Autologous NK-cell-enrichment: preclinical setting phase, Shiraz experience

: Background NK cell therapy has proven to be a promising approach for treatment of hematological malignancies and solid tumors. Masuyama et al. have recently introduced a new method for ex-vivo autologous NK cell expansion (Osaki method); resulting in the production of ample active NK cells for a promising cell therapy regimen. In order to start clinical trial phase I at Shiraz University of medical Sciences in collaboration with Masuyama clinic and St. Luck's International University Hospital, this preclinical setting study aimed to evaluate the proliferative efficacy of the method, the activation status of expanded autologous NK cells and the likely unwanted contamination of the final cell product. PBMCs were isolated from 30 ml of 5 healthy individuals' peripheral blood transferring directly to the specified initial culture bag containing antibodies for CD3, CD52 as well as IL-2 cytokine. The cells were cultured for 14-17 days in incubators; during which the cell received as indicated by in-vitro cytotoxic assay; with strong expression of NKG2D and CD16. The prepared final cell products were negative for HCV, HBV, HIV, Mycoplasma and endotoxin. In the preclinical setting phase, large numbers of activated and un-contaminated NK cells from 30 ml of healthy individuals' peripheral blood were successfully generated. The method seems to provide ample clean cell product with no contamination; suitable to be infused back to the patients in phase I clinical trial.


Background
Cancer is one of the important causes of death in Iranian population [1]. A considerable amount of literature has been published on the demonstration of the immune system crucial roles in tumor biology [2,3]. NK cells, which have been defined as CD3-CD16/56+ cells, are vital effector cells in innate immune response against tumor cells [4]. These cells have been well established to be essential in the immune surveillance; linking anti-tumor adoptive and innate immune responses [5]. Thanks to the way of recognition, as well as the natural cytotoxic potential, NK cells not only destroy target tumor cells without prior sensitization but also release crucial cytokines for engaging adoptive immunity into antitumor battle [6]. According to the conventional classification, two subsets of NK cells have been introduced; cytotoxic (CD56 dim CD16 bright ) and regulatory (CD56 bright CD16 dim ) NK cells [7]. In the more recent publications, NK cells are, however, divided, based on the differences in their phenotypes and functions, into three cell subsets including NK cytotoxic (CD56 dim CD11b + CD27 -), NK tolerant (CD56 bright CD27 -CD11b -), and NK regulatory (CD56 bright CD27 + CD11b +/-) cells [8]. Among different NK cell surface molecules, CD16 (FcγRIII) is the most well-known effecter molecule involving in NK-mediated antibody-dependent cellular cytotoxicity (ADCC) [9]. Antitumor activity of NK cells made it a promising candidate for cancer immune cell therapy. Although an ample data is available verifying the effect of NK cell based therapy in hematological malignancies [10,11] and solid tumors [12,13], an important obstacle, however, restricts the clinical use of theses natural lymphocytes. Different ex-vivo methods for NK cells expansion has been proposed but obtaining the large numbers of functional NK cells using ex-vivo culture is very difficult. Several new methods have been developed to overcome this obstacle and produce purified and active NK cells [14][15][16][17]. These protocols have some limitations for clinical applications including low scale of NK cell expansion, low purity, cytotoxic activity, high cost and complicated protocols [12,17] A recent published article has illustrated development of a simple and safe method for the ex vivo NK cells expansion.
Masuyama JI and his colleagues expanded a large number of active NK cells using anti-CD3, anti-CD52 monoclonal antibodies and IL-2 [18]. This protocol has many advantages to use for adoptive immune cell therapy of patients with cancer and in the current study, we expanded NK cells from peripheral blood mononuclear cells of healthy individuals using the mentioned protocol to check method, our facilities and expanded NK cells in our cell processing room in the preclinical setting phase.

Subjects
To assess the success of the procedure as well as the quality of our cellular products in the preclinical setting phase, peripheral blood samples were obtained from five healthy males.

Cell expansion
Cell processing, expansion and evaluation were performed by cellular Good Manufacturing Practice (cGMP) facilities in cell processing clean room at Ghadir hospital affiliated to Shiraz University of Medical Sciences, Shiraz, Iran. Adopted from Masuyama et al [18], here after Osaki method, with some in-home modifications was used for NK cells expansion and activation.
Briefly, 30-40 ml of heparinized peripheral blood was obtained from each healthy individual. After separating plasma, peripheral blood mononuclear cells (PBMCs) were isolated from blood using Ficoll®-Paque Premium (GE Healthcare, USA) gradient centrifugation. Pre-expansion analysis, killing assay and Pathogen-free evaluation tests of the separated cells (described later in this section) were performed on a part of the sample. For the rest of the sample to be expanded, PBMCs were first suspended in NKGM medium (Cellex, Japan), heat-inactivated autologous plasma and IL-2 (Novartis, Switzerland). The cells were then transferred to the initiation culture bag (Cellex, Japan) coated with anti-CD3 and anti-CD52 antibodies for 2-3 days, following by cultivating, in fresh NKGM medium and IL-2, in 225-culture flask (Corning, USA) for more 2-3 days. Cellular colonies were then transferred to one or two expansion culture bags (Cellex, Japan) for 14 to 17 days during which the cells were fed, every 2-3 days, by fresh medium plus IL-2. Three days before sample collection, 10 ml of culture medium was tested against sterility standards (described later in this section). Finally, the cells were collected between day 14 to 17 based on the cell expansion quality. Post-expansion analysis, and killing assay of the separated cells (described later in this section) were performed on the expanded samples. events. The data were subsequently analyzed by FlowJo software package (version 7.6.2, USA).

Assessment of Natural Killer cell cytotoxic activity
 Calcein release killing assay K562 cells (chronic myelogenous leukemia (CML) cell line) were selected as the target (T) cells for NK cell cytotoxic activity assay. The cells were labeled with Calcein AM (BD-USA) and cocultured with PBMCs as the effector (E) cells before and after expansion in a U bottom 96-well microtiter plate with different effector:target cells ratios in triplicate. Additional wells were used for the assessment of Calcein spontaneous release (only target cells in medium), maximum Calcein release (target cells in 10% Triton X) and background (medium only). After incubation at 37°C in 5% CO2 for 4 h and centrifugation, supernatant was harvested and measured using a microplate spectrofluorimeter. The cytotoxic activity were calculated by following formula:

 Degranulation, lytic proteins and IFN-γ production assays
PBMCs (before and after expansion) were incubated with and without K562 cells at 37∘ C, 5% USA) production were determined using appropriate antibodies in the effector cells.

Results
Five healthy males participated in the current study. The mean percentage of their age was 32.4 ± 3.1.

Calcein release killing assay
The expanded cells were highly lytic as indicated by in-vitro Calcein release cytotoxic assay. The cytotoxic activity ratio of the expanded cells to the cells before processing was revealed to be 3.19 folds, Figure 3.

Degranulation, lytic proteins and IFN-γ production assays
The mean fluorescent intensity (MFI) of perforin and IFN-γ staining in the NK and CD3+ cells were shown to be higher in E and E/T cells after expansion compared to the ones before the process. However, CD107a staining MFI was decreased following cell expansion.

Contamination test out
The prepared final cell products were shown to be negative for HCV, HBV, HIV, Mycoplasma and endotoxin.

Discussion
Convincing evidences revealed that NK cells as a significant armament of the innate immune system play undisputable roles in immune surveillance against different types of cancers [19]. Accordingly NK cells have been in the center of many therapeutic approaches by scientists from all over the world [20][21][22]. There are five main immunotherapy approaches focusing on NK lymphocytes: systemic administration of recombinant cytokines involved in NK cell activation including Interleukin (IL)-2, IL-15 and IL-12, systemic administration of recombinant monoclonal antibodies with the potency to trigger NK cell-mediated ADCC; autologous adoptive NK cells transfer, allogeneic adoptive NK cells transfer after selected KIR mismatch particularly in the patients with hematological malignancies, and administration of NK cell lines particularly NK-92 which observed to be a safe cell therapy approach. Furthermore, NK cells have recently attracted attentions for chimeric antigen receptor (CAR) genetic engineering [23]. Diverse methods for expansion and activation of NK cells have been introduced in the last decade. The majority of these methods, however, have been indicated to suffer from serious confines in terms of clinical applications including low scale cell expansion, low purity, low cytotoxic activity, high cost and the complexity of the protocol [12,17].
In the present study, and as a prerequisite to start autologous NK cell enriched therapy phase I clinical trial in Shiraz, Iran, the NK cell enrichment in whole PBMCs of five healthy individuals has been performed using a method adopted from Masuyoma et al [18] with some modifications. One of the most significant issues in immune cell processing protocols is to end with sufficient number of the target cells. Our results indicated that following applying this method, total NK cells (CD3-CD56+/CD16+/-) were expanded 510-fold in average (ranging 200-1100 fold). The mean percentage of NK cells per lymphocytes was 8.25 ± 3.21 before expansion, which extended to 43.24 ± 18.34 after expansion (Maximum 68%).  [25]. Therefore, the average expansion of the Osaki method seems worthy in comparison with the other protocols expansion rate.
Besides NK cells, the mean-percentages of the other immune cell subsets including total non-T lymphocytes, B cells, total T cells, helper T lymphocytes, cytotoxic T cells, NKT lymphocytes, as well as regulatory T (Treg) cells were observed to be 29. . Evaluation of NK cell subsets were also assessed based on the new calcification method applying the expression level of CD56, CD11b and CD27 [8]. Accordingly the percentages of both regulatory (CD3-CD56+ CD27+ CD11b+/-) and tolerant (CD3-CD56+ CD27-CD11b-) NK subsets were decreased after expansion, and the dominated population of the final NK cells after expansion was observed to be cytotoxic NK cells subset (CD3-CD56+CD11b+CD27-). Consequently it can be suggested that what we observed based on the older classical classification, i. e. the dominant expanded cells with both CD16 and CD56 overexpression (CD16 hi CD56 hi), are cytotoxic NK cells. Consistently, the cytotoxic assay of the expanded cell by Calcein released method demonstrated that the cytotoxic activity of the final cell product has raised 3-fold compared to the cells before expansion. IFN-γ were also observed to be higher in the cells after NK cell enrichment as illustrated by co-cultivation of K562 target cell line with the expanded cells following by flowcytometry analysis.
The ability of the expanded NK lymphocytes to infiltrate and to persist in the tumor microenvironment has been considered as the forth important characteristic of the cells in an expansion process. As an essential chemokine-chemokine receptor for NK cells homing, CXCL10-CXCR3 axis has been illustrated to be related in higher infiltration of NK cells into tumor microenvironment as well as tumor draining lymph nodes [26]. CD96 is an important adhesion molecule involved in NK cell-target cell adhesion by ligation with CD155 [27]. In the present study, the percentages of NK cells expressing CXCR3 and adhesion molecule and CD96 were revealed to be significantly increased in expanded cell product (27.30 ± 0.01 Vs. 93.75 ± 8.55, and 2.34 ± 2.61 Vs. 18.53 ± 16 respectively). These observations suggest that Osaki method is able to increase the ability of the expanded cells to be recruited to tumor environment and to be efficiently engaged with the target cells; a suggestion which needs more functional assays to be fully elucidated.
As the last but not the least entity, the ability of the expanded NK cells to be activated and to kill target cells efficiently should take into account in any cell therapy based approach. NKG2D, as an NK cells activating receptor, demonstrated to be involved in tumor growth suppression in several research [28]. In the present study, after NK cells enrichment, not only the production of IFN-γ and the cytolytic activity of the expanded cells against K562 cell line were increased (as previously described), but also the percentages of the NK cells positive for NKG2D, one of the main NKactivating receptor was also significantly increased (37.88 ± 1.47 Vs. 7.26 ± 2.45). Evaluating the expression of lytic proteins including Perforin and Granzyme B in enriched NK cells were illustrated not to be dramatically changed after enrichment. In fact in total NK cell population before and after expansion, the percentages of NK cells positive for theses lytic proteins was around 90%. Interestingly, and as a co-finding, the cell expansion method in the present study was observed to increase the percentages of CD8+ T lymphocytes expressing Granzyme, as well as the percentages of T lymphocytes (most likely CD8+) expressing perforin (14.43 ± 13.93 Vs. 90.56 ± 4.82 and 13.70 ± 0.34 Vs. 83.20 ±0.65). These observations collectively suggest the expanded cells are well armed, are able to be effectively activated and to effectively kill target cell by means of their effector lytic molecules.
In our study, NK cell enrichment process was performed in the clean room. Assessing the final cell product by standard clinical tests revealed negative HCV, HBV, HIV, Mycoplasma and endotoxin contamination before and after procedure. Consequently and as a prerequisite to start autologous NK cell enriched therapy phase I clinical trial in Shiraz, this method seems to provide ample clean cell product with no contamination; safe to be infused back to the patients in phase I clinical trial.

Conclusion
As a preclinical setting phase of NK cell enrichment for implication in immunotherapy, method adopted from Masuyama et al [18]; seems to expand a huge number of un-contaminated NK cells. Not only NK cells, but also cytotoxic T cells seems to be be increased after cell processing by this method. The dominant expanded NK cell subset seems to be cytotoxic NK cells and to have significant increased IFN-γ release, overexpressed chemokine receptor for NK cells homing, over expressed cell adhesion molecule involved in NK cell-target cell adhesion and over expressed NKG2D, as an NK cells activating receptor. The expanded cells seems to have enough lytic molecules Granzyme and Perforin, have their cytotoxic activity increased significantly and are able to kill target tumor cells. These observations support the idea that the expanded cells are able to be effectively recruited and to meritoriously target the tumor cells in tumor microenvironment. The expansion method seems not only to expand and to activate NK cells but also to affect the other immune cells all in favor of antitumor immunity. The data have illustrated that the final cell product processed by this method is suitable to be infused back to the patients in phase I clinical trial for refractory breast cancer.
As a limitation in the current study, the cells were from healthy donors in which NK cells are not expected to be spontaneously suppressed. The challenge is whether or not we can get the same results when the cells are extracted from end-stage patients (in clinical trial) have the same responses to expansion method.