Generation and application of TGFβ-educated human Vγ9Vδ2 T cells

Summary Clinical trials that tested the antitumor activity of γδ T cells have been mostly unsuccessful. To address this, we expanded human Vγ9Vδ2 T cells in TGFβ1, a cytokine which enhances their viability, trafficking properties, and intrinsic antitumor activity. This protocol summarizes the production and in vitro functional characterization of TGFβ1 educated human Vγ9Vδ2 cells and highlights their compatibility with chimeric antigen receptor (CAR) engineering. We also describe in vivo testing of the antitumor activity of these CAR T cells in mice. For complete details on the use and execution of this protocol, please refer to Beatson et al. (2021).


TGFb promotes viability and retards differentiation of Vg9Vd2 cells
TGFb-educated Vg9Vd2 cells achieve greater invasion and anticancer activity Specificity of these cells can be reprogrammed using chimeric antigen receptors SUMMARY Clinical trials that tested the antitumor activity of gd T cells have been mostly unsuccessful. To address this, we expanded human Vg9Vd2 T cells in TGFb1, a cytokine which enhances their viability, trafficking properties, and intrinsic antitumor activity. This protocol summarizes the production and in vitro functional characterization of TGFb1 educated human Vg9Vd2 cells and highlights their compatibility with chimeric antigen receptor (CAR) engineering. We also describe in vivo testing of the antitumor activity of these CAR T cells in mice. For complete details on the use and execution of this protocol, please refer to Beatson et al. (2021).

BEFORE YOU BEGIN
Experimental consideration 1. gd T cells play a key role in immune surveillance for malignant transformation (Kabelitz et al., 2020). Moreover, the presence of intra-tumoral gd T cells is strongly predictive of improved prognosis across a range of cancer types (Gentles et al., 2015). The dominant circulating gd T cell subtype (d2 subtype) expresses a Vg9Vd2 T cell receptor that detects phosphoantigen intermediates of mevalonate metabolism, a metabolic pathway that is upregulated in transformed cells. These cells can be selectively activated using aminobisphosphonate drugs such as zoledronic acid [ZOL]. Non d2 gd T cells are also present in the circulation, albeit at lower levels. These cells also possess intrinsic anti-tumor activity. In order to activate both gd T cell subtypes, an immobilized pan-gd T cell receptor antibody may be used. Given the intrinsic anti-tumor activity of both d2 and non d2 gd T cells and the fact that neither subtype causes graft versus host disease, these cells are the subject of great interest as a potential off-the-shelf approach to adoptive cancer immunotherapy. 2. Recombinant human TGFb1 available from different suppliers varies in its bioactivity. We use TGFb1 from Bio-Techne and have validated its activity in house using a TGFb1 reporter system

MATERIALS AND EQUIPMENT
Note: Store TexMACS TM + medium at 4 C for up to 2 weeks.
Note: Store D10 medium at 4 C for up to 2 weeks.
Note: Store Disodium edetate (0.5 M stock) for up to 1 year at 4 C.

STEP-BY-STEP METHOD DETAILS
Timing: 15 or more days, depending on expansion methodology used Expansion of activated Vg9Vd2 T cells in the presence of TGFb1.
This step describes methods used to activate and expand Vg9Vd2 T cells over a 2 week period in serum-free medium supplemented with IL-2 and TGF-b1.
Note: Cells expanded using the combination of IL-2 and TGF-b1 are designated gd[T2] cells. If cells are expanded in the same way but using IL-2 alone, they are referred to as gd[2] cells.
Note: Our RNA sequencing data suggests that cells are not phenotypically mature at day 8 (Beatson et al., 2021). We therefore recommend completing the 15 day culture period.
1. Alternative (antibody activation method only): Coat 24, 12 or 6 well plates (depending on the numbers of PBMCs expected) overnight (12-16 h) at 4 C with 0.8 mg/mL pan gd TCR antibody (11F2 or B1 clones) contained in a volume of 1 mL, 2 mL or 3 mL PBS, respectively.
CRITICAL: We have not tested other pan gd TCR antibodies for their ability to stimulate the activation of these cells.
2. Withdraw blood (typically 45 mL) from a consenting healthy volunteer and mix with 5 mL sodium citrate in pre-prepared 50 mL Falcon tubes (stored up to 6 months at 4 C).
Note: Alternatively, blood can be collected into sodium heparin blood collection tubes.
3. Gently mix blood with an equal volume of PBS (1:1 ratio). 4. Gently layer the blood/PBS mix (using the gravity setting on a pipette controller) on top of 15 mL Ficoll Paque Plus in 50 mL Falcon tubes. 5. Centrifuge the 50 mL tubes at 750 3 g for 30 min at 20 C (acceleration and brake set to 0) to separate the PBMC cell fraction. 6. Transfer the PBMC layer, seen as an interface between the plasma and the Ficoll layer, using a sterile Pasteur pipette to a fresh 50 mL Falcon tube.
Note: After centrifugation, the layer above the PBMC fraction which contains plasma can be carefully removed using a Pasteur pipette to make the PBMC layer easier to harvest.
7. In the 50 mL tube containing the PBMCs, add PBS to a final volume of 50 mL and centrifuge at 500 3 g for 5 min. The acceleration and brake can be set back to 9. 8. Aspirate and discard the supernatant leaving the pelleted cells. 9. If more than one tube has been used (due to the volume size), resuspend the pellets in TexMACS TM + and combine in a single tube. Bring the cells up to a final volume of 50 mL of PBS for a second wash. 10. Centrifuge at 500 3 g for 5 min. 11. Aspirate and discard the supernatant. 12. Resuspend the cell pellet an appropriate volume of TexMACS TM + medium (see materials and equipment Table above; usually 10 mL per 45 mL blood).
CRITICAL: We have not tested other serum free media for their ability to support the expansion of gd T-cells.
Note: Samples are normally diluted 1 in 10 with PBS prior to cell counting.
14. Alternative (antibody activation method only -to expand both d2 and non-d2 gd T cells): Aspirate the antibody-containing PBS solution from plates set up in step 1. Wash each well once with PBS. Plate PBMC immediately at a density of 3 3 10 6 cells/mL in 4 mL TexMACS TM + medium per well of a 6 well plate. Add recombinant IL-2 (100 U/mL) and recombinant TGF-b1 (5 ng/mL). After 72 h, remove the cells are removed from this plate.
Note: We do not observe consistent differences in total yield of gd T cells achieved using the two methods although the relative proportion of non d2 gd T cells that expand is greater using the antibody method.
Note: Cytokine addition is corrected to the total volume of medium present. Where necessary, split and expand cultures into additional wells on the chosen plate type, and thereafter into T75 flasks.

OPEN ACCESS
Note: Other anti-human gd TCR antibodies suitable for flow cytometry are commercially available and may potentially be used for this step.
e. Add 5 mL FITC-conjugated isotype control to the second tube containing gd[2] and gd[T2] cells. Place tubes on ice for 30 min. f. Wash each tube with 2 mL FACS buffer and centrifuge at 400 3 g for 5 min. g. Discard supernatant and resuspend cells in 100-250 mL PBS. h. Analyze using flow cytometry.
Pause Point: Cells can be maintained in culture for up to day 21. In this case, continue the feeding regimen as described in step 16.
Note: Using this method, a median purity of 92.7% gd T cells is achieved. The lowest purity achieved at which acceptable anti-tumor activity was observed was 74.9%. Flow sorting can be used to improve purity and also to remove contaminating ab T cells and B cells, which are undesirable impurities in the event that allogeneic gd T cells are infused for therapeutic purposes.
Note: The evaluation assays described in the sections that follow are performed independently of each other, rather than representing a sequentially timed protocol.

Evaluation of in vitro real time tumor cytolytic activity of gd[T2] cells
Timing: Up to 6 days This step describes methods used to quantify the tumor cell killing activity of TGF-b-educated Vg9Vd2 T cells using the xCELLigence MP impedance analyzer.
Note: It is recommended to use tumor cell lines that have been passaged 20 times or less.
Note: DMEM medium is available from many commercial suppliers and may be used for this step.
19. After 24 h, pulse the cells with ZOL (3 mg/mL) or medium alone as a control. 20. After a further 24 h, remove 100 mL medium from each well. Add gd[T2] and gd[2] cells in 100 mL TexMACS TM + medium to achieve a final 1:1 E:T ratio.
Note: Add cells gently to minimize disruption to tumor monolayers.
21. Perform dynamic monitoring of adherent tumor viability/ proliferation using an xCELLigence MP impedance analyzer.
Note: This equipment measures cellular impedance (and thereby cellular viability) in a labelfree manner. The xCELLigence MP impedance analyzer can host up to six 96-well electronic microplates. The analyzer is maintained in an incubator (37 C, 5% CO 2 ). This step describes methods used to quantify the tumor cell killing activity of TGF-b-educated Vg9Vd2 T cells using an MTT assay. Measurement of cytokine release by these cells by ELISA is also described.
Note: It is recommended to use tumor cell lines that have been passaged 20 times or less.
23. Plate tumor cells (1 3 10 5 cells each) in a 24 well microtiter plate in 1 mL D10 medium (see materials and equipment Table above). 24. After 24 h, pulse tumor cell monolayers with the indicated concentration of ZOL or pamidronic acid (PAM).
Note: These are alternative aminobisphosphonate drugs that cause phosphoantigen accumulation in tumor cells, thereby sensitizing them to gd T cells.

After a further 24 h, remove half of the medium and add gd[T2] and gd[2] cells in 500 mL
TexMACS TM + medium to achieve a final 5:1 E:T ratio. 26. After a further 24 h, collect 200 mL supernatant from each well for cytokine analysis by ELISA, as described by the manufacturers.
Note: Supernatants can be harvested 24-72 h after addition of gd T cells. While IFN-g is readily detectable at these time points, IL-2 levels are much lower and are optimally detected at 24 h after gd T cell addition.

Seventy two hours after addition of gd[T2]
and gd[2] cells, assess residual tumor cell viability using the 3-(4,5-Dimethylthiazol-2-Yl)-2,5-Diphenyltetrazolium Bromide (MTT) assay. a. Prepare MTT to a concentration of 500 mg/mL in D10 medium (13 mL/24 well plate). b. Carefully remove media from each well of one 24 well plate (leaving two plates for 48 h and 72-h analysis). c. Pipette 500 mL of 500 mg/mL MTT solution into each well. d. Incubate at 37 C & 5% CO 2 for 1 h. e. Aspirate MTT solution from each well being careful not to disturb the monolayers. f. Add 500 mL DMSO into each well. g. Gently swirl the plate to solubilize the formazan crystals. h. Read absorbance at 560 nm using a plate reader. i. Calculate residual tumor cell viability using the following equation: ðAbsorbance of tumor + T À cells = absorbance of tumor + medium controlÞ x 100 Note: Figure 4 provides examples of tumor cell killing by gd[T2] and gd[2] cells when co-cultivated with a panel of ovarian cancer cell lines that had been pre-sensitized with the indicated concentration of PAM or ZOL. Further examples involving triple negative breast cancer cell lines are provided in (Beatson et al., 2021).

Timing: 3 days
This step describes methods used to quantify the ability of TGF-b-educated Vg9Vd2 T cells to invade basement membrane extract.
CRITICAL: An icebox should be placed within a tissue culture hood to maintain sterility and low-temperature.
31. Aspirate the blocking solution (see step 1) from the 24 well plate and replace with 650 mL TexMACS TM + medium per well. 32. Mix BME with cold TexMACS TM + media at a 1:1 ratio on ice (within hood) using a cold 1.5 mL Eppendorf and pipette tips (keep on ice).
Note: The volume prepared should allow for coating of the required number of Transwell inserts with 80 mL of the mixture. A small excess should be prepared to allow for losses during pipetting.
33. Coat ThinCert TM inserts (6.5 mm, 3 mm pore size) with 80 mL BME/TexMACS TM + solution using cold pipette tips and placed in rows A and D of the 24 well plate. 34. Place the 24 well plate in an incubator (37 C) for at least 30 min or at room temperature (approximately 20 C-22 C) for >1 h to allow the BME to solidify. 35. Remove expanded gd[T2] and gd[2] cells from culture into a 15 mL Falcon tube and count live cells using trypan blue and a hemocytometer.
Note: there is no need for trypsin or any disassociation agent as the cells do not attach to plastic.

Centrifuge expanded gd[T2]
and gd[2] cells at 350 3 g for 5 min at room temperature (approximately 20 C-22 C). Remove supernatant and resuspend the cells at a density of 1 3 10 6 /mL. 37. If antagonism of cell surface proteins is being assessed, treat gd T cells with an appropriate blocking antibody or control in PBS prior to the start of the assay.

Note:
In the example shown in Figure 5, the effect of CD11a blockade has been tested, making comparison with an isotype control at the same concentration. In this example, gd T cells are treated with 10 mg/mL of anti-CD11a or isotype control immediately prior to placement of cells in ThinCert TM .
38. If migration towards a particular factor or supernatant is being assayed, add the relevant factor to (or substitute for) medium in the appropriate well in rows B-C. This has not been tested in the example shown in Figure 5.  42. At predetermined time points, carefully remove the inserts from the media and place these in the corresponding columns in rows A and D of the 24 well plate. 43. Count the number of cells present in the medium using a hemocytometer. It is advisable to take three measurements for each well, as there can be large variations.
Note: A representative example of this protocol is shown in Figure 5. Comparison is made between gd[T2] and gd[2] cells. Inhibition of invasion by CD11a blockade has also been evaluated, implicating CD11a in this process.

Engineering of gd[T2] T cells to express a chimeric antigen receptor
Timing: Minimum 2 weeks (to allow for transduction and expansion of the cells) This step describes methods used to engineer TGF-b-educated Vg9Vd2 T cells to express a chimeric antigen receptor.
Note: For vector design protocols and transient retroviral production protocols using triple transfected 293T cells, please see Larcombe-Young et al. (currently under review as a Star Protocol). Using this method, viral titer typically ranges from 5-10 3 10 5 viral particles/mL. Viral titer is not routinely determined prior to proceeding as described below.
Note: Parallel (p)CAR is a recently described technology in which a second generation (CD28containing) CAR is co-expressed with a 4-1BB containing chimeric co-stimulatory receptor.
The pCAR-H/T parallel CAR co-targets MUC1 and ErbB dimers and is shown in schematic form in Figure 6. Transduced cells are conveniently detected using anti-EGF antibody which binds specifically to one targeting moiety present in pCAR-H/T. Further details of the pCAR-H/T pCAR are provided in (Muliaditan et al., 2021).  b. Add 200 mL RetroNectinâ (1 mg/mL stock solution) to 18 mL sterile PBS in a 50 mL Falcon tube. c. Using a Pasteur pipette, aliquot 3 mL into each well of a non-tissue culture-treated 6 well plate. d. Wrap the 6 well plate in Saran Wrap and store overnight (12-16 h) in a refrigerator at 4 C.
Note: 1 million cells are transduced in each well of the RetroNectinâ-coated 6 well plate. Calculate how many wells are needed for each construct to be transduced.
CRITICAL: RetroNectinâ can stick to certain plastics. Use polypropylene Pasteur pipettes to minimize protein loss during reagent preparation.

On the day of transduction:
a. Rapidly thaw 1.5 mL aliquots of viral supernatant in a 37 C water bath for the retroviral constructs of choice.
Note: A total of 4.5 mL viral supernatant is required per 1 3 10 6 cells to be transduced, of which 1.5 mL will be used to pre-load the appropriate well of the RetroNectinâ-coated 6 well plate.
b. Spray cryovials containing viral supernatant with 70% ethanol and transfer into a laminar flow hood. c. Pre-load the RetroNectinâ-coated 6 well plate with viral vector by removal of RetroNectinâ from each individual well and prompt addition (to avoid drying) of 1.5 mL of thawed viral supernatant. d. Incubate virus coated six well plate for 2 h or overnight (12-16 h) at 4 C. e. Retrieve activated gd[T2] cells from antibody-coated plate and transfer into a Falcon tube.
After gentle mixing by inversion, count using a hemocytometer and Trypan blue. f. Aspirate the 1.5 mL viral vector from the preloaded RetroNectinâ-coated 6 well plate. g. Add 3 mL of freshly thawed viral supernatant to each well. h. Add 1 3 10 6 activated gd[T2] cells per well of the 6 well plate. i. For non-transduced control gd[T2] cells, add 3 mL D10 medium. j. Add IL-2 to each well to a final concentration of 100 IU/mL. k. Incubate cells at 37 C & 5% CO 2 for 72 h. Note: Activated T-cells should not be washed prior to counting and transduction as this will remove potentially stimulatory factors present in the medium that they have conditioned.
Note: Do not use all activated T-cells for transduction. Untransduced T-cells will later be required, for example, as a control for flow cytometry analysis and/ or functional studies.

Every 2 days (3 days if leaving over weekends), feed gd[T2] cells with 100% volume increase of
TexMACS TM + medium, with the addition of recombinant human IL-2 to a final concentration of 100 IU/mL and recombinant human TGFb1 to 5 ng/mL. Over this interval, maintain T cells at 37 C and 5% CO 2 .

On day 15 of the culture, determine the transduction efficiency of the gd T-cells by flow cytometry.
Note: The staining procedure is summarized in Table 1. a. Transfer 2.5 3 10 5 transduced and untransduced gd[T2] cells into three separate 5 mL flow cytometry tubes. b. Wash each tube with 2 mL PBS and centrifuge at 400 3 g for 5 min. c. Discard PBS and resuspend the cell pellet in 50 mL PBS. d.
Step 1: Add 5 mL anti-human gd TCR-FITC antibody to one tube containing untransduced gd [T2] cells (Ut tube 1). e. Add 5 mL FITC-conjugated isotype control to the second tube containing untransduced gd [T2] cells (Ut tube 2). f. Add 5 mL anti-human gd TCR-FITC antibody to all three tubes containing transduced gd [T2] cells (Td tubes 1-3). g. Place tubes on ice for 30 min. h. Wash each tube with 2 mL FACS buffer and centrifuge at 400 3 g for 5 min. i. Discard PBS and resuspend cells in Ut tube 1, Ut tube 2 and Td tube 1 in 250 mL PBS. These tubes are now ready for analysis. j. Add 50 mL PBS to Ut tube 3, Td tube 2 and Td tube 3. k.
Step 2: Add 3 mL biotinylated anti-human EGF antibody to Ut tube 3 and Td tube 3 and place on ice for 30 min in the dark. l. Wash each tube with 2 mL PBS and centrifuge at 400 3 g for 5 min. m. Discard PBS and add fresh 200 mL PBS to Td tube 2. This tube is now ready for analysis. n. Discard PBS and add 50 mL to Ut tube 3 and Td tube 3. o.
Step 3: Add 3 mL Streptavidin-PE to Ut tube 3, Td tube 2 and Td tube 3 and incubate on ice for 30 min in the dark. p. Wash each tube with 2 mL PBS and centrifuge at 400 3 g for 5 min and return to ice. q. Discard PBS and resuspend cells in 250 mL PBS. r. Analyze using flow cytometry.

Note: A representative example of expression of pCAR-H/T in retrovirus transduced gd[T2]
cells is shown in Figure 6, together with efficiency of in vitro expansion of these cells and untransduced control cells. cells. The method can also be extended to assess tumor re-stimulation potential. This provides a measure of resistance of the CAR-engineered gd T cells to exhaustion, induced by repeated antigen exposure (Vardhana et al., 2020).
Note: It is recommended to use tumor cell lines that have been passaged 20 times or less.

Day 1 -prepare tumor cells:
a. Seed tumor cells at 1 3 10 5 cells in 500 mL D10 medium per well in tissue culture treated 24 well plates.
Note: Ensure that triplicate wells are plated for each T-cell population to be tested.
Note: validate tumor cell surface expression of the target antigen(s) recognized by the CAR under study prior to co-culture experiments. In this example, firefly luciferase-expressing BxPC3 tumor cells have been used which co-express both MUC1 and multiple ErbB dimers (Whilding et al., 2017) and which have limited intrinsic sensitivity to killing by gd T-cells.
Note: This provides cells for a 2:1 E:T ratio. An excess should be prepared so that cells can be processed by serial 2-fold dilution. Note: No cytokine support is provided for T cells during the co-culture assay. 50. Day 3 (Optional) -collect supernatant for cytokine analysis as per step 31. 51. Day 4 -analyze cytotoxicity.
a. Add d-luciferin at 150 mg/mL immediately prior to luminescence reading. b. Measure luminescence using appropriate apparatus and settings. c. Calculate tumor cell viability using the following equation: ðluminescence of tumor cells cultured with T cells = luminescence of untreated monolayer aloneÞ x 100%: 52. To assess tumor re-stimulation potential, repeat steps 48-51 twice per week until T cells can no longer be retrieved or less than 25% tumor cell killing occurs.
Note: Donor to donor variation is a significant issue in performing functional studies with primary human CAR T-cells. See troubleshooting section. An adequate number of biological replicates are required to ensure validity of results.  Figure 8. Tumor cells co-express red fluorescent protein (RFP) and firefly luciferase (ffLuc), achieved by retroviral transduction and flow sorting to purity (Whilding et al., 2017). Comparison is made with unmodified gd[T2] cells that have negligible cytolytic activity against these tumor cells.
Note: It is recommended to use tumor cell lines that have been passaged 20 times or less.
Note: Group sizes should be determined prior to the experiment. In the example shown in Figure 8, a pilot study with 3 mice per group is shown. In definitive studies, a power calculation should be performed. One suitable online tool is found at http://www.biomath.info/power/ ttest.htm (accessed February 1 st , 2022). See troubleshooting step.
Note: To reduce bias, the experiment should be performed in a blinded fashion whereby one team member prepares gd[T2] cells in coded tubes for administration by a blinded second team member.
Note: All in vivo experimentation must comply with local regulatory requirements. The experiment described here was performed using the authority of the U.K. Home Office project license number P23115EBF. The content of that license had been reviewed and approved by the King's College London animal welfare and ethical review body (AWERB).
Note: CAR T-cell expansion data from previous experiments is required to calculate the number of gd[T2] cells required for each group. In general, at least 10 million transduced cells can be generated from 1 million transduced gd[T2] cells over 10 days when validated viral vector is used for the gene transfer step.
54. 6-10 week-old NSG mice are used for the experiment. In general, equal numbers of male and female mice should be used.

OPEN ACCESS
Note: Adequate time must be allowed for mice to acclimatize after delivery from a commercial supplier (generally up to 7 days).
55. Day 1 -Establish BxPC3 human pancreatic ductal adenocarcinoma xenograft in NSG mice. a. Take baseline weight readings and check mice for signs of ill health prior to tumor injection. b. Inject 1 3 10 5 ffLuc/RFP-expressing BxPC3 tumor cells suspended in 100 mL sterile PBS into all mice using the i.p. route. c. Weigh mice twice weekly and check for signs of ill health daily.
Note: Mice should be humanely killed using an appropriate procedure in the event of weight loss >15%, if total tumor flux exceeds 1 3 10 10 photons per second (p/s) or if tumor growth impairs normal behavior or vital function. Abdominal distention, dyspnoea, neurological dysfunction, lameness, jaundice, piloerection, hunched posture, inability to groom, inactivity, and/ or inappetence are all additional humane end points and should prompt the humane killing of the affected animal.
a. Inject mice i.p. with 150 mg/kg D-luciferin. b. Place mice into an anesthetic induction chamber (isoflurane 2.5%, with flow rate 2 L/min) until fully anesthetized. c. Fifteen minutes after the D-luciferin injection, image the mice with an IVIS Spectrum Imaging platform (Perkin Elmer) with Living Image software, using the auto-exposure function. d. Determine total flux (p/s) by creating a region of interest over each entire mouse. e. After imaging, return the mice to their cages and monitor until they have fully recovered from anesthesia. 57. Day 11 -Perform bioluminescence imaging of mice.
a. Mice that have total flux values greater than the enrollment threshold (usually >1 3 10 7 photons/second) are distributed into treatment groups with similar average tumor burden. 58. Day 12 -Inject the appropriate dose of transduced and untransduced gd[T2] cells into each mouse using the i.p. route (10 3 10 6 cells in this example). 59. Monitor the mice daily for signs of ill health. 60. Weigh the animals twice weekly. Note: Reduced weight, piloerection, ruffled coat, poor appetite, and reduced locomotion can all be signs of graft versus host disease (GvHD). Animals should be humanely killed if they show these signs without improvement for 48 h; if they lose >15% of body weight or if they develop other GvHD related signs such as diarrhea (>48 h). 61. To monitor tumor status, weekly bioluminescence imaging is performed for the duration of the study. 62. Mice with total flux (p/s) values reaching the humane endpoint threshold (1 3 10 10 p/s) are humanely sacrificed by cervical dislocation or asphyxiation in CO 2 . 63. Tumors are fixed in 10% formalin and paraffin embedded for future analysis.
CRITICAL: Prior to starting experiments, follow Perkin Elmer protocols to determine the D-luciferin kinetic curve for your model (https://resources.perkinelmer.com/lab-solutions/ resources/docs/SOP_Determine_Luciferin_Kinetic_Curve.pdf, accessed January 2 nd , 2022). Because bioluminescence emission is tissue dependent, optimal timings between injection of D-luciferin substrate and imaging should be determined prior to starting experiments. Serial imaging is performed every 5 min for a total of 40 min after injection of D-luciferin substrate. Total flux is plotted against time and this curve allows the determination of an imaging time at which maximum total flux is observed (most commonly at 10-20 min post injection).

EXPECTED OUTCOMES
It is anticipated that gd[T2] cells will have enhanced viability and yield, accompanied by increased intrinsic anti-tumor cytolytic activity. Activation of gd[T2] cells is accompanied by increased cytokine production compared to counterparts expanded in IL-2 alone. In vivo anti-tumor activity of these cells is also expected to be increased and in a manner that can be targeted using a chimeric antigen receptor (Beatson et al., 2021).

QUANTIFICATION AND STATISTICAL ANALYSIS
Statistical significance is determined using GraphPad Prism 9 software. Statistical comparison between two groups is undertaken using an unpaired Student t-test, assuming normally distributed data. Comparison of three or more groups is performed using one-way or two-way ANOVA with multiple comparisons, when one or two independent variables were present respectively.

LIMITATIONS
Donor variation, in terms of number, purity and phenotype of ex vivo expanded gd T cells is substantial and there is a need to identify attributes/ biomarkers of the starting material that are predictive of robust expansion of cells with the greatest intrinsic anti-tumor activity. Figure 8. In vivo anti-tumor activity of pCAR H/Tengineered gd[T2] cells 1 3 10 5 ffLuc + BxPC3 tumor cells were injected i.p. in NSG mice. After 12 days, mice were treated with 10 3 10 6 untrans(duced) or pCAR-H/T transduced gd [T2] cells, making comparison with PBS. Serial BLI emission is shown (mean +/-SD, n=3). Statistical analysis was performed using an unpaired Student's t test.

OPEN ACCESS
In vivo analysis of human CAR T cell functionality using immunocompromised mice is limited by many factors. Since NSG mice do not have human hematopoietic cells or an intact immune system, it is difficult to predict how CAR T cells might overcome immunosuppressive cellular factors, particularly in the case of solid malignancies. Moreover, in an NSG model, CAR T cells may not be exposed to the same immune suppressive checkpoint molecules, cytokines, and metabolic stresses which may be present in a naturally occurring tumor.
Using healthy donors as a source of PMBCs often results in successful transduction and expansion of CAR gd[T2] cells. However, in a clinical setting when using autologous patient T cells, prior exposure to treatments such as chemotherapy and radiotherapy may compromise the quality and yield of CAR gd[T2] cells.

TROUBLESHOOTING
Problem 1 Low yields of ex vivo expanded gd T cells (steps 1-17).

Potential solution 1
It is appropriate to check starting material to quantify the number of gd T cells that are present. Upon culture on immobilized anti-gd TCR antibody, clustering and enlargement of T cell size are generally visible microscopically. Similar attributes are observed in ZOL-activated cultures after a few days. We have found that freshly isolated PBMC from healthy donors are the most reliable source of these cells. Feeder cells that provide mitogenic and/or co-stimulatory signals to the cells may also be used to enhance cell yield (Xiao et al., 2018).

Problem 2
Lack of distinctive phenotype associated with gd[T2] cells (step 17). gd[T2] cells typically have a highly distinctive phenotype with high expression of CD103, CXCR3, E-selectin binding activity and high levels of cutaneous leukocyte antigen (Beatson et al., 2021).

Potential solution 2
The quality of the TGFb1 used to supplement the culture should be checked, if necessary, using a functional assay (Abe et al., 1994). It is also important to closely follow manufacturer's instructions, especially the addition of hydrochloric acid, when reconstituting this cytokine.

Problem 3
Tumor engraftment may fail in NSG mice following missed i.p. injection. Similarly, false negative BLI emission values may be obtained for the same reason (see steps 55-63).

Potential solution 3
Very few tumor cell lines will have 100% engraftment rates. Tumor cells left in PBS for extended periods of time before injection will greatly reduce engraftment rates. It may be useful to inject tumor in 3-4 additional mice to ensure that group sizes are sufficient for meaningful analysis. Tumor growth in PBS-treated mice tends to be most homogeneous meaning that a smaller group size may be tolerated than is the case for T cell-treated mice. If all animals engraft tumor successfully, it is recommended to add additional mice to the key T cell treatment groups. Suspected failed i.p. injection of D-luciferin is generally readily identifiable because of the trend of BLI emission in affected mice. If repeat luciferin injection yields a positive reading, the spurious negative value should be omitted from data analysis.  Potential solution 4 High quality plasmids are required for efficient transfection of HEK293T cells (see Larcombe-Young et al., currently under review as a STAR Protocol). HEK293T cells should be of low passage number when transfected. gd[T2] cells should be activated on the day of gene transfer, indicated by formation of clusters/ clumps of enlarged T cells. If problems persist, it may be appropriate to titrate viral vector in order to ensure that batches contain a high titer of virus.

Problem 5
Undertaking functional in vivo studies using CAR-engineered gd[T2] cells can be unpredictable due to donor-to-donor variability (see steps 53-63). T cell fitness, state of differentiation and yield can vary significantly between donors.

Potential solution 5
The primary solution to this issue is to perform each experiment with multiple biological replicates.

Problem 6
Cytotoxicity assays are conveniently performed using luciferase assays as described in steps 48-51 of this protocol. However, this method requires the stable expression of firefly luciferase in the tumor cells.

Potential solution 6
If luciferase-expressing tumor cells are unavailable, MTT assays can be performed to quantify cytotoxicity of T cells against adherent tumor cell monolayers. This measures the reduction of tetrazolium in a manner that is dependent on intact mitochondrial function. An example of the use of this assay is provided in steps 23-27.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact Dr John Maher (john.maher@kcl.ac.uk).

Materials availability
Constructs and other reagents generated or described in this study will be made available from the lead contact for academic/noncommercial research purposes on request. Commercial use of the constructs generated, or derivatives, would be subject to a licensing agreement as intellectual property rights are in place.

Data and code availability
The datasets supporting this protocol, and constructs described or used in Figures, have not been deposited in a public repository but are available from the corresponding author upon request.