1. Isolation of PBMCs from Buffy Coat Peripheral blood mononuclear cells (PBMC) are obtained by buoyant density centrifugation on Ficoll-Paque from healthy donor buffy coat samples derived by leukapheresis. The Ficoll-Paque centrifugation is done as per manufacturer’s protocol with minor modifications. Add PBS to a normal-donor blood-bank buffy coat to a final volume of 140 mL (typical buffy coat volume is 40-70 mL). Layer 35 mL of buffy coat sample on 15 mL of Ficoll-Paque (4 tubes). Centrifuge at 400g for 20 minutes without brake. Recover the PBMCs from the Ficoll-Paque: plasma interface, do not discard the RBCs at the bottom of the Ficoll-Paque. Wash PBMCs three times with PBS, centrifuging each time at 400g for 10 minutes. PBMCs can be used directly for NK cell expansion at this stage or NK cells can be isolated by RosetteSep (Section 4) Remaining PBMCs can be frozen in FBS containing 10% DMSO in liquid nitrogen. Aspirate the Ficoll and collect the RBCs from step 5 in to two 50 mL centrifuge tube, wash three times with PBS (added to 50 mL mark), each time aspirate the supernatant skimming the top of the RBC layer to remove granulocytes. The RBCs can be used immediately for RosetteSep purification of NK cells (refer to Section 4) or stored in equal volume of Alsever’s solution at 4°C for later use (The RBCs can be store for a maximum of 4 weeks). 2. NK Cell Expansion The NK cell expansion can be initiated using PBMCs or purified NK cells. The amount of PBMCs used for expansion can be varied based on the amount of NK cells desired at the end of a three week expansion, refer to representative results section for details. (See Note 1) STIMULATION 1 Day 0 For each 5×106 PBMCs to be expanded, count and irradiate 10×106 K562 Cl9 mIL21 using a gamma irradiator at 100 Gy. Post irradiation, wash the cells with PBS and resuspend in NK cell expansion media (NKEM). Seed 5×106 PBMCs with 10×106 irradiated K562 Cl9 mIL21 (1:2 ratio) in 40 mL of NKEM in a T75 flask and place it upright in an incubator at 37°C and 5% CO2. Days 3 and 5 Recover cells by centrifugation at 400g for 5 min and replace half of the media with fresh NKEM (adding fresh IL2 for the entire media volume) and continue culture. STIMULATION 2 Day 7 Count the number of cells in culture at the end of one week. Set aside 5×105 cells for phenotyping by flow cytometry (See Note 2) For each 5×106 cells to be restimulated, count and irradiate 5×106 K562 Cl9 mIL21 using a gamma irradiator at 100 Gy. Add an equal number of irradiated K562 Cl9 mIL21 (1:1 ratio) and resuspend in NKEM at 2.5×105 total cells/mL (See Note 3). Seed cells in T75 flasks (maximum 50 mL per flask). Days 10 and 12 Count the number of cells. Change entire media with fresh NKEM based on the cell numbers (See Note 3). Day 14 At the end of two weeks of expansion count the number of cells in culture. Set aside 5×105 cells for phenotyping by flow cytometry (See Note 2) If expansion was started from PBMCs the NK cells can be purified at this stage of expansion using the RosetteSep purification protocol (refer to Section IV). If expansion was started from purified NK cells proceed to stimulation 3. After purification set aside 5×105 cells for phenotyping by flow cytometry to verify purity of the NK cells (as in Step 8). Proceed with Stimulation 3 using all of the purified NK cells (See Note 4). STIMULATION 3 Resuspend NK cells with irradiated K562 Cl9 mIL21 (1:1 ratio) in NKEM based on cell numbers (See Note 3). Days 17 and 19 Count the number of cells. Change media with fresh NKEM based on the cell numbers (See Note 3). Day 21 At the end of three weeks of expansion count the number of cells in culture. Recover 1×106 cells for flow cytometry analysis for full NK cell phenotyping antibody panel (See Table 1). Freeze cells in FBS containing 10% DMSO at a maximum density of 5×107 cells per vial for future use. 3. NK Cell Cytotoxicity Assay Thaw a vial of NK cells and seed in NKEM, one day prior to performing cytotoxicity assay to allow recovery. For each NK cell cytotoxicity assay using a single target cell line, 6×105 NK cells and 3×105 target cells are required (See Note 5). Prepare CAM-Media by diluting Calcein-AM (stock 1 mg/mL in DMSO) in NKEM (See Note 6). Resuspend 106 target cells in 1 mL of CAM-media (See Note 7). Incubate for 1 h at 37 °C, with occasional shaking. Resuspend NK cells at 1×106 cells/ mL and add 200uL of NK cell suspension to each of the 3 wells of a U-bottom 96-well plate corresponding to10:1 E:T ratio shown in Figure 1. (See Note 8) Add 100uL of complete media to all remaining wells except for “Maximum”. Add 100uL of 2% Triton X-100 to “Maximum”. Perform serial dilutions of the NK cells for the 5 subsequent E:T ratios by transferring 100uL of cells each time, mix well. Discard 100uL from the last wells (E:T ratio of 0.3125:1). After 1 hour of calcein loading, wash target cells in NKEM twice, centrifuging for 5 min at 1200 rpm. (See Note 9) Re-count the target cells and resuspend at 1×105 cells/ mL. Add 100 μL of target cells to each well (1×104/well). Centrifuge for 1 min at 100g to initiate cell contact. Incubate at 37 °C and 5% CO2 for 4 hours. Mix the culture gently by pipetting with a 100 μL pipetter in order to uniformly suspend the released calcein, spin down plate at 100g for 5 minutes to pellet the cells and transfer 100 μL of the supernatant to a new plate taking care to avoid bubbles. Pop any bubbles that may form using fine needle. Read the plate using a fluorescent plate reader (excitation filter 485 nm, emission filter 530 nm). Bottom read is recommended. Calculate Percent Specific Lysis according to the formula [(test release-spontaneous release)/(maximum release – spontaneous release)] x 100. 4. NK Cell Purification by RosetteSep Take 100-fold excess of RBCs to that of PBMCs or expanded cells, into a 50 mL tube (100:1 RBC:PBMC). If using fresh RBCs proceed directly to next step or if the RBCs were stored in Alsever s solution, count the number of RBCs and wash appropriate (100 fold excess) amount of RBCs with PBS supplemented with 2% FBS three times, centrifuging at 1200 rpm for 10 minutes each time. Combine RBCs with PBMCs from step 1.7 or expanded cells from step 2.10 in PBS + 2% FBS to a final volume of 1 mL per 5×107 of PBMCs or expanded cells. Add 1μL of RosetteSep Human NK Cell Enrichment Cocktail per 1×106 of PBMCs or expanded cells. Mix well and incubate at room temperature for 20 minutes with gentle mixing every 5 minutes. Add equal volume of PBS + 2% FBS mix gently and layer on top of Ficoll-Paque. Repeat the steps of Ficoll-Paque centrifugation described for PBMC isolation (Section I). Count the NK cells recovered after purification and set aside 5×105 cells for phenotpying by flow cytometry for NK cell purity (Step 8). 5. Notes NOTE 1. NK cells can be expanded directly from PBMCs, or from RosetteSep purified NK cells. We have noted similar expansion efficiency, but some donors may have very low NK cell numbers resulting in difficulty purifying by RosetteSep prior to expansion. NOTE 2. We routinely use CD56-FITC, CD16-PE, and CD3-PE-Cy5 for phenotyping during the expansion, enumerating NK cells as those which are CD3-negative and CD16- or CD56-positive. NOTE 3. At each media change or stimulation, resuspend cells at 2.5 x 105/mL to keep PBMC/NK cell numbers at or under 2 million per mL at peak stages of expansion. This will prevent depletion of nutrients and help achieve maximal expansion and survival. NOTE 4. The NK expansion rate is donor dependent and at the end of stimulations 1 or 2 part of the cells can be frozen and part expanded further depending on the experimental need. We have had good success in using the frozen cells for expansions at a later time. NOTE 5. In order to allow room for error we recommend using a minimum of 7×105 NK cells resuspended in 700 uls of NKEM and 4×105 Calcein-AM stained target cells resuspended in 4 mL of NKEM for setting up the cytotoxicity assay. If using multichannel pipette for seeding target cells higher volumes of cells (up to 6×105 in 6 mL) may be required based on the size of media basin being used. Also the recommended NK cell numbers are specifically for the E:T ratios show in the protocol, for using higher E:T ratios increase the NK cell numbers per mL accordingly (eg. For a 40:1 E:T ratio use 4×106 cells/ mL) NOTE 6. We recommend performing a preliminary Calcein-AM loading titration for the target cell line of choice, using the following dilutions of 1:500, 1:400. 1:300, 1:200 and 1:100 to achieve optimal difference between maximum and spontaneous release. NOTE 7. When using an adherent cell line as target, first prepare single cell suspension using non-enzymatic cell dissociation buffer. If performing ADCC, prepare a duplicate tube of target cells in CAM-media. NOTE 8. If performing ADCC, add same NK cells to 3 wells corresponding to 10:1 E:T for ADCC. Repeat for additional donors. Repeat for additional target cells. NOTE 9. If performing an ADCC experiment, after 45 minutes of calcein loading, add 10ug of antibody specific for inducing ADCC against the target cells. After 15 more minutes, wash target cells in complete medium twice, centrifuging for 5 min at 1200 rpm. Resuspend cells at 1×105 cells/ mL and proceed with the next step in the protocol. 6. REPRESENTATIVE RESULTS Figure 2. When the expansion is performed as per the scheme described above, using 5×106 PBMCs as starting material, typical NK cell yields range from 1×109 to 1010 cells (donor dependent variability). The figure shows NK cell fold expansion (n=19) compared to NK cells present in the original product (median +/- quartile). Figure 3. The expanded NK cells express various NK cell receptors that are comparable to the unexpanded primary NK cells with a few exceptions (CD11b, CD160 and CD244). Figure 4. PBMCs recovery from Buffy coat is donor dependent and can range from 300×106 to 800×106. NK cells may comprise 2% – 18% of the PBMCs. For RosetteSep purification of expanded cells, recovery of pure NK cells on day 14 ranges from 40-70%. By following the recommended protocol of expansion and purification, NK cell purity of 99% can be expected. Figure 5. Expanded NK cells have demonstrated cytotoxicity against a range of tumor cell lines including neuroblastoma, AML, osteosarcoma and melanoma (representative AML killing shown as percent specific lysis).