of any one of embodiments 1 to 9, further comprising storing the NK cell formulation at a temperature between 0° and 10°C and / or freezing the NK cell formulation.
[0125] Embodiment 11. The method of any one of embodiments 1 to 10, wherein the first culturing step is a perfusion, fed-batch or batch process; preferably wherein the first culturing step is a perfusion or fed-batch process, and / or wherein the lactate concentration during the first culturing exceeds a threshold of 5 mM, preferably 7.5 mM, more preferably 10 mM or even 12.5 mM, yet more preferably 15 mM or even 17.5 mM, especially 20 mM.
[0126] Embodiment 12. The method of any one of embodiments 1 to 11, wherein, in the expanding step, NK cells are expanded to a viable cell count (VCC) of at least 107, preferably at least 108, especially at least 109.
[0127] Embodiment 13. The method of any one of embodiments 1 to 12, the total VCC of the formulation is at least 107, preferably at least 108, especially at least 109and / or wherein cytotoxic capacity of the formulation is at least 106, preferably at least 107, especially at least 108.
[0128] Embodiment 14. The method of any one of embodiments 1 to 13, wherein the viable cell count (VCC) of NK cells in the formulation is at least 70%, preferably at least 80%, more preferably at least 90%, even more preferably at least 95% or even at least 97%, especially at least 98% or even at least 99% of the total VCC of the formulation.
[0129] Embodiment 15. An NK cell formulation for cancer therapy, obtainable according to the method of any one of embodiments 1 to 14, preferably wherein FasL is transiently enriched on the surface of NK cells.
[0130] Embodiment 16. An NK cell formulation for cancer therapy, wherein FasL is transiently enriched on the surface of NK cells.
[0131] Embodiment 17. The NK cell formulation of embodiment 15 or 16, wherein the surface of NK cells is decorated with exogenous FasL, preferably wherein the FasL protein sequence is the native FasL sequence.
[0132] Embodiment 18. The NK cell formulation of any one of embodiments 15 to 17, wherein NK cells transiently overexpress FasL, preferably wherein the FasL protein sequence is the native FasL sequence.
[0133] Embodiment 19. The NK cell formulation of any one of embodiments 15 to 18, wherein NK cells comprise a transient expression system for FasL overexpression, preferably wherein the transient expression system is selected from the group consisting of non-chromosomal DNA vectors such as plasmids,and exogenous RNA constructs such as exogenous messenger RNA (mRNA), in particular exogenous mRNA.
[0134] Embodiment 20. The NK cell formulation of any one of embodiments 15 to 19, wherein the VCC of NK cells in the formulation is at least 70%, preferably at least 80%, more preferably at least 90%, even more preferably at least 95% or even at least 97%, especially at least 98% or even at least 99% of the total VCC of the formulation.
[0135] Embodiment 21. The NK cell formulation of any one of embodiments 15 to 20, wherein the cell formulation comprises an NK cell dose of at least 107, preferably at least 108, especially at least 109VCC.
[0136] Embodiment 22. The NK cell formulation of any one of embodiments 15 to 21, wherein the cell formulation has a cytotoxic capacity of at least 106, preferably at least 107, especially at least 108.
[0137] Embodiment 23. The NK cell formulation of any one of embodiments 15 to 22, wherein FAM13B is expressed in the NK cells.
[0138] Embodiment 24. The NK cell formulation of any one of embodiments 15 to 23, wherein an antiapoptotic protein such as Fas-associated death domain-like interleukin-l-beta-converting enzyme-inhibitory protein (FLIP) is expressed, preferably transiently overexpressed, in the NK cells.
[0139] Embodiment 25. The NK cell formulation of any one of embodiments 15 to 24, wherein Fas receptor is downregulated, preferably transiently downregulated, in the NK cells.
[0140] Embodiment 26. The NK cell formulation of any one of embodiments 15 to 25, wherein FasL is native FasL protein, in particular wherein both the NK cells and the FasL protein sequence is human.
[0141] Embodiment 27. The NK cell formulation of any one of embodiments 15 to 26 for use in cancer therapy.
[0142] Embodiment 28. The NK cell formulation for use according to embodiment 27, for use in treatment of a solid malignancy.
[0143] Embodiment 29. The NK cell formulation for use according to embodiment 28, wherein the solid malignancy is selected from the group consisting of breast cancer, in particular triplenegative breast cancer, colorectal cancer, glioma, in particular glioblastoma, head and neck squamous cell carcinoma, lung cancer, in particular non-small cell lung cancer, melanoma, ovarian cancer, cervical cancer, pancreatic cancer, prostate cancer, and sarcoma, especially head and neck squamous cell carcinoma, lung cancer, in particular non-small cell lung cancer, melanoma, ovarian cancer, pancreatic cancer, prostate cancer, and sarcoma.
[0144] Embodiment 30. The NK cell formulation for use according to embodiment 28 or 29, wherein the malignancy is metastatic.
[0145] Embodiment 31. The NK cell formulation for use according to any one of embodiments 28 to 30, wherein the malignancy is unresectable.
[0146] Embodiment 32. The NK cell formulation for use according to any one of embodiments 28 to 31, wherein the malignancy is refractory to anti immune-checkpoint-inhibitor, in particular anti-programmed death 1 (PD1 ), anti-programmed death ligand 1 (PD-L1 ) or anti- cytotoxic T-lymphocyte-associated protein 4 (CTLA-4 ), treatment and / or refractory to a targeted therapy.
[0147] Embodiment 33. The NK cell formulation for use according to any one of embodiments 28 to 32, wherein the solid malignancy is of the highest grade.
[0148] Embodiment 34. The NK cell formulation for use according to any one of embodiments 28 to 33, wherein the solid malignancy hasa lactate concentration of at least 5 µmol / g, preferably at least 6 µmol / g, more preferably at least 7 µmol / g, even more preferably at least 8 µmol / g, yet even more preferably at least 9 µmol / g or even at least 10 µmol / g, or of at least 5 mM, preferably at least 6 mM, more preferably at least 7 mM, even more preferably at least 8 mM, yet even more preferably at least 9 mM or even at least 10 mM.
[0149] Figures and example
[0150] The present invention is further illustrated by the following figures and example, without being restricted thereto.
[0151] Figures 1A-1F. Total VCC, viability ( 1A), cytotoxicity, specific growth rate ( IB), cytotoxic capacity ( 1C), lactate concentration and production rate ( ID), glucose concentration and consumption rate ( IE), glutamine concentration and consumption rate ( IF) of the NK-92 cell batch expansion process in static T-flasks. Error bars represent the standard deviation of three, independent biological replicates.
[0152] Figure 2. Correlation analysis of lactate concentrations from small scale experiments and the pH. Spearman' s correlation shown as the R2of three different process modes (batch, fed-batch, and, as a special case of a batch process, repetitive batch) with three independent biological replicates.
[0153] Figures 3A-3C. Total viable cell count and viability (3A), cytotoxicity and lactate concentration (3B), and cytotoxic capacity and specific growth rate (3C) of NK92 cells during a fed-batch production process in a 2-liter working volume stirred-tank bioreactor. Error bars represent technical replicates.
[0154] Figures 4A-4C. Total VCC and viability (4A), cytotoxicity and lactate concentration (4B), and cytotoxic capacity and
[0155] specific growth rate (4C) of NK92 cells during the recovery phase following expansion in the same bioreactor are shown. Due to the limited working volume of the bioreactor, only 6.6% of all cells produced during the fed-batch phase could be retained for the recovery phase. Error bars represent technical replicates.
[0156] Figures 5A-5C. Total viable cell count and viability (5A), cytotoxicity and lactate concentration (5B), and cytotoxic capacity and specific growth rate (5C) of NK92 cells during a fed-batch production process in a 2-liter working volume stirred-tank bioreactor. Feed rate was controlled by a pH-auxostat. Error bars represent technical replicates.
[0157] Figures 6A-6F. Proteomic analysis of NK-92 cells harvested from small-scale batch format on day 3 upon exposure to L-Lactate. Lactate concentration (A), viability (B), cytotoxicity (C), specific growth rate (D), and the label-free quantification (LFQ) of the protein FAM13B (E) and (F) FasL (TNFL6), after 3 days of culture were compared using unpaired t test where p>0, 05 is ns, p<0.0332 is *, p<0.0021 is **, p<0.0002 is ***, and p<0.0001 is ****. Error bars represent the standard deviations of six independent biological replicates.
[0158] Figure 7. FasL supplementation assay. Supplementation with soluble FasLG protein (" FasLG") during cytotoxicity assay lead to significant increase of cytotoxicity, in particular when the NK cells had been pre-incubated with 15 mM lactate. E: T ratio: effector to target ratio. Measurements were compared using unpaired t test where p>0, 05 is ns, p<0.0332 is *, p<0.0021 is **, p<0.0002 is ***, and p<0.0001 is ****. Error bars represent the standard deviations of six independent biological replicates.
[0159] Materials and Methods
[0160] Small-scale cultivation
[0161] NK-92 cells (ACC 488, Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures GmbH, Braunschweig, Germany) were cultured at standard culturing conditions (37 °C and 5% CO2 in a humidified incubator), in a minimal essential medium alpha modification (aMEM) (Gibco, Paisley, United Kingdom) supplemented to a final concentration of 1.15 mM L-Arginine (Carl Roth, Karlsruhe, Germany), 2.31 mM L-Glutamine (Carl Roth, Karlsruhe, Germany), 0.286 mM L-Serine (Thermo Fisher Scientific, Waltham, USA), 0.194 mM i-inositol (Sigma-Aldrich, St. Louis, USA), 23.2 mM D-Glucose (Carl Roth, Karlsruhe, Germany), 24 pM beta-mercaptoethanol (Sigma-Aldrich, St. Louis), Insulin-Transf errin-Selenium (Gibco, Grand Island, NY, USA) 17.21 pM, 0. 69 pM and 0.39 pM, respectively, 500 lU / mL premium grade interleukin-2 ( IL-2 ) (Miltenyi Biotech, Bergisch Gladbach, Germany) and 10% type
[0162] AB, off-the-clot human serum (Pan Biotech, Aidenbach, Germany). A batch culture was seeded in T-175 flasks (Starlab, Brussels, Belgium) at a concentration of 2.5xl05cells / mL and was supplemented with IL-2 every second day without media exchange or addition.
[0163] To assess the cytotoxicity of the NK-92 cells, K-562-GFP (K562 ) cells (American Type Culture Collection, Manassas, USA, CCL-243-GFP) were maintained at standard culturing conditions in a Roswell Park Memorial Institute 1640 medium (RPMI ) (Gibco, Grand Island, NY, USA) containing 5% heat-inactivated fetal bovine serum (FBS) (Gibco, Thermo Fisher Scientific, Waltham, USA). Regular subculture of K562 cells was done at a seeding concentration of IxlO6cells / mL.
[0164] Primary human NK cells were sourced from buffy coat of healthy donors (Osterreichisches Rotes Kreuz, Vienna, Austria).
[0165] Peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation (Lymphoprep, STEMCELL Technologies, Vancouver, Canada) followed by negative magnetic isolation (Miltenyi Biotec, Bergisch-Gladbach, Germany). The enriched NK cell population was cultured in CTS NK Expander media (Thermo Fisher Scientific, Waltham, USA) supplemented with 5% type AB, off-the-clot human serum (Pan Biotech, Aidenbach, Germany). 500 TU / mL premium grade IL-2 (Miltenyi Biotech, Bergisch Gladbach, Germany) and 120 lU / mL premium grade interleukin 15 ( IL-15) (Miltenyi Biotech, Bergisch Gladbach, Germany) were supplemented every second day. On the day of isolation, degradable expansion prototype beads (Allegrow, Hong Kong) were added to the culture at a 1: 1 ratio of cells-to-beads when seeding the cells. NK cells were expanded for a period of 14 days, when 2xl06cells were seeded in 2mL of media containing OmM, lOmM, 20mM or 30mM L-Lactate, all at pH 4, in duplicates. Two days after seeding, one replicate per donor was analyzed for cytotoxicity, while the other replicate was re-seeded in fresh media without lactate. The re-seeded replicate was cultured for another 48 hours and analyzed for cytotoxicity afterwards.
[0166] Benchtop bioreactor cultivation
[0167] Bioreactor cultivations were conducted in 2-liter working volume UniVessel glass vessels (Sartorius, Gottingen, Germany) controlled by Labfors 5 ( Infers HT, Bottmingen, Switzerland) controllers. Each vessel was equipped with two, 30° 3-blade pitch-blade impellers, a microsparger for aeration, pH and optical dissolved oxygen (DO) sensors (both Hamilton, Bonaduz, Switzerland). The cultures were heated with external heating mats and the temperature monitoring was performed using the built-in PtlOO probe. The inoculum was expanded in 250 mLshake flasks (Corning, Corning, USA) on an orbital shaker under standard culturing conditions at 120 rpm. During preculture expansion NK-92 cells were seeded at 2.5xl05cells / mL, IL-2 was supplemented every second day, and medium was replaced every 4 days. The culture in the bioreactor was also seeded at 2.5xl05NK-92 cells / mL in 600 mL starting volume ( 80 rpm, 37 °C) with IL-2 supplementation every second day. A total gas flow rate of 12 mL / min ( 0.02 vessel volumes per minute) was achieved by a constant pressurized air flow of 2 mL / min, varying pure oxygen flow to maintain a dO level of 60%, constant pure CO2 flow of 0. 6 mL / min to sustain 5% CO2level and varying pure nitrogen flow to reach the total gas flow rate, all controlled by mass flow controllers. The offgas left the reactor through a chilled off-gas cooler and its O2, CO2 and humidity content were analyzed by a BluelnOne Cell gas analyzer (BlueSense, Herten, Germany). The pH was maintained at 7.40 (±0.05) with the addition of 0.5 M NaHCO3, delivered through the integrated pump system of the controller in the case of non-automated feeding. The batch phase was terminated in non-automated cultures, when the lactate concentration exceeded a threshold of 12 mM by feeding a total of 1000 mL fresh medium in an exponential feed regime (cf. Li et al, 2015), using PreciFlow (Lambda Instruments, Baar, Switzerland) peristaltic pumps, with the goal to maintain a lactate concentration below this threshold. The initial federate was determined based on the lactate concentration (Equation 1 and Equation 2 ) and an exponential federate was set according to the growth rate of the cells (Equation 3 and Equation 4 ). The culture was maintained after all feed was added to the culture, until no further growth was observed (cf. Costariol et al, 2019). Automated pH auxostat based feeding was performed by maintaining the pH at 7.40 (±0.05) with the addition fresh culture medium. A total of 1000 mLfresh medium was fed into the vessel, and the culture was maintained after all feed was added to the culture, until no further growth was observed. This correlated with a lactate level of about 20 mM (see Results and Discussion) and concluded the expansion phase.
[0168] After the culture ended, cells were separated from culture medium by centrifugation at 300xg for 5 min at room temperature, and the recovery phase was initiated by
[0169] re-seeding a fraction ( 6%) of the cells in 90% (v / v) of fresh medium at a seeding density of 2.5xl05cells / mL in a total volume of 600 mL. For the recovery phase, the bioreactor was maintained at 37.2 °C, 5% CO2in the off-gas, 52.86% DO, and a pH of 7.4, with a stirring speed of 80 rpm and a total gas flow rate of 12 mL / min, as the predicted optimum of the previous DoE analysis in T-Flasks. Throughout the culture and the cytotoxicity recovery process, process parameters (DPs) were monitored and recorded by the Lucullus process information management system (Securecell, Urdorf, Switzerland).
[0170]
[0171] reactor volume (Vreactor ) •
[0172]
[0173] Equation 2 Calculation of the current lactate concentration (cLactate ) r determined as the product of the current lactate concentration (cLactate) and the reactor volume at the time ofsampling (Vreactor), and the specific lactate production rate (^Lactate ) multiplied by the total viable cell count.
[0174]
[0175] Equation 3 Calculation of the specific growth rate g where Nt is the total viable cell count at the time, and Nto is the initial viable cell count. Ti is the current time point, and To is the time point of seeding.
[0176] F(t) = Fo■ e^
[0177] Equation 4 Calculation of the exponential feed rate based on the initial feeding volume multiplied by the exponent of the specific growth rate and time.
[0178] Process analytics
[0179] Cell counting and viability
[0180] Cell count and viability were measured by flow cytometry on a CytoFLEX S (Beckman Coulter, Brea, USA). 150 pL of cell suspension was mixed with 50 pL of cell counting beads ( Invitrogen, Thermo Fisher Scientific, Waltham, USA) and 50 pL of 5 pg / mL 7-AAD ( Invitrogen, Thermo Fisher Scientific, Waltham, USA) fluorescent stain for live-dead cell discrimination. The viable cell concentration was calculated as described in Equation 5.
[0181] cells'
[0182] Viable cell concentration
[0183] . pL.
[0184] (7A4D event count x counting bead volume)
[0185] - - - -; -; - - -; - - - x counting bead concentration counting bead count x cell volume)
[0186] Equation 5 Formula for the calculation of the absolute cell count as instructed by the manufacturer of the counting beads. Cell count and bead count are determined by the respective event count in the gates by flow cytometry.
[0187] Size calibration beads for flow cytometry ( Invitrogen, Thermo Fisher Scientific, Waltham, USA) were used to perform a size calibration that enabled the gating of particles based on the beads' reference size. This calibration step minimized background signals in live-dead cell discrimination, which primarily resulted from protein aggregation in the medium, by excluding particles smaller than 8 pm, which is approximately half the size of an NK-92 cell. Afterwards, viable cells were gated with the side-scatter signal, including only 7-AAD-events. The bead count was gated according to the manufacturer ' s protocol in a fluorescent channel ( 660110 nm detector) unaffected by other stains used in the analysis.
[0188] Cytotoxicity and cytotoxic capacity determination
[0189] Cytotoxicity of NK-92 cells was assessed by co-culturing them with K562 cells for a duration of 4 hours, at an E: T ratio of 1: 1. Both cell types were seeded in triplicates at a concentration of IxlO5cells / well in 96-well V-bottom plates (Thermo Fisher Scientific, Waltham, USA) in 200 µL aMEM medium without serum and IL-2. Additionally, one negative control of both cell types containing the same cell amount
[0190] (2xl05cells / well) was seeded to assess spontaneous death.
[0191] After 4 hours, the cells were collected by centrifugation (300 x g, 5 min. ), washed in a FACS buffer (containing PBS + 1% bovine serum albumin (Carl Roth, Karlsruhe, Germany) + 50 pM NaNs (Thermo Fisher Scientific, Waltham, USA) ), collected again (300 x g, 5 min. ), then stained with Annexin-V (BD Biosciences, Franklin Lakes, USA) for apoptosis. The reaction was carried out by adding 100 pL annexin binding buffer (ABB) ( Invitrogen, Thermo Fisher Scientific, Waltham, USA) to eachwell, (which contained 0. 625 pL Annexin-V) and incubated for 15 minutes at room temperature in the dark. Afterwards 150 pL ABB was added and cells were collected by centrifugation.
[0192] After removal of the supernatant cells were resuspended in 250 pL 7-AAD in ABB to a final concentration of 1 pg / mL for the second staining step, where Annexin+ 7-AAD-events indicated early apoptotic cells, Annexin+ 7-AAD+ events indicated late apoptotic cells, and Annexin- 7-AAD+ events indicated necrotic cells. After the second staining, the cells were immediately analyzed by flow cytometry. For each measurement, a minimum of 10, 000 events were recorded in the gate of live K562 or NK-92 cells, respectively.
[0193] Cytotoxicity was calculated by quantifying the loss of viable target cells in the co-culture with NK-92 cells, compared to the negative controls (Equation 6); see also Penney & Clinton, 2022. Therefore, the count of early apoptotic, late apoptotic, and necrotic K562 cells were used from the co-culture sample and the negative control sample.
[0194] (100% - live K562 [%]) - (100% - live K 62control [%]) Cytotoxicity [%] = - - - - 77-: - xlOO live K 62control [%]
[0195] Equation 6 Formula for the determination of absolute cytotoxicity from flow cytometry measurements where "live K562" are 7-AAD and Annexin-V double-negative effector events in co-culture wells, and "live K562COntroi" are the 7-AAD and Annexin-V double-negative effector cells in the control wells.
[0196] In addition, the cytotoxic capacity was calculated to describe the potential of the whole culture to kill a target cell within 4 hours of co-culture (Equation 7 ).
[0197] Cytotoxic capacity [cells] = cytotoxicity [%] x total viable cells [cells ]
[0198] Equation 7 Formula for the calculation of the cytotoxic capacity of the culture, where the cytotoxicity is multipliedby the total viable cell count of the culture. The result is a number of cells that have the capacity to eliminate one target cell.
[0199] Medium analysis
[0200] From the spent medium, the concentration of glucose, lactate, glutamine, and ammonia was determined by a CEDEX BioHT analyzer (Roche, Basel, Switzerland). The pH was also measured but only after re-equilibrating the spent medium in a 5% C02incubator environment for at least 1 hour to minimize pH shifts caused by the degassing of C02. Afterwards, the pH was measured immediately after removing the sample from the incubator, using a freshly-calibrated pH sensor (Xylem Analytics, San Diego, USA).
[0201] Protein analysis
[0202] Static small-scale cultivation
[0203] Analysis of proteins in NK-92 cells upon exposure to lactate above 20mM was performed in small scale batch experiments. Cells were seeded in T-75 flasks (Starlab, Brussels, Belgium) under standard culturing conditions at a concentration of 2.5xl05cells / mL and were supplemented with IL-2 every second day without media exchange. The initial L-lactate concentration was adjusted to 0.915 mM and 23.440 mM in the conditions without lactate and with lactate, respectively. The pH was adjusted to 7.4 in both conditions. Following culture, cells were analyzed for growth, metabolite production, and cytotoxicity. From each replicate 4.5xl06cells were frozen at -80°C without any buffer after washing two times in PBS.
[0204] Sample preparation
[0205] Cells were lysed in 100 mM Tris-HCl pH=8.5 containing 1 % SDS, 10 mM TCEP ( Tris ( 2-carboxyethyl ) phosphine ) and 40 mM
[0206] chloroacetamide using probe sonication. Then samples were reduced and alkylated for 10 min at 95 °C and centrifuged at 7000 xg for 5 min at room temperature (RT) to remove cell debris followed by protein estimation using Pierce™ BCA Protein Assay. 100 pg each was acetone precipitated, dissolved with 25 % 2, 2, 2-Trif luoroethanol in 75 mM Tris-HCl pH=8.5 and digested by adding Promega Trypsin at a ratio of 50: 1 and shaking overnight with 550 rpm at 37 °C. 4 pg of the resulting peptide solution was acidified to a final concentration of 1 % Trifluoroacetic acid and desalted using in-house made stagetips with SDB-RPS (Styrene Divinylbenzene - Reversed Phase Sulfonate) as material.
[0207] LC-MS / MS analysis
[0208] Peptides were separated with the UltiMate 3000 RSLCnano Dionex system (ThermoFisher Scientific, MA, USA) using an Tonopticks Aurora Series UHPLC C18 column (250 mm x 75 pm, 1. 6 pm) ( Tonopticks, Australia). LC-MS / MS run per sample was 86.5 min with a flow rate of 400 nl / min at 40°C, using solvent A: 0.1 % formic acid in water; solvent B: acetonitrile with 0.1 % formic acid, with the following gradient: 0-5.5 min: 2 % B; 5.5-25.5 min: 2-10 % B; 25.5-45.5 min: 10-25 % B, 45.5-55.5 min: 25-37 % B, 55.5-65.5 min: 37-80 % B, 65.5-75.5 min: 80 % B; 75.5-76.5 min: 80-2% B; 76.5-86.5 min: 2 % B. The timsTOF HT mass spectrometer (Bruker Daltonics, Germany) was operated as follows: positive mode, enabled trapped ion mobility spectrometry (TIMS), 100% duty cycle (ramp 100 ms); source capillary voltage (V): 1600 V; dry gas flow: 3 L / min, 180 °C; scan mode: data independent parallel accumulation-serial fragmentation (diaPASEF), previously described by Meier et al., 2020, using 21 x 25 Th isolation windows, m / z 475 -1, 000; 0 Th overlap between windows. Two and three isolation windows were fragmented per TIMS ramp after the MSI scan, respectively (overall DIA cycle time: 0.95 s).
[0209] Data processing
[0210] Raw data files were analysed and proteins were quantified using DIA-NN software (version 2.0), see Demichev et al, 2020 and Demichev et al, 2022. The SwissProt human proteome database in fasta format (containing common contaminants; downloaded on 12.02.2025, 52, 410 sequences) was used for a library-free search with FDR set to 1 %. Deep learning-based spectra and retention time prediction was enabled, minimum fragment m / z was set to 200 and max fragment m / z set to 1800. N-terminal methionine excision and cysteine carbamidomethylation was enabled and maximum number of trypsin missed cleavages set to 2. Minimum peptide length was set to 7 and the maximum to 30 amino acids. Cysteine carbamidomethylation was set as a fixed and methionine oxidation as a variable modification. DIA-NN optimised the mass accuracy automatically using the first run in the experiment. Data processing using protein group quantities was done with Perseus software version 2.0.10 (see Tyanova et al, 2016). Intensities were log2 transformed in order to lower the effect of the outlier values and filtered for 100 % valid values. A two-sample t-test with subsequent multiple testing correction by Permutation-based FDR method was used to identify significantly changed proteins.
[0211] Data analysis
[0212] Time-resolved analysis of the culture data from batch expansion was performed by combining three independent biological replicates and presenting their standard deviation. The data obtained from the fed-batch expansion in stirred-tank bioreactors were shown with standard deviations of three technical replicates. Protein analysis of the effect from lactate on NK-92 cells was performed with six independentbiological replicates per lactate condition. Statistical analysis was performed in GraphPad Prism (version 10.4.0 for Windows, GraphPad Software, Boston, Massachusetts USA). A correlation analysis was performed using Spearman' s correlation. For comparison of two datasets, an ordinary t-test was chosen. The following levels of significance were assigned if the p-values were less than 0.0332 (* ), 0.0021 (** ), 0.0002 (*** ), and 0.0001 (****) • Bar and line charts were created with the GraphPad Prism software.
[0213] FasL supplementation assay
[0214] NK-92 cells were cultured under standard conditions as previously described. Cells were maintained in two independent static T-Flasks for a duration of three days. One flask was supplemented with an initial L-Lactate concentration of 15 mM, while the other flask was not supplemented with L-Lactate. The pH in both flasks was adjusted to 7.4. Following the three-day pre-culture period, cytotoxicity was assessed under both conditions using increasing concentrations of soluble recombinant Fas Ligand (sFasL) protein ( Peprotech, Cranbury, USA). The final concentrations of sFasL in the co-culture wells were set at 0, 1, 5, 10, 20, 40, 80, 120, and 160 ng / mL, with a total volume of 200 µL per well. The cytotoxicity assay was conducted as described above to evaluate the effect of sFasL on NK-92 cells under the specified conditions.
[0215] Results & Discussion
[0216] Batch culture
[0217] Cultivating NK-92 cells in a typical batch fashion for 8 days resulted in a classical growth curve, with the four
[0218] distinctive (lag, exponential, plateau, decline) phases of growth (Figure 1A). Viability stabilized around 90% during exponential growth after adaptation to the culture conditions (Figure 1A) but it decreased rapidly during plateau and decline phase of the culture. Cytotoxicity showed a non-linear behavior and it reached its maximum of 43.7 ± 2.92% on day 2, following an adaptation to the new culture conditions, just at the start of the exponential growth phase (Figure 1B). It remained stable for two days and started to rapidly decline from day 4 onwards to reach its minimum of 1.23 ± 0.08% on day 8. The growth rate strongly increased after just 1 day of lag phase and stayed close to its maximum of 0.02 ± 0.003h-1 until day 5 and then decreased continuously until day 7 when it turned into death until the end of the culture, following the trend of viability (Figure 1A, 1B). Cytotoxic capacity increased until day 5, up to 9.98x107 ± 1.98x107 cytotoxic cells at the intersection of high cell count and high cytotoxicity, followed by a rapid drop to only 4.07x106 ± 2.2x105 cytotoxic cells on day 8, driven mostly by the fall in cytotoxicity (Figure 1C).
[0219] The purpose to study the small-scale batch expansion of NK-92 cells was to better understand the expansion process and determine potential dynamics of cytotoxicity. The results demonstrated that despite the expected growth behavior of the cells, cytotoxicity showed unexpected dynamics that was so far not reported. Similarly, cytotoxic capacity also showed a nonlinear trend. These findings are particularly important for cell therapy production, where the cell itself is the product, thus the ideal point of harvest is not solely defined by the correct number of viable cells but also by a high cytotoxicity, hence the cytotoxic capacity. Unlike as in current laboratory practice that frequently relies only on endpoint-measurements of CQAs, we performed a time-resolvedprocess analysis of CQAs, which included VCC, cytotoxicity and cytotoxic capacity. Growth and cytotoxic phenotypes were decoupled. The dynamic changes of cytotoxicity occurred one day prior to the cells' growth rate response, as cytotoxicity increased earlier than the growth rate. Later in the expansion process it also decreased faster than the growth rate. If a time resolved analysis of cytotoxicity and cytotoxic capacity would not have been performed, the point of harvest would have been determined to be on day 7 of the culture, the peak value of the VCC. However, it was evident by this time that cytotoxicity and cytotoxic capacity were lost entirely in this particular setup. These observations demonstrated the importance of a closer monitoring of all CQAs to design a process that robustly produces a high-quality product.
[0220] However, due to the unexpected high dynamics of cytotoxicity and cytotoxic capacity, it was hypothesized that recovery of the lost cytotoxicity potential would be possible by the selection of the correct PPs.
[0221] Lactate production, as well as glucose and glutamine consumption exhibited a correlation with cell growth (Figure 1D-F). Lactate was produced continuously at specific production rates of up to 79.1 ± 4.4 pmol / cell / day until day 5, accumulating to almost 30 mM by the end of culture, followed by a stop of production due to reduced VCC and stop of cell growth from day 7 onwards. Glucose was consumed until day 7, not taken up anymore after cell growth terminated, while glutamine displayed a similar pattern, but overall a more linear consumption. Since glucose was still present in 10 mM concentration, the depletion of glutamine on day 7 seemed to cause a limitation that prevented the cells to grow further. Seemingly, lactate accumulation did not impact proliferation, however considering that cell growth was already almost zero by day 6 ( 0.005 ± 0.004h-l ), when glucoseand glutamine were not depleted yet, it is most probable that lactate accumulation above 20 mM inhibited growth. However, our goal was not only to produce more but also highly functional cells. Thus, it turned out that an additional culturing step was advantageous when using fed-batch in the expansion step, with the aim to recover the lost cytotoxicity.
[0222] Lactate was identified as the effector that inhibited cytotoxicity and growth during small-scale batch expansion. It was showcased in our study for the first time, that the same PP was responsible for the limitation of two, decoupled QAs. This new observation is important for the production process of cell therapies, as it defined lactate concentration as a critical process parameter (CPP) for cytotoxicity and growth, independently. Further, the preferred ranges of lactate for higher cytotoxicity ( 0-10 mM) and higher growth rates ( 0-15 mM) in the present setup were now defined, which helped bring robustness to the cell expansion process. Interestingly, neither glucose or glutamine, commonly discussed as the limiting factor in cell culture media, were limited throughout the entire growth phase. With the definition of cytotoxic capacity, a metric that combines the VCC and the cells' cytotoxicity, a valuable tool to determine the ideal timepoint of harvest was constructed.
[0223] pH-lactate correlation
[0224] The relationship between culture media pH and lactate concentration was analyzed using data from three process modes— batch, fed-batch, and repetitive batch (as a special case of a batch process) —each with three independent biological replicates in small-scale static cultures. The correlation plot (Figure 2 ) revealed a strong negative relationship, with a Spearman' s correlation coefficient of R2= 0.94, indicating that as lactate concentration increases, pH decreases. This strong correlation enabled the development of a lactate soft sensor using a pH probe, allowing real-time monitoring and control of lactate levels.
[0225] Phase separation and cytotoxicity recovery in a stirred-tank bioreactor (STR)
[0226] We demonstrated quick cytotoxicity recovery even at production scale, as outlined in the following.
[0227] A fed-batch NK-92 expansion process ("expansion phase") was conducted in a 2-liter STR, in which a clinically-relevant dose of cells was produced. Cells doubled exponentially, reaching a similar maximum growth rate of 0.026 h-1as recorded in T-flasks, which yielded a total VCC maximum of
[0228] 2. 02 X 109± 9. 02xl07cells (Figure 3A). Viability was constantly high, only dropping to around 85% during the early lag phase after inoculation and after growth termination on day 8 (Figure 3A). Cells showed maximum cytotoxicity of 56.44 ± 0.46% on day 4, but by the end of expansion on day 8, cytotoxicity was lost entirely as it declined below the detection limit (Figure 3B). This behavior was very similar to what we observed in small-scale T-flasks, however, cytotoxicity maximum was almost 13% higher in the bioreactor, most probably as a result of the beneficial dynamic culturing environment. Feeding was initiated on day 4. From day 4 to 6, lactate concentration stabilized at 11.9 ± 0.1 mM, while cytotoxicity decreased to approximately half of its maximum to 30.4 ± 2.0%, highlighting again the negative influence of lactate on cytotoxicity (Figure 3B). Cytotoxic capacity increased constantly until day 4, when it dropped for 1 day, following the decrease in cytotoxicity (Figure 3C). Due to the highest growth rate on day 6, cytotoxic capacity increasedonce more, before it rapidly fell below the detection limit on day 8, as a consequence of low cytotoxicity and the end of cell growth.
[0229] The scale-up of the production of NK-92 cells to a 2-liter STR showed comparability of results generated during small-scale static expansion experiments. It was demonstrated that scale-up to production scale was indeed possible while further improving performance. During this first phase the goal was not to maintain high cytotoxicity, but rather high growth rate, which was achieved, even exceeding the growth rates in static cultures.
[0230] Following expansion, a recovery of cytotoxicity was carried out ("recovery phase"), using the following setpoints determined to be optimal: 90% fresh medium and 37. 2 °C incubation. Cytotoxicity recovered and cells remained alive. Following a lag phase of 2 days, the VCC increased and cells grew continuously until the end of the recovery phase (Figure 4A, 4B). There were some kinetic differences, as cytotoxicity recovered quicker than in T-flasks, reaching its maximum (43.06 ± 8.18%) on day 2 (Figure 4B), however, a rapid decline followed afterwards (i. e. when lactate concentration came close to 10 mM) (Figure 4B), while cytotoxic capacity still kept increasing (Figure 4C). Similar to the trend observed in small-scale experiments, cytotoxic capacity continued to rise until day 3, resulting in 6. 69xl07± 5.35xl06cytotoxic cells. However, contrary to the observations from T-flasks, cytotoxic capacity could be maintained at high level until day 4, before its rapid decrease was observed (Figure 4C).
[0231] Unlike current processes, by conducting a fed-batch expansion the need for fresh medium addition during the expansion phase was significantly reduced. The subsequent recovery phase led to a rapid and successful cytotoxicity recuperation. Contrary to the small-scale experiments, the optimized and well-
[0232] controlled PPs in the STR, in particular the pH value, even promoted cell growth which was critical for cytotoxic capacity enhancement. In this study, due to vessel size limitations, only 6% of the expanded cells could be transferred to the recovery phase. At a manufacturing scale, where this limitation would not apply, the final VCC and cytotoxic capacity would be approximately 15-fold higher than observed in this example. All in all, our example showed the applicability and great potential of NK-92 cell production at scale by separating the process into production phase and recovery of cytotoxicity ("recovery phase").
[0233] pH auxostat controlled fed-batch expansion in a stirred-tank bioreactor
[0234] To demonstrate the feasibility of the pH-auxostat approach for automated feed control, a fed-batch expansion was performed in a 2L STR in which a clinically relevant dose was produced within 8 days. During the culture, cells doubled exponentially, reaching a similar maximum growth rate of 0.026 h-1as recorded in T-flasks and non-automated feeding in STR, which yielded a total VCC maximum of 2.06xl09± 4.78xl07cells on day 8 (Figure 5A, 5C). Viability remained high, only dropping briefly to around 80% in the early lag phase after inoculation and again after the end of the fed-batch phase on day 6 (Figure 5A). The lactate concentration increased steadily until the start of the two-day fed-batch phase from day 4 to 6 when it was maintained stable at around
[0235] 11.2 ± 1.71 mM (Figure 5B). Cells exhibited maximum cytotoxicity of 38.3 ± 1.75% on day 1, which decreased to
[0236] 25.3 ± 3.7% later (Figure 5B). By the end of the fed-batch phase on day 6, cytotoxicity was lost entirely in this particular setup as it declined below the detection limit(Figure 5B). Cytotoxic capacity increased constantly until day 6, when it dropped and remained low, following the decrease in cytotoxicity (Figure 5C).
[0237] The automation of feed control using a lactate soft-sensor in a pH-auxostat setup yielded results comparable to nonautomated expansion experiments in an STR. It demonstrated the possibility to automate the expansion phase of NK cell, and made it applicable to all vessels equipped with a pH probe. Cytotoxic capacity could be maximized within 6 days of expansion marking the ideal point of harvest with the most potent cells.
[0238] Protein analysis
[0239] L-lactate concentrations increased in both conditions during the 2-day cultivation period, but to markedly different levels. In cultures seeded without initial lactate, L-lactate reached 12.54 ± 0.22 mM, whereas in cultures seeded with lactate [23.44 mM], levels rose significantly to 30.97 ± 0.19 mM (p < 0.0001 ) (Figure 6A). Notably, cell viability was similar between conditions (p = 0.894 ), indicating that lactate did not adversely affect cell survival at this level in this static small-scale setup (Figure 6B). As expected from the experiments described above, cytotoxicity was reduced in high lactate conditions, dropping to 1.34 ± 0. 65%, while it remained high at 49.34 ± 1.58% in the absence of initial lactate (p < 0.0001 ) (Figure 6C). Similarly, the specific growth rate decreased significantly from 0.008 ± 0.0 h-1in lactate-free conditions to 0.006 ± 0.0 h-1under high lactate conditions (p < 0.0001 ) (Figure 6D). Analysis of expression in revealed a total of 7992 ± 38 identified proteins across all replicates including 647 differently expressed proteins, consisting of 370 proteins upregulated and 374 proteins
[0240] downregulated in absence of initial lactate. Surprisingly, among all identified proteins, FAM13B was abundantly expressed in conditions without lactate (25,0637.88 ± 3,233.7 LFQ and 16,259.93 ± 3,374.81 LFQ for FAM13B and TNFL6, respectively), whereas it was undetectable in high lactate conditions (p < 0.0001 ) (Figure 6E). FAM13B is poorly studied and not clearly connected to cytotoxicity in the prior art. By contrast several other proteins linked to cytotoxicity, such as FasL (TNFL6), Perforin-1, KIR3DX1, or CCR1 were found significantly less abundant in the presence of lactate, which confirmed our functional findings in regard to cytotoxicity and cytotoxic capacity.
[0241] In particular, the highly significant downregulation of FasL in the presence of lactate (see Fig. 6F) was very surprising. This contrast with the observations in Huang et al, 2023, which did not find significant downregulation of FasL in the presence of lactate (see Fig 3A of that document).
[0242] Next, we tested the effect of supplementation with soluble FasL on the cytotoxicity of the NK cells (either pre-incubated with 15 mM lactate or without lactate). The results of this FasL supplementation assay, as depicted in Fig. 7, demonstrate a statistically significant and dose-dependent enhancement of cytotoxicity, strikingly also in the samples which were subj ected to lactate pre-incubation.
[0243] The effect was very pronounced in the high lactate environment, which are characteristic for many TMEs in solid malignancies.
[0244] While the soluble FasL supplementation in the assay demonstrated efficacy, it is expected that membrane-bound, full-length FasL elicits an even more robust and sustained effect, due to amplification of the apoptotic signal by
[0245] downstream signalling and / or facilitating Fas receptor clustering.
[0246] In conclusion, the FasL supplementation assay provides compelling evidence that transient enrichment of FasL on NK cells significantly enhances their resistance to lactate-induced dysfunction and restores cytotoxic activity in high lactate environments
[0247] Non-patent references
[0248] Andreola, Giovanna, et al. " Induction of lymphocyte apoptosis by tumor cell secretion of FasL-bearing microvesicles. " Journal of Experimental Medi cine 195. 10 ( 2002 ): 1303- 1316.
[0249] Azambuja, Juliana Hofstatter, et al. " Molecular profiles and immunomodulatory activities of glioblastoma-derived exosomes. " Neuro-oncol ogy advances 2. 1 ( 2020 ): vdaa056.
[0250] Berg, et al. " Clinical Grade Ex Vivo-Expanded Human Natural Killer Cells Upregulate Activating Receptors and Death Receptor Ligands and Have Enhanced Cytolytic Activity against Tumor Cells", Cytotherapy, vol. 11, no. 3, 1 January 2009
[0251] Berrien-Elliott, et al. " Allogeneic natural killer cell therapy. " Bl ood 141. 8 ( 2023 ): 856- 868.
[0252] Böhme, Ines, and Anja Katrin Bosserhoff. " Acidic tumor microenvironment in human melanoma. " Pigment cell & melanoma research 29. 5 ( 2016 ): 508-523.
[0253] Broker, Katharina, et al. " Mass production of highly active NK cells for cancer immunotherapy in a GMP conform perfusion bioreactor. " Frontiers in Bioengineering and Biotechnology 7 ( 2019 ): 194.
[0254] Costariol, Elena, et al. " Establishing the scalable manufacture of primary human T-cells in an automated stirred-tank bioreactor. " Biotechnology and bioengineering 116. 10 ( 2019 ): 2488-2502.
[0255] de la Cruz-Lopez, Karen G., et al. " Lactate in the regulation of tumor microenvironment and therapeutic approaches. " Frontiers in oncology 9 ( 2019 ): 1143.
[0256] Demichev, V., et al., DIA-NN: neural networks and interference correction enable deep proteome coverage in high throughput. Nat Methods, 2020. 17 ( 1 ): p. 41-44.
[0257] Demichev, V., et al., dia-PASEF data analysis using FragPipe and DIA-NN for deep proteomics of low sample amounts. Nat Commun, 2022. 13 ( 1 ): p. 3944.
[0258] Fang, Fang, et al. " Advances in NK cell production. " Cellular & molecular immunology 19. 4 ( 2022 ): 460-481.
[0259] Granzin et al. " Fully automated expansion and activation of clinical-grade natural killer cells for adoptive immunotherapy", Cytotherapy, vol. 17, no. 5, 1 May 2015
[0260] Gu, Xuan-Yu, et al. " Impact of lactate on immune cell function in the tumor microenvironment: mechanisms and therapeutic perspectives. " Frontiers in Immunology 16 ( 2025 ): 1563303.
[0261] Guia, Sophie, et al. " Genome-wide CRISPR / Cas 9 screen reveals factors that influence the susceptibility of tumor cells to NK cell-mediated killing. " Journal for Immunotherapy of Cancer 13. 3 ( 2025 ): e010699.
[0262] Huang, Y., et al. " Lactic Acid Inhibits the Cytotoxicity of NK- 92 Cells via Activating the Lactate Receptor GPR81. " Annual Research & Review in Biology, 2023: p. 1- 10.
[0263] Jia, Hao, et al. " NK cell exhaustion in the tumor microenvironment. " Frontiers in Immunology 14 ( 2023 ): 1303605.
[0264] Klingemann, Hans. " Challenges of cancer therapy with natural killer cells. " Cytotherapy 17. 3 ( 2015 ): 245-249.
[0265] Li, Jian, et al. " Scale-up bioprocess development for production of the antibiotic valinomycin in Escherichia coli based on consistent fed-batch cultivations. " Microbial Cell Factories 14 ( 2015 ): 1- 13.
[0266] Motallebnejad, Pedram, et al. " Process engineering of natural killer cell-based immunotherapy. " Trends in biotechnology 41. 10 ( 2023 ): 1314- 1326.
[0267] Meier, F., et al., diaPASEF: parallel accumulation-serial fragmentation combined with data-independent acquisition. Nat Methods, 2020. 17 ( 12 ): p. 1229- 1236.
[0268] Myers, Jacob A., and Jeffrey S. Miller. " Exploring the NK cell platform for cancer immunotherapy. " Nature reviews Clinical oncology 18. 2 ( 2021 ): 85- 100
[0269] Oyer, J. L., et al., Generation of highly cytotoxic natural killer cells for treatment of acute myelogenous leukemia using a feeder- free, particle-based approach. Biol Blood Marrow Transplant, 2015. 21 ( 4 ): p. 632- 9.
[0270] Penney McWilliams-Koeppen, M. S., and James Clinton. " A simple and rapid alternative to 51chrAomium or fluorescent dye loading for quanti fication of natural killer cell activity. " Application Note, American Type Culture Collection (ATCC ), 2022. https: / / www.atcc.org / - / media / resources / application-notes / a-simple-and-rapid-alternative-to-51chromium.pdf, accessed on 17 March 2025
[0271] Rossin, Aurelie, Giorgia Miloro, and Anne-Odile Hueber. " TRAIL and FasL functions in cancer and autoimmune diseases: towards an increasing complexity. " Cancers 11. 5 ( 2019 ): 639.
[0272] Schroeder, U., et al. " Decline of lactate in tumor tissue after ketogenic diet: in vivo microdialysis study in patients with head and neck cancer. " Nutrition and cancer 65. 6 ( 2013 ): 843- 849.
[0273] Schwane, Vera, et al. " Distinct signatures in the receptor repertoire discriminate CD56bright and CD56dim natural killer cells. " Frontiers in Immunology 11 ( 2020 ): 568927.
[0274] Screpanti, Valentina, et al. " A central role for death receptor-mediated apoptosis in the rejection of tumors by NK cells. " The Journal of Immunology 167. 4 ( 2001 ): 2068-2073.
[0275] Spanholtz, J., et al., Clinical-grade generation of active NK cells from cord blood hematopoietic progenitor cells for immunotherapy using a closed-system culture process. PLoS One, 2011. 6 ( 6 ): p. e20740.
[0276] Terrén, Iñigo, et al. " NK cell metabolism and tumor microenvironment. " Frontiers in immunology 10 (2019): 2278.
[0277] Tyanova, S., et al., The Perseus computational platform for comprehensive analysis of (prote ) omics data. Nat Methods, 2016. 13 ( 9 ): p. 731-40.
[0278] Valeri, Antonio, et al. " Overcoming tumor resistance mechanisms in CAR-NK cell therapy. " Frontiers in Immunology 13 ( 2022 ): 953849.
[0279] Villalba, Martin, et al. " Non-genetically improving the natural cytotoxicity of natural killer (NK) cells. " Frontiers in Immunology 10 (2020): 3026.
[0280] Vivier, Eric, et al. " Natural killer cell therapies. " Nature 626. 8000 ( 2024 ): 727-736.
[0281] van Vliet, Amanda A., et al. " Adoptive NK cell therapy: a promising treatment prospect for metastatic melanoma. " Cancers 13. 18 ( 2021 ): 4722.
[0282] Walenta, et al. " Lactate in solid malignant tumors: potential basis of a metabolic classi fication in clinical oncology. " Current medicinal chemistry 11. 16 ( 2004 ): 2195-2204.
[0283] Wang, Xuening, et al. " Scalable process development of NK and CAR-NK expansion in a closed bioreactor. " Frontiers in immunology 15 ( 2024 ): 1412378.
[0284] Yolcu, Esma S., et al. " Cell membrane modification for rapid display of proteins as a novel means of immunomodulation:
[0285] FasL-decorated cells prevent islet graft rejection." Immunity 17. 6 ( 2002 ): 795- 808.
Claims
Claims1. An in-vitro method for producing natural killer (NK) cell formulations for cancer therapy, comprising at least the following steps:- expanding at least one NK cell in a first culturing step,- transiently enriching Fas ligand (FasL) on the surface of the expanded NK cells, and- providing the expanded and transiently FasL-enriched NK cells in an NK cell formulation for cancer therapy.
2. The method of claim 1, wherein said transiently enriching FasL comprises decorating the surface of the expanded NK cells with exogenous FasL, preferably wherein the FasL protein sequence is the native FasL sequence, in particular wherein both the NK cells and FasL protein sequence are human.
3. The method of claim 1 or 2, wherein said transiently enriching FasL comprises transient overexpression of FasL in the expanded NK cells, preferably wherein the transient overexpression is started only after said expanding step, and / or preferably wherein the FasL protein sequence is the native FasL sequence, in particular wherein both the NK cells and FasL protein sequence are human.
4. The method of any one of claims 1 to 3, wherein the CD56brightNK cell subpopulation is selectively enriched before and / or during the expanding step, preferably wherein the CD56dimNK cell subpopulation is selectively enrichedthereafter, in particular before the transiently enriching step; or wherein the CD56dimNK cell subpopulation is selectively enriched after the expanding step, in particular before the transiently enriching step.
5. An NK cell formulation for cancer therapy, obtainable according to the method of any one of claims 1 to 4.
6. An NK cell formulation for cancer therapy, wherein FasL is transiently enriched on the surface of NK cells; preferably wherein the surface of NK cells is decorated with exogenous FasL or wherein NK cells transiently overexpress FasL.
7. The NK cell formulation of claim 5 or 6, wherein the FasL protein sequence is the native FasL sequence.
8. The NK cell formulation of any one of claims 5 to 7, for use in treatment of a solid malignancy.
9. The NK cell formulation for use according to claim 8, wherein the solid malignancy is selected from the group consisting of breast cancer, in particular triple-negative breast cancer, colorectal cancer, glioma, in particular glioblastoma, head and neck squamous cell carcinoma, lung cancer, in particular non-small cell lung cancer, melanoma, ovarian cancer, cervical cancer, pancreatic cancer, prostate cancer, and sarcoma, especially head and neck squamous cell carcinoma, lung cancer, in particular non-small cell lung cancer, melanoma, ovarian cancer, pancreatic cancer, prostate cancer, and sarcoma.
10. The NK cell formulation for use according to claim 8 o 9, wherein the solid malignancy has a lactate concentration of at least 5 µmol / g, preferably at least 6 µmol / g, more preferably at least 7 µmol / g, even more preferably at least 8 µmol / g, yet even more preferably at least 9 µmol / g or even at least 10 µmol / g, or of at least 5 mM, preferably at least 6 mM, more preferably at least 7 mM, even more preferably at least 8 mM, yet even more preferably at least 9 mM or even at least 10 mM.
