In Vivo 19F-Magnetic Resonance Imaging of Adoptively Transferred NK Cells

  • Srinivas S. Somanchi
  • Bridget A. Kennis
  • Vidya Gopalakrishnan
  • Dean A. Lee
  • James A. Bankson
Part of the Methods in Molecular Biology book series (MIMB, volume 1441)

Abstract

In order to assess the biodistribution, homing, and persistence of adoptively transferred natural killer (NK) cell immunotherapies, there is a need for imaging methodology suitable for use in preclinical studies with relevance to clinical translation. Amongst the available approaches, 19F-MRI is very appealing for in vivo imaging due to the absence of background signal, enabling clear detection of 19F labeled cells in vivo. Here we describe a methodology for in vivo imaging of adoptively transferred NK cells labeled with 19F nano-emulsion, using clinically translatable technology of 19F/1H magnetic resonance imaging.

Key words

NK cells Adoptive immunotherapy In vivo imaging 19Magnetic resonance imaging 

1 Introduction

The preclinical and clinical development of pharmacologic cancer therapeutics generally requires in vivo measures of biodistribution and pharmacokinetics/dynamics. For cancer immunotherapy with adoptively transferred effector cells, the parameters of persistence, homing, trafficking, and proliferation constitute the analogous measures of immunokinetics. Blood sampling with cell quantitation by flow cytometry is the most commonly used method to assess persistence and proliferation, but ignores tissue localization and may under or overrepresent what is actually occurring at the tumor site. This is particularly true for tumors that are not readily accessible for repeated tissue sampling, such as brain tumors.

We developed a robust system for expanding NK cells from peripheral blood using a genetically engineered feeder cell K562 Cl9.mbIL21 [1, 2], and are now using this approach to generate NK cells for clinical trials of adoptive immunotherapy in leukemias, solid tumors, and brain tumors. Radiologic methods of localizing and quantifying infused cells in vivo include nonclinical methods such as fluorescent proteins [3] or small-molecule labels [4, 5, 6, 7] and bioluminescence (e.g., firefly luciferase/luciferin) [8], and clinically applicable methods such as radioactive labeling for PET imaging using 111Indium [9, 10, 11], 18FDG [12, 13] or thymidine kinase and 18F-FEAU [14], and magnetic resonance imaging (MRI) using iron oxide [15, 16, 17, 18]. For imaging NK cells during NK cell adoptive immunotherapy, the PET approaches are toxic to NK cells (111Indium) or require gene modification (thymidine kinase) that has proven difficult in these cells, and difficulty in loading iron oxide nanoparticles may yield insufficient contrast-to-noise for imaging a low concentration of these small cells using traditional MRI. Ideally, methods that can be adapted for both preclinical and clinical studies are needed. With a drive to develop hotspot imaging with heteronuclear MRI contrast agent, 19F has emerged as a promising imaging target. Although 19F/1H imaging was first described in 1977 [19], this method has gained attention for tracking and imaging cells in recent years [20, 21, 22, 23, 24]. The main advantage of 19F as contrast agent for MRI imaging is the absence of background signal (noise) in tissue, which enables clear visualization of the labeled cells. 19F perfluorocarbon nano-emulsions have been used for in vivo tracking of immune cells such as DCs [22, 25] and T cells [26, 27]. The main innovation of the nano-emulsion, in addition to improving imaging sensitivity, is that it eliminates the need for transfection to label cells [28] in favor of simple co-incubation. This makes the method significantly more appealing for NK cell applications.

Here we describe the preclinical proof-of-principle methodology for 19F labeling and in vivo imaging of expanded NK cells by 19F MRI in a mouse model of intracranial NK cell adoptive immunotherapy. For the methods described here, we established an orthotopic model of brain tumor in mice, using cranial guide screw. The cranial guide screw method has been effectively used for establishing xenograft models for a number of brain tumor types [29, 30, 31]. For additional visual guidance on the guide screw method for intracranial xenografts, refer to the video article in the Journal of Visual Experiments [32].

2 Materials

  1. 1.

    NK cells (see Note 1).

     
  2. 2.

    NK cell medium: RPMI 1640, 10 % FBS, 1× GlutaMAX, 1× Pen/Strep. Add 50 IU/ml of IL2 to medium just before using for NK cell culture.

     
  3. 3.

    Tissue culture flasks (T25, T75).

     
  4. 4.

    Ficoll-Paque Plus (GE Healthcare Life sciences).

     
  5. 5.

    DAOY (Medulloblastoma cell line) ATCC.

     
  6. 6.

    MEM complete medium: Supplement MEM medium with 10 % FBS, 2× l-glutamine, 0.075 % sodium bicarbonate, 1 mM sodium pyruvate, 1× antibiotic–antimycotic, and 1× MEM nonessential amino acids.

     
  7. 7.

    Cell Sense 19F reagent: CS-ATM-DM Green (Celsense Inc) (see Note 2).

     
  8. 8.

    2× Lysis buffer: 2 % Triton X-100 in D2O (Deuterium Oxide; Sigma).

     
  9. 9.

    Reference standard: 0.1 % TFA (Trifluoroacetic acid; Sigma) in D2O.

     
  10. 10.

    Deuterium oxide (Sigma).

     
  11. 11.

    5 mm NMR tubes (Sigma).

     
  12. 12.

    Bruker 300 MHz DPX NMR spectrometer.

     
  13. 13.

    15 ml and 50 ml sterile conical tubes.

     
  14. 14.

    K562 cell line (ATCC).

     
  15. 15.

    Complete culture medium: RPMI 1640, 10 % FBS, 1× GlutaMAX, 1× Pen/Strep.

     
  16. 16.

    Calcein-AM (1 mg/ml stock) (Life Technologies).

     
  17. 17.

    2 % Triton X-100: Prepare a 2 % Triton X-100 solution in complete culture medium and store at 4 °C.

     
  18. 18.

    4–8-week-old NOD scid gamma (NSG) Mice (Jackson Laboratories).

     
  19. 19.

    SD-80 screwdriver—Part: 26030 (Plastics One, Inc).

     
  20. 20.

    3 M® tissue adhesive.

     
  21. 21.

    Drill Bit #56 (High Speed Steel)—Part: D#56 (Plastics One, Inc).

     
  22. 22.

    DH-1 Pin Vise Starrett (0.030–0.062)—Part: 50600 (Plastics One, Inc).

     
  23. 23.

    SU1 Guide Screw W-0.5 mm hole; Nylon—Part: C212GN (Plastics One, Inc).

     
  24. 24.

    Dummy (stylet) 0.018 in-0.45 mm; Nylon—Part: C212SDN (Plastics One, Inc).

     
  25. 25.

    7T Biospec Small Animal MRI system with BG6 gradients and a 1H MRI volume coil with 35 mm inner diameter (ID) (Bruker Biospin Corp., Billerica, MA).

     
  26. 26.

    Quadrature 19F volume coil with 35 mm ID (Rapid MR International LLC, Columbus, OH).

     
  27. 27.

    Spherical NMR Bulb (529-A, Wilmad-Labglass, Vineland, NJ).

     

3 Methods

3.1 Establishing Murine Intracranial Tumor Model

3.1.1 Placing Guide Screws in Cranium

  1. 1.

    The number of mice to be used for any experiment should be determined based upon statistical criteria.

     
  2. 2.

    Anesthetize the mice using 3–4 % inhalant isoflurane (see Note 3).

     
  3. 3.

    Shave the scalps of mice from the nape of the neck to between the eyes using an electric shaver, clean the surgical area once with a sterile betadine swab followed by wiping one time with a sterile 70 % isopropyl alcohol pad.

     
  4. 4.

    With a sterile scalpel make an incision of approximately 0.5 inch in caudal to cranial direction, over the site of guide screw placement (see Note 4).

     
  5. 5.

    Carefully drill a hole in the skull at the site of guide screw placement using DH-1 Pin Vise Starrett hand drill and drill bit.

     
  6. 6.

    Secure the nylon guide screw using a screwdriver until the guide screw is flush with the surface of the skull. It is important to stop rotating the screwdriver once the screw has met the resistance of the surface of the skull in order to prevent stripping or collapse of the burr hole.

     
  7. 7.

    Then, place the stylet in the shaft of the guide screw to close the opening.

     
  8. 8.

    Close the incision using forceps and seal the wound with 3 M® tissue adhesive, covering the guide screw with the scalp.

     
  9. 9.

    Allow the animals to rest for 7 days after surgery.

     

3.1.2 Intracranial Injection of Tumor Cells

Culture and maintain the tumor cell line of choice in appropriate culture medium in sterile culture condition and periodically test for mycoplasma. Use tumor cell lines expressing bioluminescence or fluorescence markers for convenience of imaging the tumor during the experiments. The protocol is described here for DAOY medulloblastoma cell line. If a different cell line is used, an initial tumor cell dose titration is recommended for establishing the tumor model, as different cell lines have different in vivo growth kinetics.
  1. 1.

    Culture and maintain DAOY cells in MEM complete medium. Passage the cells at 80 % confluency at a split ratio of 1:6.

     
  2. 2.

    On the day of injection (day 7 after guide screw placement), trypsinize tumor cells and wash twice in PBS. Resuspend cells in PBS and pass through 70 μm nylon mesh to remove cell aggregates, and perform cell counts.

     
  3. 3.

    Recover desired number of cells for injection into mice into a fresh tube (see Note 5), spin at 400 × g for 5 min and resuspend the pellet at 10 × 106 cells/ml in PBS so that 5 μl would contain 50,000 cells for intracranial injection.

     
  4. 4.

    To inject the tumor cells, anesthetize and prepare mice for surgery as described in Subheading 3.1.1, step 2.

     
  5. 5.

    Use a sterile scalpel to make an incision directly over the guide screw of enough length to expose the guide screw.

     
  6. 6.

    Remove the stylet.

     
  7. 7.

    Use a blunt-tipped Hamilton syringe and stereotactic device to inject up to 5 μL total volume of tumor cells at a steady rate of 0.5 μL/min. It is important that the injected cells are placed at least one millimeter beyond the end of the guide screw.

     
  8. 8.

    Once the entire volume has been delivered, rest the syringe in position for 1 min before removal in order to allow all injected cells to settle.

     
  9. 9.

    Remove the Hamilton syringe and replace the stylet. Close the incision using forceps and seal the wound with 3 M® tissue adhesive, covering the guide screw with the scalp.

     
  10. 10.

    Allow the tumor to establish for about 7 days (see Note 6), and randomly separate the mice into control and treatment groups.

     

3.2 In Vivo Infusion of 19F Labeled NK Cells

3.2.1 Labeling NK Cells with 19F

  1. 1.

    One day before NK cell infusion to mice, thaw a vial of expanded NK cells in a 37 °C water bath and transfer cells to a 15 ml conical tube containing 10 ml of pre-warmed NK cell medium. Spin the cells at 400 × g for 5 min. (If using NK cells that are in culture, start the process from step 5).

     
  2. 2.

    Remove the supernatant and resuspend cells in 10 ml of NK cell medium. Count and check for viability of NK cells using trypan blue exclusion method.

     
  3. 3.

    Transfer NK cells to an appropriate culture flask (see Note 7) and add NK cell medium to adjust cell density to 1 × 106/ml (see Note 8).

     
  4. 4.

    Incubate the cells overnight in CO2 incubator at 37 °C.

     
  5. 5.

    Next day, prior to 19F labeling, count the NK cells and check for viability by trypan blue method.

     
  6. 6.

    If the viability of NK cells is above 90 %, proceed to step 14.

     
  7. 7.

    If the viability of NK cells is below 90 %, perform a Ficoll-Paque density centrifugation to remove nonviable cells and debris.

     
  8. 8.

    In a 50 ml conical tube add 15 ml of Ficoll-Paque and carefully layer up to 35 ml of NK cells on the Ficoll-Paque layer (for smaller NK cell culture volumes use a 15 ml conical tube to maximize recovery).

     
  9. 9.

    Spin at 400 × g for 20 min without brakes.

     
  10. 10.

    Recover the NK cells from the Ficoll–medium interface and transfer to a fresh 15 ml conical tube.

     
  11. 11.

    To wash cells, add PBS to fill the tube and spin at 400 × g for 5 min.

     
  12. 12.

    Repeat the wash two more times.

     
  13. 13.

    Resuspend NK cells in NK cell medium, count and check viability by trypan blue exclusion.

     
  14. 14.

    Recover the desired number of NK cells (see Note 9) for labeling with 19F, spin at 400 × g for 5 min.

     
  15. 15.

    Discard supernatant and resuspend at 3 × 106/ml using NK cell medium (see Note 10).

     
  16. 16.

    Seed cells in appropriate culture plate based on cell volume (see Note 11).

     
  17. 17.

    Add Cell Sense 19F reagent (CS-ATM DM Green) to the cells at 5 mg/ml (see Note 12) and mix gently and thoroughly using a P-1000 pipette.

     
  18. 18.

    Incubate the cells overnight (16–18 h) at 37 °C in a CO2 incubator.

     
  19. 19.

    On the day of infusion, mix and transfer the NK cells to a 15 ml conical tube, rinse the wells thoroughly with NK cell medium and transfer to conical tube to completely recover the NK cells.

     
  20. 20.

    To wash off excess 19F, fill the tube with NK cell medium and spin at 400 × g for 5 min.

     
  21. 21.

    Repeat the wash step 2 more times.

     
  22. 22.

    An aliquot of the 19F labeled NK cells can be used to determine the number of fluorine atoms incorporated per cell by NMR (see Subheading 3.4) and to determine the effect of 19F labeling on NK cell cytolytic function (see Subheading 3.5).

     

3.2.2 Intracranial NK Cell Infusion

  1. 1.

    Resuspend the 19F labeled NK cell pellet at 1 × 106 cells/3 μl of PBS for intracranial injection into mice.

     
  2. 2.

    Shave the scalps of mice and clean scalp with sterile betadine swab followed by 70 % isopropyl alcohol.

     
  3. 3.

    Make an incision using sterile scalpel to expose the guide screw.

     
  4. 4.

    Infuse 19F labeled NK cells intratumorally using the steps 58 described in Subheading 3.1.2.

     
  5. 5.

    Rest the mice for 2 h to recover from the surgery.

     
  6. 6.

    Image mice to detect the infused 19F labeled NK cell by MRI.

     

3.3 In Vivo MRI Imaging of 19F Labeled NK Cells

Magnetic resonance imaging (MRI) is a well-known diagnostic modality that provides good image resolution and exquisite soft-tissue contrast. MRI is based on measurement of the spatial distribution of hydrogen nuclei. Sensitivity is high because 1H is the most abundant nucleus in vivo and because its nuclear magnetic resonance (NMR) signal strength is highest of all stable isotopes. 19F has the second highest relative NMR signal strength, and low abundance in normal tissue ensures little background signal against which 19F-labeled cells must be detected. This procedure outlines a method that can be used for coregistered imaging of labeled cells with anatomic reference using traditional MRI.
  1. 1.

    Anesthetize the mouse that is to be imaged (see Note 13) and place on an imaging sled that includes a method for delivering anesthesia (if an inhalable anesthetic is used) while maintaining body temperature.

     
  2. 2.

    Attach sensors for remote monitoring of physiological status (see Note 14). Make sure that the region of the body that will be scanned is immobilized to minimize MRI artifacts that are caused by motion. Place an external reference containing a small amount of 19F (Cell Sense reagent) near the target imaging site, in a small NMR bulb.

     
  3. 3.

    Place the mouse at isocenter inside the MRI magnet, at the center of the imaging gradients and within the RF coil(s) (see Note 15).

     
  4. 4.
    Acquire localizing scans to confirm that the animal is positioned correctly. A 3-plane fast spin-echo (FSE) sequence (TEeff = 46 ms, TR = 2 s, ETL = 8, 128 × 128 image matrix over 45 mm × 45 mm field of view), for example, allows visualization of animal position after; see Fig. 1). Adjust as necessary to center the target anatomy within the imaging system.
    Fig. 1

    Representative localizer scans. Interleaved (a) axial, (b) sagittal, and (c) coronal images show that target anatomy (brain) is positioned correctly within the imaging system

     
  5. 5.

    Once position is confirmed, make sure that the RF coil is tuned and matched. Then adjust the center frequency, shims, and calibrate the RF excitation power for the MRI scanner. Many of these optimizations are performed automatically during prescan.

     
  6. 6.

    Acquire anatomic reference images using traditional 1H MRI. The exact configuration of sequences that highlights target anatomy will depend on the anatomic region of interest and the field strength for the scanner. For visualizing NK cells within the murine brain, a single 3D T2-weighted fast spin-echo imaging sequence (TEeff = 72 ms, TR = 1500 ms, ETL = 16, 128 × 128 × 32 image matrix covering a 3 cm × 3 cm × 3 cm field-of-view) provided excellent anatomic context with approximately 6.5 min scan time.

     
  7. 7.

    Without moving the animal, slide the 1H MRI RF coil out of the magnet and replace it with the 19F volume coil. Change the reference frequency of the MRI scanner to correspond with 19F, and confirm that the coil is tuned and matched. Set the excitation calibration (see Note 16). Fine-tune the 19F reference frequency by running a quick pulse-acquire spectroscopy sequence (90 excitation, TR = 1000 ms, 2048 point readout over 25 kHz bandwidth, approximately 10 averages to ensure sufficient signal-to-noise to measure the resonance frequency of the reagent in tissue and/or within the external NMR bulb).

     
  8. 8.

    Acquire 19F images that are co-registered with anatomic reference images using the 3D fast spin-echo sequence, tailored for 19F (TEeff = 71.5 ms, TR = 700 ms, ETL = 24, 80 signal averages, 32 × 32 × 32 matrix size covering 3 cm × 3 cm × 3 cm field-of-view) (see Note 17). Make sure that the slice prescription and imaging geometry match the anatomic references.

     
  9. 9.
    Overlay 19F images (step 8) on the anatomic reference images (step 6) using image-processing software such as Matlab (Mathworks) or Photoshop (Adobe Systems Inc.) see Fig. 2.
    Fig. 2

    Representative overlay of 19F images on T2-weighted anatomic references. Left image shows intracranial signal after injection of 19F-labeled NK cells. At right, agreement between proton and 19F images in external reference demonstrates good co-registration

     

3.4 Determining 19F Labeling of NK Cells

The NK cells labeled by the dual-mode imaging reagent CS-ATM DM Green described in Subheading 3.2.1 can be assessed for 19F labeling qualitatively by flow cytometry and quantitatively by nuclear magnetic resonance (NMR).
  1. 1.
    To qualitatively assess labeling by flow cytometry, recover about 5 × 105 unlabeled NK (as control) and 19F labeled NK cells in NK cell medium into separate FACS tubes. Run the samples on flow cytometer in the FITC channel (see Note 18) (Fig. 3a).
    Fig. 3

    Determining 19F labeling of NK cells. (a) Qualitative assessment of CS-ATM DM green (19F) labeling of NK cells by flow cytometry compared to unstained control NK cells. (b) Quantitative analysis of 19F labeling of NK cells, showing the number of fluorine atoms/NK cell after overnight incubation with CS-ATM DM Green

     
  2. 2.

    To determine the number of fluorine atoms incorporated per cell by NMR, collect 3 × 106 19F labeled NK cells in a 15 ml conical tube add PBS to fill the tube.

     
  3. 3.

    Spin cells at 400 × g for 5 min.

     
  4. 4.

    Aspirate the supernatant and repeat wash step four more times for a total of five washes.

     
  5. 5.

    Aspirate the PBS as completely as possible taking care not to lose NK cells.

     
  6. 6.

    To the NK cell pellet add 100 μl of 2× lysis buffer, and mix well using P-200 pipette to promote complete lysis of cells.

     
  7. 7.

    To the lysate, add 200 μl of D2O.

     
  8. 8.

    Prior to running the sample on NMR, add 200 μl of 0.1 % TFA as an internal reference standard for fluorine, and mix thoroughly using a P-200 pipette.

     
  9. 9.

    Transfer the sample to 5 mm NMR tubes and acquire 19F NMR spectrum (see Note 19).

     
  10. 10.

    Calculate the number of fluorine atoms per cell using the following formula: (Fig. 3b).

     
$$ {}^{19}\mathrm{F}_{\mathrm{a}\mathrm{toms}/\mathrm{cell}}=3{I}_{\mathrm{s}}{M}_{\mathrm{r}}{N}_{\mathrm{a}}/{I}_{\mathrm{r}}{N}_{\mathrm{c}} $$

Is = Integrated area of major peak of the cell pellet

Mr = Moles of TFA reference (1.75 × 10−6 moles for 200 μl of 0.1 % TFA)

Na = Avogadro’s number

Ir = Integrated area under TFA reference peak

Nc = number of cells in the pellet

3.5 Effect of 19F Labeling on NK Cell Function

It is essential to understand the effect of 19F labeling on NK cells cytolytic function, as this could impact their therapeutic potential. Numerous methods can be used to determine NK cell cytotoxicity such as Chromium-51 release assay, LDH release assay, nonradioactive Europium release and flow cytometry based methods, to name a few. This protocol describes calcein release assay for quantitating cytotoxicity of NK cell with and without 19F labeling using K562 as the target tumor cell line.
  1. 1.

    Maintain K562 cell line in complete culture medium and passage at 1:6 split ratio when the cell density exceeds 0.6 × 106 cells/ml.

     
  2. 2.

    To stain K562 with calcein, count and transfer 2 × 106 K562 cells (see Note 20) into a 15 ml conical tube and spin at 400 × g for 5 min. Resuspend cell pellet in 2 ml of fresh complete culture medium (cell density of 1 × 106/ml).

     
  3. 3.

    Add stock calcein (1 mg/ml) to the cells at 1:333 fold dilution (3 μl/ml). Incubate for 30 min at 37 °C in a CO2 incubator, and vortex cells every 5 min.

     
  4. 4.

    Wash the cells three times with 10 ml of complete culture medium to remove excess calcein-AM, spin at 400 × g for 5 min. Resuspend calcein labeled K562 in 10 ml medium and transfer to a 50 ml conical tube, perform a cell count and add medium to adjust cell density to 1 × 105 cells/ml.

     
  5. 5.

    Count and resuspend NK cells with and without 19F labeling in complete culture medium at 1 × 106 cells/ml (see Note 21).

     
  6. 6.

    Seed 200 μl of NK cells per well in 3 wells, in a “U” bottom 96-well plate (e.g., unlabeled NK cells in A1, B1, C1, and 19F labeled NK cells in A7, B7, C7).

     
  7. 7.

    Add 100 μl of complete culture medium to wells A2–C6 and A8–C12.

     
  8. 8.

    Perform a twofold serial dilution of NK cells (by transferring 100 μl of NK cells serially across columns) (see Note 22).

     
  9. 9.

    Add 100 μl of complete culture medium to 6 wells (e.g., E1–E6) for spontaneous release control.

     
  10. 10.

    Add 100 μl of 2 % Triton X -100 to 6 wells (e.g., G1–G6) for maximum release control (see Note 23).

     
  11. 11.

    Add 100 μl of calcein loaded K562 to the wells containing NK cells and the spontaneous control and maximum release control wells. Mix gently.

     
  12. 12.

    Spin the plate at 100 × g for 1 min.

     
  13. 13.

    Incubate the plate for 4 h at 37 °C in a CO2 incubator.

     
  14. 14.

    After 4 h, remove the plate, gently mix the contents of each well, in order to evenly distribute the released calcein using a P-200 pipette and spin the plate at 400 × g for 1 min.

     
  15. 15.

    Carefully aspirate 100 μl of the supernatant using P-200 pipette and transfer into a clear flat bottom 96-well plate (see Note 23).

     
  16. 16.

    Read fluorescence intensity using a fluorescence spectrophotometer (excitation filter 485 nm: emission filter 530 nm).

     
  17. 17.
    Calculate percent specific lysis of NK cells using the formula [(Test release—Spontaneous release)/(Maximum release—Spontaneous release)] × 100 (see Note 24) (Fig. 4).
    Fig. 4

    Representative NK cell cytotoxicity assay. Cytotoxicity of NK cells labeled with 7.5 mg/ml of 19F, note that the NK cells continue to be highly cytotoxic against K562 cell line after labeling with 19F (CS-ATM DM Green) reagent

     

4 Notes

  1. 1.

    NK cell lines, primary or expanded human NK cells can be used for this study depending on the user’s interest or research goal. The protocol described here is for expanded NK cells. If primary NK cells or NK cell lines are used for experiments, optimize 19F labeling as per see Note 10.

     
  2. 2.

    Dual mode MRI/Optical reagent is used for this protocol. This reagent contains 19F as well as fluorescein within the emulsion to facilitate qualitative assessment of cell labeling by flow cytometry as well as quantitative assessment of 19F labeling (number of fluorine atoms per cell) by fluorine Nuclear Magnetic Resonance (NMR).

     
  3. 3.

    Animals should be continually monitored for anesthetic depth with toe or tail pinch throughout the duration of anesthesia and surgical procedures.

     
  4. 4.
    Surgical coordinates used for various locations in the mouse brain:
    • Cerebellum: 2 mm posterior to lambdoid suture, 2 mm lateral of midline, 2–3 mm below skull surface

    • Cerebrum: 2 mm anterior to lambdoid suture, 2 mm lateral of midline, 2–3 mm below skull surface

    • 4th Ventricle: 1.5 mm posterior to lambdoid suture, 0.5 mm lateral of midline, 4 mm below skull surface

    • Brainstem/Pons: 1.5 mm posterior to lambdoid suture, 0.5 mm lateral of midline, 5 mm below skull surface

     
  5. 5.

    It is strongly suggested that a tumor growth curve is established prior to any therapeutic experiment for each separate cell line and at various numbers of injected cells. This will assist in determining the number of cells to be injected for the calculated experiment timeline and tumor burden.

    For the experiment detailed in this chapter, 50,000 DAOY medulloblastoma cells were injected because these cells establish tumors and progress extremely quickly. However, in cell lines that take significantly longer to establish tumors, cell numbers will be much higher based on established growth curves.

     
  6. 6.

    Criteria for tumor establishment can be measured in a number of ways. If the tumor cells injected are not labeled for fluorescence or luminescence (e.g., firefly luciferase) imaging, magnetic resonance imaging can be used to monitor tumor establishment. However, whether or not imaging is done to observe tumor establishment, mice should be monitored for neurological, behavioral, and overall health effects of tumor burden. Tumors are allowed to grow until tumor burden is clinically visible. Mice are then euthanized and the brain can be collected for histopathological analyses.

     
  7. 7.

    Use T25 flask for culture volume from 5 to 10 ml, and for volume between 10 and 40 ml, use a T75 flask. Over 40 ml of culture volume use multiple flasks. Incubate the flasks upright in the CO2 incubator.

     
  8. 8.

    We expand NK cells in the presence of 50 IU/ml of IL2; therefore, we supplement the medium with 50 IU/ml of IL2 to provide the cytokine support to NK cells after thawing. If a different amount of IL2 was used for activating or expanding NK cells prior to freezing them, then use the same concentration of IL2 in the medium after thawing.

     
  9. 9.

    Always stain more cells than absolutely needed for injection/experiments as loss of cells is expected during wash steps or due to loss of viability in culture.

     
  10. 10.

    We tested various cell densities and reagent concentration to optimize 19F labeling of the cells in the minimum possible volume (that did not affect the NK cell viability and labeling efficiency). If primary NK cells or NK cell lines are used, we recommend optimizing the 19F labeling initially with cell numbers ranging from 1 × 106/ml to 5 × 106/ml.

     
  11. 11.

    For staining NK cells with 19F reagent, we normally seed up to 1 ml of cells/well of 24-well plate; 2 ml of cells/well of 12-well plate; and 3 ml/well of 6-well plate. For volume more than 3 ml use multiple wells of a 6-well plate.

     
  12. 12.

    We performed labeling optimization (at 3 × 106 cells/ml) using 19F at concentrations of 0, 2.5, 5, and 7.5 mg/ml for NK cells expanded on K562 Cl9.mbIL21 feeder cell platform. The 19F labeling of these cells saturated at 5 mg/ml as analyzed by NMR, using this protocol. Similar optimization should be performed if primary NK cells, NK cell activated and expanded by other methods or NK cell lines are used for 19F labeling studies.

     
  13. 13.

    Inhalable anesthetics such as isoflurane are convenient because they allow remote adjustment, without having to interrupt the imaging protocol or move the animal. Unfortunately, isoflurane contains 19F, which will concentrate in tissue to give a competing signal. Injectable anesthetics do not add a 19F signal, but dosing must be carefully administered to ensure that the animal is sedated long enough for the imaging measurement.

     
  14. 14.

    Sensors may include bellows for monitoring respiratory rate, electrocardiogram (ECG) for monitoring heart rate, and/or a thermocouple to monitor body temperature. All of these must be MRI compatible. Physiological monitoring systems for animal models are commercially available, such as the Model 1030 Monitoring & Gating System from Small Animal Instruments, Inc.

     
  15. 15.

    The sensitivity of 19F scans needs to be enhanced as much as possible by using good quality coils for signal excitation and detection. There is a wide range of possible coil configurations, including the use of large “volume” coils that give relatively uniform sensitivity, small “surface” coils that provide high sensitivity over a small region, and arrays of surface coils that provide good sensitivity over an extended range. Multinuclear measurements can be particularly complex, since coils are needed that are sensitive to the resonance frequencies of both 1H, for anatomic reference using traditional MRI, and 19F for cell tracking. The approach described in this work utilizes volume coils for both nuclear targets. The two volume coils have the same dimensions and are simply exchanged in order to swap between imaging targets.

     
  16. 16.

    The 19F signal is generally not strong enough to permit use of automatic prescan adjustments for calibration of the excitation power. Calibration measurements can be performed prior to imaging using a 19F phantom, or if a reference standard is present during the imaging measurement, signal may be strong enough for manual calibration. A simple pulse-acquire MR spectroscopy sequence can be used for this purpose. The repetition time for the sequence should be set at greater than several times the characteristic spin–lattice (T1) relaxation time of the 19F agent at the field strength for the MRI system. Then, the excitation power for the pulse-acquire sequence can be set very low and then increased until the observed signal reaches a maximum, which corresponds to a 90° excitation. A slightly more accurate approach would involve several measurements ranging from low power to approximately 180° excitation (when the signal vanishes) and fitting observations to the expected sinusoidal signal response curve. Calibration may need to be adjusted if coil loading changes substantially between measurement conditions.

     
  17. 17.

    The relaxation characteristics of the 19F agent should be measured prior to imaging. The spin–lattice (T1) and spin–spin (T2) relaxation times will vary with the field strength of the MRI magnet. There are a number of methods that can be used for making these measurements. We used a saturation-recovery imaging sequence to measure the T1 of CelSense at 7 T, and a Carr–Purcell–Meiboom–Gill (CPMG) spin-echo sequence to measure T2. With T1 ≅ 550 ms and T2 ≅ 240 ms, efficient sequences with a long echo train and relatively short repetition times (TR) can be used for imaging with good sensitivity.

     
  18. 18.

    Select live cell population in forward and side scatter plot and display this population on FITC channel of the flow cytometer. The fluorescence method is strictly a qualitative assessment of labeling and is not a substitute for quantitative estimation of 19F labeling by NMR analysis, which can provide number of fluorine atoms incorporated per cell, which could subsequently be used to calculate the number of NK cells at a given location based on 19F signal strength in vivo.

     
  19. 19.

    Since most institutions offer NMR as a core service, to acquire and analyze 19F spectrum by NMR, we recommend the users to refer to their Institutional core resources. The NMR should be equipped with probe for observing 19F.

     
  20. 20.

    The recommended number of K562 cells is for one full 96-well plate cytotoxicity assay. Since 10,000 target cells are seeded per well—a minimum of 1 × 106 cells are needed for a full 96-well plate, including +4 for pipetting errors; however, 2 × 106 cells are recommended here as a twofold excess to accommodate for cell loss during wash steps and for accommodating multichannel pipetting of target cells from a media basin.

     
  21. 21.

    Note that the target cells are resuspended at 1 × 105 cells/ml and NK cells are resuspended at 1 × 106 cells/ml, this recommendation is for setting up the cytotoxicity assay at effector to target (E:T) ratios starting at 10:1. If primary NK cells or NK cell lines are used, the E:T ratio can be changed as desired, which will then dictate the density for resuspending NK cells (e.g., 2 × 106 NK cells/ml for a 20:1 E:T ratio). For the recommended assay 600,000 NK cells will be used per cytotoxicity condition; therefore, about 650,000 NK cells will be needed per condition to accommodate for errors.

     
  22. 22.

    After serial dilution the wells A6, B6, C6 and A12, B12, C12 will have 200 μl of cells. Discard 100 μl of cells from these wells.

     
  23. 23.

    A blank row is recommended between test and controls to prevent accidental spill over during setup and fluorescence bleed-over during plate reading.

     
  24. 24.

    A well optimized 19F labeling of NK cells should have sufficient fluorine atoms to provide strong signal in MRI for in vivo imaging of NK cells, with minimal impact on NK cell cytolytic function.

     

Notes

Acknowledgments

This work was supported in part by funding from the Addis Faith Foundation to VG, CURE Childhood Cancer funding to DAL, and by MD Anderson Cancer Center’s Core Grant (P30-CA016672).

References

  1. 1.
    Somanchi SS, Senyukov VV, Denman CJ et al (2011) Expansion, purification, and functional assessment of human peripheral blood NK cells. J Vis Exp 48, e2540. doi: 10.3791/2540 Google Scholar
  2. 2.
    Denman CJ, Senyukov VV, Somanchi SS et al (2012) Membrane-bound IL-21 promotes sustained ex vivo proliferation of human natural killer cells. PLoS One 7, e30264CrossRefPubMedPubMedCentralGoogle Scholar
  3. 3.
    Edinger M, Cao YA, Verneris MR et al (2003) Revealing lymphoma growth and the efficacy of immune cell therapies using in vivo bioluminescence imaging. Blood 101:640–648CrossRefPubMedGoogle Scholar
  4. 4.
    Tavri S, Jha P, Meier R et al (2009) Optical imaging of cellular immunotherapy against prostate cancer. Mol Imaging 8:15–26Google Scholar
  5. 5.
    Leung K (2009) DiD-labeled anti-EpCAM-directed NK-92-scFv(MOC31) zeta cells. 2009 Aug 28 [Updated 2009 Sep 30]. In: Molecular imaging and contrast agent database (MICAD) [Internet]. National Center for Biotechnology Information (US), Bethesda; 2004–2013. http://www.ncbi.nlm.nih.gov/books/NBK23559/
  6. 6.
    Lim YT, Cho MY, Noh YW et al (2009) Near-infrared emitting fluorescent nanocrystals-labeled natural killer cells as a platform technology for the optical imaging of immunotherapeutic cells-based cancer therapy. Nanotechnology 20:475102CrossRefPubMedGoogle Scholar
  7. 7.
    Youniss FM, Sundaresan G, Graham LJ et al (2014) Near-infrared imaging of adoptive immune cell therapy in breast cancer model using cell membrane labeling. PLoS One 9, e109162CrossRefPubMedPubMedCentralGoogle Scholar
  8. 8.
    Olson JA, Zeiser R, Beilhack A et al (2009) Tissue-specific homing and expansion of donor NK cells in allogeneic bone marrow transplantation. J Immunol 183:3219–3228CrossRefPubMedPubMedCentralGoogle Scholar
  9. 9.
    Meller B, Frohn C, Brand JM et al (2004) Monitoring of a new approach of immunotherapy with allogenic (111)In-labelled NK cells in patients with renal cell carcinoma. Eur J Nucl Med Mol Imaging 31:403–407CrossRefPubMedGoogle Scholar
  10. 10.
    Matera L, Galetto A, Bello M et al (2006) In vivo migration of labeled autologous natural killer cells to liver metastases in patients with colon carcinoma. J Transl Med 4:49CrossRefPubMedPubMedCentralGoogle Scholar
  11. 11.
    Wagstaff J, Gibson C, Thatcher N et al (1982) A method for studying the dynamics of the primary migration of human lymphocytes using Indium-iii oxine cell labelling. Adv Exp Med Biol 149:153–160CrossRefPubMedGoogle Scholar
  12. 12.
    Meier R, Piert M, Piontek G et al (2008) Tracking of [18F]FDG-labeled natural killer cells to HER2/neu-positive tumors. Nucl Med Biol 35:579–588CrossRefPubMedGoogle Scholar
  13. 13.
    Melder RJ, Elmaleh D, Brownell AL et al (1994) A method for labeling cells for positron emission tomography (PET) studies. J Immunol Methods 175:79–87CrossRefPubMedGoogle Scholar
  14. 14.
    Dotti G, Tian M, Savoldo B et al (2009) Repetitive noninvasive monitoring of HSV1-tk-expressing T cells intravenously infused into nonhuman primates using positron emission tomography and computed tomography with 18F-FEAU. Mol Imaging 8:230–237PubMedPubMedCentralGoogle Scholar
  15. 15.
    Sheu AY, Zhang Z, Omary RA et al (2013) MRI-monitored transcatheter intra-arterial delivery of SPIO-labeled natural killer cells to hepatocellular carcinoma: preclinical studies in a rodent model. Invest Radiol 48:492–499CrossRefPubMedPubMedCentralGoogle Scholar
  16. 16.
    Mallett CL, McFadden C, Chen Y et al (2012) Migration of iron-labeled KHYG-1 natural killer cells to subcutaneous tumors in nude mice, as detected by magnetic resonance imaging. Cytotherapy 14:743–751CrossRefPubMedGoogle Scholar
  17. 17.
    Daldrup-Link HE, Meier R, Rudelius M et al (2005) In vivo tracking of genetically engineered, anti-HER2/neu directed natural killer cells to HER2/neu positive mammary tumors with magnetic resonance imaging. Eur Radiol 15:4–13CrossRefPubMedGoogle Scholar
  18. 18.
    Meier R, Golovko D, Tavri S et al (2011) Depicting adoptive immunotherapy for prostate cancer in an animal model with magnetic resonance imaging. Magn Reson Med 65:756–763CrossRefPubMedGoogle Scholar
  19. 19.
    Holland GN, Bottomley PA, Hinshaw WS (1977) 19F magnetic resonance imaging. J Magn Reson 28:133–136Google Scholar
  20. 20.
    Partlow KC, Chen J, Brant JA et al (2007) 19F magnetic resonance imaging for stem/progenitor cell tracking with multiple unique perfluorocarbon nanobeacons. FASEB J 21:1647–1654CrossRefPubMedGoogle Scholar
  21. 21.
    Boehm-Sturm P, Mengler L, Wecker S et al (2011) In vivo tracking of human neural stem cells with 19F magnetic resonance imaging. PLoS One 6, e29040CrossRefPubMedPubMedCentralGoogle Scholar
  22. 22.
    Ahrens ET, Flores R, Xu H et al (2005) In vivo imaging platform for tracking immunotherapeutic cells. Nat Biotechnol 23:983–987CrossRefPubMedGoogle Scholar
  23. 23.
    Bulte JW (2005) Hot spot MRI emerges from the background. Nat Biotechnol 23:945–946CrossRefPubMedGoogle Scholar
  24. 24.
    Srinivas M, Heerschap A, Ahrens ET et al (2010) (19)F MRI for quantitative in vivo cell tracking. Trends Biotechnol 28:363–370CrossRefPubMedPubMedCentralGoogle Scholar
  25. 25.
    Helfer BM, Balducci A, Nelson AD et al (2010) Functional assessment of human dendritic cells labeled for in vivo (19)F magnetic resonance imaging cell tracking. Cytotherapy 12:238–250CrossRefPubMedGoogle Scholar
  26. 26.
    Srinivas M, Morel PA, Ernst LA et al (2007) Fluorine-19 MRI for visualization and quantification of cell migration in a diabetes model. Magn Reson Med 58:725–734CrossRefPubMedGoogle Scholar
  27. 27.
    Kadayakkara DK, Beatty PL, Turner MS et al (2010) Inflammation driven by overexpression of the hypoglycosylated abnormal mucin 1 (MUC1) links inflammatory bowel disease and pancreatitis. Pancreas 39:510–515CrossRefPubMedPubMedCentralGoogle Scholar
  28. 28.
    Janjic JM, Srinivas M, Kadayakkara DK et al (2008) Self-delivering nanoemulsions for dual fluorine-19 MRI and fluorescence detection. J Am Chem Soc 130:2832–2841CrossRefPubMedGoogle Scholar
  29. 29.
    Aoki Y, Hashizume R, Ozawa T et al (2012) An experimental xenograft mouse model of diffuse pontine glioma designed for therapeutic testing. J Neurooncol 108:29–35CrossRefPubMedGoogle Scholar
  30. 30.
    Brockmann MA, Westphal M, Lamszus K (2003) Improved method for the intracerebral engraftment of tumour cells and intratumoural treatment using a guide screw system in mice. Acta Neurochir (Wien) 145:777–781CrossRefGoogle Scholar
  31. 31.
    Lal S, Lacroix M, Tofilon P et al (2000) An implantable guide-screw system for brain tumor studies in small animals. J Neurosurg 92:326–333CrossRefPubMedGoogle Scholar
  32. 32.
    Donoghue JF, Bogler O, Johns TG (2011) A simple guide screw method for intracranial xenograft studies in mice. J Vis Exp 55, e3157. doi: 10.3791/3157 Google Scholar

Copyright information

© Springer Science+Business Media New York 2016

Authors and Affiliations

  • Srinivas S. Somanchi
    • 1
  • Bridget A. Kennis
    • 1
  • Vidya Gopalakrishnan
    • 1
  • Dean A. Lee
    • 1
  • James A. Bankson
    • 2
  1. 1.Division of PediatricsThe University of Texas MD Anderson Cancer CenterHoustonUSA
  2. 2.Department of Imaging Physics, Division of Diagnostic ImagingThe University of Texas MD Anderson Cancer CenterHoustonUSA

Personalised recommendations