Mouse cancer models are important for establishing basic mechanisms of tumor growth and immune escape. However, cancer treatment studies in mouse models have yielded finite translation to the clinic due to limited syngeneic models and species-specific differences^1,2. The emergence of immune therapies as a dominant approach to control tumors has reiterated the need for an in vivo model with a functional human immune system. Advancements in human immune system mice (HIS mice) over the past decade have made it possible to study immuno-oncology in vivo in a wide variety of cancer types and immunotherapeutic agents^3,4,5,6. Human tumor models, including cell-line derived and patient-derived xenografts (CDX and PDX, respectively), grow well in HIS mice and in most cases are nearly identical to their growth in the immunodeficient host lacking human hematopoietic engraftment^7,8. Based on this key finding, researchers have been using the HIS mouse model to study human immunotherapies, including combination therapies designed to alter the tumor microenvironment (TME) to decrease immunosuppression and thus enhance immune-directed tumor killing. These preclinical models help address the issues of heterogeneity of human cancers, and can also predict treatment success as well as monitor immune related drug toxicities^9,10.

The production of a mouse model with a human immune system through the introduction of human hematopoietic stem cells requires a recipient immunodeficient mouse that will not reject the xenograft. Current HIS mouse models are derived from immunodeficient mouse strains that were reported over 30 years ago. The first immunodeficient mouse strain described was SCID mice that lacked T and B cells¹¹, followed by a hybrid NOD-SCID with an SIRPα polymorphism responsible for mouse macrophage tolerance to human cells, due to increased binding for the NOD SIRPα allele to the human CD47 molecule^12,13. In the early 2000s, the deletion of the common gamma chain of the IL-2 receptor (IL-2Rγc) on both BALB/c and NOD immunodeficient strains was a game changer for enhanced human engraftment, due to genetic deletions forbidding host NK cell development^14,15,16,17. Alternative models, such as BRG and NRG mice, achieve T and B cell deficiency through deletion of the Rag1 or Rag2 gene, required for T and B cell receptor gene rearrangements and thus the maturation and survival of lymphocytes^18,19. The BRGS (BALB/c -Rag2^nullIl2RγC^nullSirpα^NOD) mouse used herein combines the IL-2Rγ chain deficiency and the NOD SIRPα allele on the Rag2-/- background, resulting in a highly immunodeficient mouse without T, B, or NK cells, yet with sufficient vigor and health to allow for long term engraftment of more than 30 weeks¹³.

HIS mice can be generated in multiple ways, with human PBMC injection being the most direct method^15,18,20. However, these mice have a pronounced expansion of activated human T cells that results in graft versus host disease (GVHD) by 12 weeks of age, preventing long-term studies. Alternatively, human hematopoietic stem cells from umbilical cord blood (CB), bone marrow, and fetal liver can also be used for engraftment and production of the human immune system de novo. In this system, the hematopoietic stem cells produce a multi-lineage human immune system with the generation of T, B, and innate immune cells that are importantly tolerant of the mouse host, compared to the PBMC mice that develop mostly T cells. Therefore, GVHD is absent or greatly delayed, and studies can be extended to mice up to 10 months of age. CB provides an easy, accessible, and noninvasive source of CD34+ human hematopoietic stem cells that facilitates the engraftment of multiple HIS mice with genetically identical immune systems^17,18,20,21. Over the past few years, HIS mouse models have been used extensively to study immunotherapy and the TME^3,4,5,6. Despite the development of human derived immune systems in these mice, human xenograft tumors grow at similar rates compared to the control immunodeficient mice and allow for the complex interplay between the cancer cells and immune cells, which is important for maintaining the microenvironment of the engrafted PDX^3,7,8. This protocol has been used to perform over 50 studies testing treatments in HIS-BRGS mice with PDXs and CDXs. An important conclusion is that human tumors in the HIS mice maintain their unique TME as defined by molecular evaluation of the tumor relative to the initial patient sample and immune infiltrate characteristics^3,22,23. Our group focuses on in-depth evaluation of the HIS in both immune organs and the tumor using multi-parameter flow cytometry. Herein, we describe a protocol for the humanization of BRGS mice, evaluation of chimerism, implantation of human tumors, tumor growth measurements, cancer treatment administration, and analysis of the HIS cells by flow cytometry.

Protocol

All animal work was performed under animal protocols approved by the University of Colorado Denver Institutional Animal Care and Use Committee (IACUC Protocols #00593 and #00021). All animal work was performed in accordance with the Office of Laboratory Animal Resources (OLAR), an accredited facility by the American Association for Laboratory Animal Care, at the University of Colorado Denver Anschutz Medical Campus. All human cord blood samples were obtained as donations from de-identified donors and are thus not subject to approval by the human research ethics committee.

NOTE: Compositions of all media and solutions mentioned in the protocol are included in Supplemental File 1. Figure 1 illustrates the overall protocol for generation and analysis of immune responses to tumors in HIS-BRGS mice.

1. Generation of HIS mice

1. Mouse husbandry of BALB/c -Rag2^null Il2RγC^null Sirpα^NOD (BRGS) mice
   NOTE: This strain is extremely immunodeficient, with no T, B, or NK cells. Therefore, rigorous measures must be used to prevent opportunistic infections. Maintain the colony on a diet containing trimethoprim and sulfadiazine on an alternating 2 week schedule with a normal diet. Maintain in the highest level of precaution housing room possible (e.g., a barrier shower-in facility with limited access).
   1. Maintain colonies of BALB/c -Rag2^null Il2RγC^null Sirpα^NOD (BRGS) and BALB/c Rag2^null Il2RγC^null Sirpa^Balb/c (BRG) homozygous mice as breeders.
   2. Breed BRGS^N/N×BRG^B/Bto generate BRGS^B/N pups, to be used as recipients of human stem cells. In this colony, BRGS^B/N are healthier than BRGS^N/N, and engraft at equivalent levels (more than BRG).
2. CD34+ human stem cell isolation from umbilical CB
   NOTE: No antibiotics are used for this procedure. Therefore, good sterile technique is imperative.
   1. Place a 50 mL conical tube rack, as well as 3x 15 mL and ~10x 50 mL conical tubes in a sterilized biosafety cabinet (BSC). Spray a blood collection bag with 70% ethanol and let it dry in the BSC.
   2. Calculate the number of 50 mL conical tubes required for CB density gradient isolation = CB volume/15, rounded up and to an even tube number. Calculate blood volume per tube = CB volume/number of tubes. Pour blood carefully from the CB bag into each conical tube; this is a maximum of 15 mL per tube. Use an automatic pipettor and a 25 mL serological pipette to mix the blood 1:1 with sterile PBS by pipetting up and down.
   3. Use an automatic pipettor on low speed and a 10 mL serological pipette to slowly underlay the blood with room temperature (RT) 1.077 g/mL density gradient solution (see Table of Materials) without disturbing the interface. Keep the pipette tip from touching the tube bottom. Repeat for all tubes. Then centrifuge for 30 min at 850 x g, with no braking, at RT to ensure maintenance of the density gradient.
   4. Visualize the cellular buffy coat on top of the 1.077 g/mL density gradient as a cloudy white layer. Remove and discard the plasma layer down to about 10 mL above the buffy coat using a 25 mL serological pipette and an automatic pipettor.
   5. Collect the buffy coat with a sterile transfer or serological pipette. Use the pipette like a spatula to scrape the cells off the side of the conical tube while releasing the bulb (or pipetting slowly) to draw the cells up. Combine the buffy coats from two 50 mL conical tubes into one new 50 mL conical tube.
   6. Wash the cells by pouring 45 mL of sterile HBSS containing 2% FBS into each conical tube. Centrifuge for 11 min at 360 x g at RT.
   7. Aspirate the wash media down to the pellet in all tubes. Use a 10 mL serological pipette and an automatic pipettor to resuspend the first pellet in 10 mL of HBSS containing 2% FBS. Resuspend each pellet in the same 10 mL of HBSS, and rinse each tube with an additional 10 mL of HBSS to collect all the cells into a single tube.
   8. Pour 45 mL of sterile HBSS containing 2% FBS in the conical tube. Centrifuge for 10 min at 360 x g, at 4 °C.
   9. Aspirate the wash buffer down to the cell pellet and resuspend the pellet in 20 mL of magnetic cell separator buffer (see Table of Materials). Remove a small aliquot to count the cells with a hemacytometer at a 1:20 dilution in methylene blue. Add the number of blue and white cells. Centrifuge at 360 x g, for 10 min at 4 °C.
      NOTE: This protocol uses magnetic bead technology (see Table of Materials). The protocol can be modified for use with any cell-separation technology, with sufficient purity and yield of the CD34+ stem cells.
   10. Aspirate the supernatant and resuspend the CD34+ cell pellet isolated from cord blood in 300 µL of magnetic cell separator buffer per 1 x 10⁸ cells. Add 100 µL of FcR blocking reagent first, and then 100 µL of CD34+ magnetic beads per 1 x 10⁸ cells. Incubate at 4 °C for 30 min (no ice).
   11. Add 5 mL of magnetic cell separator buffer per 1 x 10⁸ cells and spin at 360 x g for 10 min at 4 °C. Repeat the wash step and resuspend the pellet in 500 µL of magnetic cell separator buffer per 1 x 10⁸ cells in a 15 mL conical tube labeled "unfractionated".
   12. Label two more 15 mL conical tubes "CD34-" and "CD34+". Place the three 15 mL conical tubes (unfractionated, CD34-, and CD34+) to slots A1, B1, and C1, respectively, on a cooling rack (see Table of Materials). Separate the cells using the two-column positive selection program on an automatic magnetic cell separator (see Table of Materials) in a BSC, according to the manufacturer's instrument instructions.
3. Expanding and freezing CD34+ human stem cells
   1. Aliquot 10 µL of the recovered CD34+ cell suspension (2 mL) on a hemocytometer slide and count the cells under 10x magnification. Calculate the total number of CD34+ cells by multiplying the cell count by 2 x 10⁴. Divide the total CD34+ cell number by 250,000 to calculate the number of vials to freeze (50,000 cells per mouse pup prior to in vitro expansion).
   2. Prepare CB medium Iscove's 10% FCS (plus 1 mL extrafor filtering loss), supplemented with 40 ng/mL stem cell factor, 20 ng/mL Flt3L, and 10 ng/mL IL-6, and pass through a 0.22 µm filter. Resuspend the CD34+ cells at 100,000 per mL of CB medium, and incubate at 37 °C. On day 3, add an equivalent volume of CB medium without cytokines to the flask containing the cells and the CB medium with cytokines.
      NOTE: Addition of these cytokines to the CB medium promotes survival and expansion of the CD34+ cells while preventing differentiation.
   3. Harvest the expanded CD34+ cells on day 5. Pipette the cell suspension up and down and collect in a 50 mL conical tube. Add enough CB medium to cover the bottom of the flask. Using a cell scraper, scrape the entire bottom of the flask. Collect all the media into the same 50 mL tube, and centrifuge at 360 x g for 11 min.
   4. Resuspend the cells in 2 mL of CB medium. Save the final drop from the pipette into a 96-well plate for counting. Dilute the cells 1:1 in trypan blue and add 10 µL to the hemacytometer, then count and average cells from four quadrants. Calculate the total number of CD34+ cells by multiplying the cell count by 4 x 10⁴, and record the viability.
   5. Make n+1 mL of freezing medium, where n is the number of freezing vials calculated in step 1.3.1. Prepare freezing medium by adding 10% (v/v) DMSO to FBS and keep on ice. Label the cryovials with CB#, CD34+ d5, and date.
   6. Spin down the CD34+ cells at 360 x g for 10 min at 4 °C. Aspirate the medium down to the pellet and resuspend the cell pellet in freezing medium. Aliquot 1 mL of cell suspension to each vial and divide any remainder evenly between the vials. Add the vials to an isopropanol cell freezer chilled to 4 °C, place at -80 °C, and transfer to liquid nitrogen to store for >90 days.
4. Irradiation of mouse pups
   1. Collect the BRGS^B/N pups, 1-3 days after birth, into an autoclaved plastic box with padding. Add a small amount of bedding with the pups. Label the box with the cage number and the number of pups.
   2. Set up the irradiator (see Table of Materials) for a dose of 300 rad. Set the box of pups into the irradiator and expose them to 300 rad. Take the pups back to their cage, place them in a pile, and cover with bedding.
5. Pup injections and CD34+ cell preparation
   1. Begin CD34+ cell preparation ~3 h post irradiation. Warm 10 mL of CB media in a 50 mL conical tube. Perform all steps in a sterile BSC.
   2. Retrieve one vial of in vitro expanded and frozen CD34+ cells for every four to six pups to inject. Rapidly thaw at 55 °C, until just a small amount of ice is visible, and add the cells to the warmed CB medium (the vial should still be cool to the touch.) Use 1 mL of medium to rinse each vial and spin the cells at 360 x g for 12 min at 4 °C.
      NOTE: Rapid thawing at 55 °C was found to yield better cell viability (90%-95%) than thawing at 37 °C.
   3. Aspirate the medium carefully. Resuspend the (small) pellet in 2 mL of CB media, gently mix, and add ~30 µL of the cell suspension to a single well in a counting plate. Dilute 1:1 in trypan blue, then add 10 µL to a hemocytometer and count and average the cells from four quadrants.
   4. Calculate the total number of CD34+ cells by multiplying the cell count by 4 x 10⁴ and then record the viability. Spin at 360 x g for 12 min at 4 °C.
   5. Aspirate the medium carefully and resuspend the cell pellet in 100 µL of sterile PBS per n+1 pups to inject, resulting in 250,000-450,000 CD34+ cells per mouse. Place the conical tube on ice in a transport container and travel to the vivarium for pup injection.
   6. Bring a heat lamp, diapers, 1 mL syringe, 18 G needle, 30 G needle, and the CD34+ cell preparation in sterile containers to the vivarium BSC. Place a sterile diaper ~2 ft underneath the heat lamp. Retrieve the cage with the litter to be injected and place in the BSC.
   7. Assemble the syringe with an 18 G beveled needle. Matching the angle of the conical tube with the bevel of the needle, gently mix and draw up the cell suspension. Place the pups on the diaper to warm (watch for overheating). Remove air from the syringe and replace the 18 G needle with the 30 G needle, and then carefully push the syringe until the cell suspension is just at the needle tip.
      NOTE: Alternatively, an insulin syringe can be used.
   8. Take one pup at a time to the edge of the diaper away from the heat lamp. Immobilize the pup on its side under the thumb and forefinger, allowing for a clear view of the face. Notice the vein across the cheek under where the ear will be. Insert the needle shallowly into the vein (IV) nearest to the eye, and slowly inject 50 µL of cells.
   9. Check if a bubble is forming from a subcutaneous injection. If so, insert the needle deeper and proceed to inject cells. A small drop of blood/hematoma will be visible when successful.
      NOTE: Refer to the procedure by Gombash Lampe et al.²⁴ regarding pup injection.
   10. With the pup still immobilized, perform an intrahepatic injection (IH) with another 50 µL of cells (100 µL total per pup of IV+IH). The liver can be visualized as a dark spot between the white milk band and the thoracic cage. Place the injected pup on a diaper further away from the un-injected pups and heat.
       NOTE: IV injection results in better long-term chimerism than IH injection alone²⁵, but IV injections are not always successful. Therefore, splitting the injection between IV and IH ensures engraftment into a larger percentage of mice.
   11. Repeat both facial-vein and IH injections for all pups. Clean off any blood, return pups to the nest in their cage, and cover with bedding.

2. Testing human chimerism in blood

1. Test chimerism in the blood of HIS mice at both 10 and 14 weeks of age. Collect 50 µL of blood via the retro-orbital vein or using an alternative IACUC-approved method.
   1. For retro-orbital bleeds, anesthetize mice with an isoflurane vaporizer set to 5 for 1-2 min and then turn the vaporizer setting down to 4. Turn down the vaporizer as needed to allow mice to stay sufficiently oxygenated while under anesthesia. Do not keep the mice under isoflurane for more than 5 min.
   2. Place the nose of the anesthetized mouse into the isoflurane vaporizer nose cone attachment and place a drop of the analgesic (0.5% proparacaine HCl ophthalmic solution USP) on the eye.
   3. After 1 min, remove the proparacaine using a sterile gauze, proptose the eye, and insert a 75 mm heparinized hematocrit tube retro-orbitally to collect 50 µL of blood. Eject the blood into a 1.5 mL microfuge tube containing 50 µL of heparin and gently mix.
   4. Pinch the eye closed with sterile gauze to stop the bleeding and apply a drop of proparacaine. Remove the mouse from isoflurane and recover in a clean cage.
2. PBMC isolation from mouse blood
   1. Mix the blood/heparin by gently pipetting up and down and slowly overlay on top of 500 µL of 1.077 g/mL density gradient, being careful not to disturb the interface. Centrifuge the tubes at 1,220 x g for 20 min at RT with no brakes.
   2. Visualize the cellular buffy coat as a cloudy layer on top of the 1.077 g/mL density gradient, below the plasma. Remove as many of the cells as possible from the buffy coat with a 200 µL pipette and add to new 1.5 mL tubes containing 750 µL of harvest medium. Centrifuge at 360 x g for 11 min at RT.
   3. Aspirate the medium down to 50 µL and resuspend the pellet in 750 µL of harvest medium. Centrifuge at 360 x g for 10 min at 4 °C, and aspirate down again to 50 µL. The cells are ready to be resuspended in the surface stain.
3. Surface staining and flow cytometric analysis
   1. Fill out the staining panel worksheet (Table 1; "spectral flow bleed panel" worksheet) with mouse numbers and prepare the antibody staining cocktail by adding all fluorescent antibodies to the staining buffer (Supplemental File 1). Make a plate layout to add the samples to a 96-well U-bottom plate. Mark the bottoms of wells using a permanent marker. Titrate antibodies using a standard procedure prior to staining to determine the appropriate concentration²⁶.
   2. Add 62 µL of the surface stain cocktail to each well of the 96-well U-bottom plate. Resuspend the cells in the 50 µL remaining in the tube and add each sample to its corresponding well. Include a well for a positive staining control (human PBMCs + mouse splenocytes, 1 x 10⁶ cells each) to stain alongside the samples. Mix by pipetting and incubate the mixture for 15 min at 4 °C.
   3. Centrifuge at 680 x g for 3 min at 4 °C. Flick out the plate into the sink to remove the supernatant, and wash the cells by resuspending in 150 µL of staining buffer by gently pipetting ~10x. Spin and resuspend each well in 150 µL of staining buffer.
   4. Acquire the data for 100 µL of each sample on a flow cytometer (see Table of Materials) and export the .fcs files. Import the .fcs files to the flow data editing software (see Table of Materials). Apply a polygon gate to an FSC-A x SSC-A plot surrounding the cells, excluding any debris. Select the cells within the gate (see Table 1; "spectral flow bleed gating" worksheet).
   5. Change the axes to (FSC-A x FSC-H) and gate the cells contained on the linear diagonal excluding the doublets that protrude from the line. Select these cells and change the axes to (hCD45 x mCD45).
   6. Apply a polygon gate to the hCD45+ population and apply the name "human". Apply a polygon gate to the mCD45+ population and apply the name "mouse". Create a count statistic for both the human and mouse populations.
   7. Select the human population and change the axes to (CD19 x CD3). Apply a polygon gate to the CD19+ cells and name it "B cells". Apply a polygon gate to the CD3+ population and name it "T cells". Apply a polygon gate to the double negative population and name it "NonTB".
   8. Select the T cell population and change the axes to (CD8 x CD4). Apply a polygon gate to the CD4 and CD8 positive populations and name them "CD4+" and "CD8+", respectively.
   9. Select the NonTB population and change the axes to (CD56 x myeloid). Apply a polygon gate to the total CD56 positive population including the double positive events and name it "NK cells". Apply a polygon gate to the CD56 negative and myeloid positive population and name it "myeloid".
   10. Create a table for the percentage and count statistics of all populations and export to a spreadsheet software (see Table of Materials). Calculate %hCD45 chimerism = %hCD45/(%hCD45 + mCD45).
   11. Exclude HIS mice that are <20% hCD45+ (of mCD45 + hCD45) for further experiments.
       ​NOTE: In this study, PBMCs were prepared using a 1.077 g/mL density gradient separation for a cleaner RBC depletion. This procedure excluded human and mouse granulocytes from the PBMC layer. Alternatively, RBC lysis can be used.

3. Injection of tumors into mice

1. Prior to tumor injection, confirm T cell numbers from the 14 week bleed data. Tumors are injected to be ready for harvest between 20-26 weeks if the T cells are >20%, and between 24-28 weeks if the T cells are <20% of the total immune cell population. This injection timing ensures that the mice are 20-28 weeks of age and have sufficient numbers of T cells at the end of the study.
2. Injection of cell-line derived xenografts (CDX)
   NOTE: The procedure is described using MDA-MB-231 breast adenocarcinoma cells (see Table of Materials) as an example.
   1. Thaw an aliquot of frozen cells, wash 1x by resuspending in 10 mL of DMEM supplemented with 10% FBS, 1% PenStrep, and 1% nonessential amino acids, and centrifuge at 360 x g for 10 min at 4 °C. Aspirate the medium and resuspend in 10 mL of DMEM supplemented with 10% FBS, 1% PenStrep, and 1% nonessential amino acids in a T25 flask and incubate at 37 °C with 5% CO₂.
      NOTE: The cell lines are authenticated by PCR to ensure correct cell type. Cell supernatants are tested for mycoplasma via a biochemical assay before injection.
   2. Passage and expand the cells during exponential growth phase at about 80% confluency.
      1. Aspirate the medium, rinse the cells with 5-10 mL of sterile PBS (pH 7.2), and incubate with 1 mL of 0.25% trypsin-EDTA for 1-5 min until the cells detach from the flask.
      2. Add 5 mL of the same DMEM and mix by pipetting up and down. Passage the entire 6 mL of cell suspension into a new tissue culture treated T75 flask (1:3 dilution) and add an additional 10 mL of DMEM.
      3. Passage the cells every 2-3 days at 80% confluency at 1:3 dilutional plating in DMEM to expand the cells.
   3. Harvest the cells with 0.25% trypsin-EDTA during the exponential growth phase within six passages and wash with PBS. Mix PBS and basement membrane extract (see Table of Materials) in a 1:1 ratio and add to the cells at a final concentration of 5 x 10⁷ cells/mL.
   4. Anesthetize the mice with isoflurane as described in step 2.1.1. Subcutaneously inject 100 µL (5 x 10⁶ cells) of tumor cell suspension into each flank using a 23 G needle.
3. Perform injection of patient-derived xenografts (PDX) using trocars. The injection procedure and other development and maintenance instructions for the PDX model are available in literature²⁷.

4. Tumor growth measurement

1. Check tumor progress once a week after implantation by feeling along the flank for tumor growth. Once the tumors are palpable, anesthetize the mice with isoflurane (as described in step 2.1.1) and shave on each flank with an electric trimmer, taking care around the tumor to prevent ulcerations as the tumor grows.
   NOTE: Mice can be shaved before tumor injection to prevent injuries to the skin around the tumors; however, the rate of hair regrowth could obstruct tumor measurements.
2. Measure the length and width of the tumors twice per week using calipers and record the measurements in mm. Report tumor measurements as tumor volume (mm³) using the formula . Take care not to allow the tumor burden to exceed 2,000 mm³ per single tumor or a combined volume of 3,000 mm³.
   NOTE: Guidelines for the use of mice in cancer research vary by location. Please refer to Animal Care and Use policies at the researcher's institution.

5. Drug treatments

1. Allocate the HIS mice into equivalent treatment groups based on hCD45 chimerism, hCD3 chimerism, and hCD8 chimerism. Once the tumors reach 100 mm³ on average, begin drug treatments.
   NOTE: The number of mice per group is based on number of HIS mice generated from the CB. A minimum of four mice per group is recommended. Non-bipartite multigroup matching (e.g., in R-vivo Manila software) is useful for this grouping²⁸.
2. Drug route and frequency
   1. Inject anti-PD-1 inhibitors (nivolumab, pembrolizumab) intraperitonially (i.p.) at 30 mg/kg 1x per week or 20 mg/kg 2x per week for single treatments, or 10-15 mg/kg 2x per week for combination treatments.
   2. Dose the combination therapies, such as targeted therapies, chemotherapies, and irradiation, in accordance with the experimental design. Targeted therapies and chemotherapies can be given through oral gavage, i.p. injection, or food.
      ​NOTE: Doses can be varied and optimized according to published data and the tumor model. Dose studies are often tested in immunodeficient recipients prior to HIS mouse studies.
3. Monitor the mice at least 3x per week for health changes such as weight loss, loose feces, hunched posture, reduced mobility, and fur loss. Some symptoms could be signs of drug toxicities or GVHD, and drug dosing may need reduced or stopped. Euthanize mice as needed according to Animal Care and Use policies.

6. Harvesting of mouse tissues and tumors at the end of the study

1. Euthanize the mice singly according to institutional and veterinary guidelines using compressed CO₂ gas with a flow rate of 2.75 L/min. Monitor the mice held in CO₂ for 1 min past death and then perform cervical dislocation as a secondary form of euthanasia.
2. Blood collection
   1. Collect blood via intracardiac puncture. Hold the mouse in an upright position and insert a 1 mL syringe with a 25 G needle directly into the heart from a line just left of the midline and below the ribs. Collect the blood into a syringe ideally through one long draw and transfer into a labeled 1.5 mL tube.
      NOTE: Additional needle insertions can be performed to collect additional blood from the corners of the lung cavity with care.
   2. Leave the blood at 4 °C for 1 h, then centrifuge the blood in a microcentrifuge for 6 min at 7,800 x g. Collect the clear liquid (sera) above the blood interface into a second clean and labeled 1.5 mL tube. Store the serum at -20 °C for downstream analysis, such as human and mouse inflammatory cytokine or immunoglobulin titers.
3. Tissue dissection
   1. After the injection and growth of tumor cells in the mice and the administering of drug treatments, harvest the tissues. Place the mice on a foam dissection board, with pins to hold them in place, and the arms and legs extended at 45° angles. Make an incision up the middle of the torso, starting near the pelvis and extending to the chin, trying to avoid cutting the peritoneum (although this is not essential). Pull the skin to the edge and hold in place with pins.
   2. Extract the lymph nodes (LNs) using fine forceps in the following order: inguinal, axillary, cervical, mesenteric, hiatal.
      NOTE: In HIS mice, the LNs are often very small, resembling anlage, or "inflamed" with a distinct appearance unlike that from a wild-type mouse. Therefore, tissue resembling LNs is taken from each site. The peripheral LNs often appear as fluid-filled "balls", while the mesenteric LNs are more dense. The mesenteric LNs are the most obvious and consistent and are observed as one or two distinct denser nodes, as opposed to a string.
   3. Place the LNs on one side of frosted glass slides in 8 mL of harvest medium in a Petri dish. Holding the slides at perpendicular angles with the frosted edges inward, gently press the tissues until the cellular contents are released.
   4. Rinse the slides several times by pulling them apart and together to release the maximal amount of cells. Collect the cells with a 5 in glass pipette and filter them through a 9 in cotton-plugged pipette into a labeled 15 mL conical tube.
   5. Extract the spleen from the upper left side of the abdomen using either two pairs of forceps or forceps and scissors. Note the size of the spleen as an estimate of the volume of resuspension media. Collect and filter the spleen by mechanical digestion using frosted glass slides, as with the LNs.
      NOTE: Any technique for the preparation of tissues for single cell suspensions can be used.
   6. Perform counting and resuspension of the cells from tissue samples as described below.
      1. Centrifuge the lymph cells at 360 x g for 10 min at 4 °C. Aspirate the liquid and resuspend the cell pellet in 1 mL of harvest medium with DNase.
      2. Remove erythrocytes from the spleen cells by incubating with 3 mL of ACK lysis buffer at RT for 3 min, followed by the addition of 10 mL of harvest medium/DNase. Centrifuge the spleen cells again, aspirate the supernatant, and resuspend the cells in 1-10 mL of harvest media/DNase, based on the spleen size (e.g., 1 mL for very small spleens and up to 10 mL for the largest spleens).
      3. Add a 10 µL aliquot of cell suspension to 90 µL of media and count on a hemocytometer. Centrifuge the LNs and spleen cells, aspirate the supernatants, and resuspend the cells in harvest medium at a concentration of 1 x 10⁸ cells/mL, or a minimum of 80 µL.
   7. Extract the tumors.
      1. Remove the tumor from the open flank by holding the tumor with forceps while slowly snipping at the tumor margins with dissection scissors.
      2. Once the tumor is removed, weigh it and remove 1/4 for RNA and immunohistochemistry (IHC) processing. Divide the 1/4 tumor in half; place one half (1/8 of the whole tumor) in a cryovial, flash freeze in liquid N₂, and store at -80 °C for downstream genomic studies.
      3. Place the other 1/8 of the tumor in a labeled specimen tube containing 10% formalin. The next day, rinse and resuspend the tissue in 70% ethanol until future use. Place the remaining 3/4 of the tumor into a 6 cm dish and mince into ~1 mm pieces using a scalpel blade.
      4. Transport the tumor pieces into a dissociation tube (see Table of Materials), rinse the dish with 5 mL of incomplete tumor infiltrating leukocyte (TIL) medium, and add to the dissociation tube.
         NOTE: For tumors weighing >0.4 g, rinse the dish with 10 mL of TIL incomplete media and add to dissociation tube.
      5. Add collagenase preparation (see Table of Materials) at a final concentration of 50 mg/mL to the tissue in the dissociation tube. Dissociate the tissue using mechanical dissociation at 37 °C for 30 min to 1 h, depending on the firmness of the tumor.
      6. After dissociation, pass the suspension over a 100 µm filter into a 50 mL conical tube and rinse the filter with 10 mL of TIL complete media. The serum in this media will stop the collagenase reaction and protect the cells from further degradation.
      7. Centrifuge the single cell suspension at 360 x g for 10 min at 4 °C. Resuspend the pellet in just enough harvest medium with DNase so that the cell suspension can easily pass through a P1000 pipette tip and record the volume for downstream analysis.
         ​NOTE: A smaller volume increases the collection of tumor immune cells but is more susceptible to flow cytometer clogs.

7. Cell staining and flow cytometric analyses

1. Stain preparation and cell plating
   1. Prepare staining cocktails: On the day of harvest, add the number of samples to the staining worksheet (Table 2; "Conventional Flow A panel", "Spectral Flow B panel", and "Conventional Flow C panel" worksheets) for all stains and print. Prepare surface stain cocktails for stains A and B in staining buffer (SB) by adding each antibody individually with a new tip, marking off each reagent on the go, and store at 4 °C until needed. Prepare the appropriate viability dyes in azide-free PBS and store at 4 °C until needed (warm to RT before use). Prepare surface stain C cocktail on day 2 in SB along with intracellular stains B and C in their respective permeabilization buffers.
   2. Create a plate layout for all samples in the staining worksheet and aliquot 100 µL of azide-free PBS to the wells. Include unstained controls for each tissue and stain (for spectral cytometry).
   3. Add the cells to 96-well plates containing 50 µL of PBS. For stain A, resuspend each tissue group in the appropriate volume, as noted on the staining worksheet, and store at 4 °C until acquisition on flow cytometer on same day. For stains B and C, add 25 µL of lymph and 60 µL of non-lymph tissue cell suspensions to the wells with PBS.
   4. Perform in vitro stimulation of cytokines for detection by intracellular staining using stain C.
      1. Centrifuge the stain C 96-well plate from step 7.1.3 at 680 x g for 3 min at 4 °C. Add 200 µL of TIL complete media (RPMI 1640, 10 mM HEPES [pH 7], 10% FBS) to each well. Store the plate with the cell suspension at 4 °C overnight.
      2. Early the following morning, dilute the cell stimulation cocktail (see Table of Materials) 1:500 in complete TIL media. Centrifuge the plate containing the cell suspension at 680 x g to pellet the cells and flick to remove the media. Resuspend the cells in 200 µL of the prepared cell stimulation cocktail and incubate at 37 °C for 1 h.
      3. Add 25 µL of protein transport inhibitor solution containing monensin at a 1:1,000 dilution in complete TIL media, mix the cells, and incubate at 37 °C for an additional 4 h to allow intracellular accumulation of cytokines.
   5. Perform cell staining.
      1. For stains A and B (Day 1), and for stain C (Day 2), centrifuge at 680 x g for 3 min at 4 °C. Flick the plates and add the appropriate viability dyes to the wells. Mix well by gently pipetting up and down with a multichannel pipette and incubate for 15 min at RT.
      2. Centrifuge the plates (as in step 7.1.5.1), then add appropriate surface stains to the plates and mix gently by pipetting with a multi-channel pipette. Incubate for 15 min at 4 °C and centrifuge again. Flick the plates, wash the cells by gently pipetting up and down with 150 µL of SB, and centrifuge.
      3. Flick the plates, repeat the wash by gently pipetting up and down with 150 µL of SB, and centrifuge at 680 x g for 3 min at 4 °C. For stain A, resuspend each tissue group in the appropriate volume, as noted on the staining worksheet, and store at 4 °C. For stain B, fix with FoxP3 transcription factor kit fixative (see Table of Materials) for 30 min at RT. For stain C, fix in 1% (v/v) paraformaldehyde in SB for 30 min at RT.
      4. Centrifuge the fixed plates at 480 x g for 3 min at 4 °C. Wash 1x in SB. For Stain B, the cells can be left overnight.
      5. Permeabilize the cells: resuspend the wells in 150 µL of FoxP3 transcription factor kit permeabilization buffer for stain B and in 0.5% (w/v) saponin in SB for Stain C. Incubate for 15 min at RT. Centrifuge the fixed plates at 480 x g for 3 min at 4 °C.
      6. Flick the plates and add the respective intracellular staining cocktails for stain B and stain C. Incubate for 30 min at RT.
      7. Centrifuge the fixed plates at 480 x g for 3 min at 4 °C and flick the plates. Wash the cells with 150 µL of the corresponding permeabilization buffers (stain B, FoxP3 transcription factor kit permeabilization buffer; stain C, saponin).
      8. Flick the plates and wash by pipetting up and down in 150 µL of SB. Centrifuge at 480 x g for 3 min at 4 °C. Flick the plates and resuspend the cells in SB by tissue group in the appropriate volumes noted on the staining worksheet.
         NOTE: The samples are now ready for acquisition by spectral flow cytometry.
      9. Set up the flow cytometer (see Table of Materials) with appropriate single stained controls for each stain and unstained samples for each tissue group to account for differences in autofluorescence during unmixing. Acquire the appropriate volumes per tissue group as defined in the staining worksheet (Table 2) and export the .fcs files.
2. Flow cytometry data analysis
   1. Using the flow cytometry analysis software (see Table of Materials), create a new workspace. Create new groups for each organ (LN, spleen, and TIL). Import the .fcs files for each organ into the group.
   2. Create a bivariate dot plot, set the axes to (FSC-A x SSC-A), and apply a polygon gate to cellular events, avoiding events on the edges (all cells gate). Select the cellular events, change the axes to (FSC-A x FSC-H), and apply a polygon gate to the events on the linear diagonal, excluding events that deviate from the diagonal to generate the "singlets" gate. Select the "singlets" gate and change the axes to (hCD45 x mCD45). Apply polygon gates to the hCD45 and mCD45 positive events, and name them "human" and "mouse", respectively.
   3. Select the human population and change the Y-axis to Live/Dead Aqua. Apply a polygon gate to the Live/Dead negative, hCD45 positive population and name it "live human". Select the mouse population and change the X-axis to Live/Dead Aqua. Apply a polygon gate to the Live/Dead negative, mCD45 positive population and name it "live mouse".
   4. In a similar fashion, select the parent gate and the X- and Y-axes as outlined in the "Conventional Flow A Gating", "Spectral Flow B Gating", and "Conventional Flow C Gating" worksheets (Table 2) to isolate the indicated populations (e.g., human B cells, activated T cells).
      NOTE: Representative gating for each stain (Bleed, A, B, C) is included in Supplemental File 2.
   5. Create count and frequency export tables in the flow cytometry analysis software for all populations and export to a spreadsheet software. The parent population for frequencies is indicated in Table 2.
   6. Use the data to generate graphs based on experimental treatment groups.
      NOTE: The data can also be analyzed using R packages in the analysis software. A single .fcs file from all samples to be analyzed can be created using the concatenate feature. This data can be dimensionally reduced with the T-SNE algorithm while adding keyword parameters for tissue type and treatment group. The FlowSOM algorithm can then be used to cluster populations and the ClusterExplorer tool can be used to identify the populations. Novel cell populations can be identified in this manner, compared visually, and quantified between treatment groups or within various tissues.
   7. Correlate immune parameters for tumors within the same treatment group, with the growth for that tumor to define immunotypes that correlate with tumor growth inhibition. Quantify tumor growth by the specific growth rate (SGR) for that tumor, a measurement that takes into account the difference in tumor volumes over a specified time. This measurement normalizes tumors harvested on different days due to mouse health and treatment start dates.
      

Following the flank tumor protocol and experimental timeline (Figure 1), the tumor growth and immune response to a targeted tyrosine kinase inhibitor (TKI) therapy and nivolumab combination treatment was studied in two distinct human colorectal cancer (CRC) PDXs. The TKI drugs have been studied in immunodeficient hosts to evaluate tumor growth only²⁹. This model enabled the study of changes in the immune response of the TKI alone, and more importantly, in combination with anti-PD-1. This study was focused on the combination treated cohorts in two distinct experiments that represent a successful evaluation using HIS-BRGS mice and a technically flawed experiment. For PDX CRC307P, the combination treatment slowed the growth of the tumor, as determined by tumor growth volumes over time, tumor weights at harvest, and the SGR (Figure 2A-C). On the other hand, the growth of a PDX developed from a metastatic tumor from the same patient, PDX CRC307M, tested in a different cohort of HIS-BRGS mice, was less affected by the same combination treatment in HIS-BRGS mice (Figure 2D).

Despite the allocation of HIS-BRGS mice into equivalent experimental groups based on overall human (hCD45+) and human T cell (hCD3+) chimerism in the blood prior to tumor implantation (Figure 3A-C), both parameters increased in the peripheral immune system (spleens) and the tumor-infiltrating leukocytes (TIL) in combination treated CRC307P-bearing mice, but not the CRC307M model (Figure 3D-F). Notably, although both HIS-BRGS cohorts had comparable human and T cell chimerism in the blood prior to tumor implantation, the CRC307M cohort had very little lymph node chimerism and failed to develop appreciable levels of T cells in the spleen in the treated cohort (Figure 3D-F). The CRC307M experiment represents a technically flawed example due to overall insufficient splenic T cell chimerism and lymph node chimerism. We suggest the exclusion of HIS mice that have <20% hCD3+ chimerism in the spleens and/or <1.5 x 10⁶ hCD45+ cells in the lymph nodes at the end of the study. For the CRC307M model, the majority of mice were excluded, mostly due to very small lymph nodes, leaving only two mice per cohort.

Further investigation of the human T cells (Figure 4A) revealed more activated (HLA-DR+) T cells in the CR307P tumors, but not the lymph, from combination-treated mice (Figure 4B). In the CRC307M "negative" experiment, there was no significant increase of HLA-DR+ T cells in the TILs of treated HIS-BRGS mice when analyzing all the mice, suggesting this non-optimal HIS-BRGS cohort influenced the data significance due to HIS mice with very few T cells and no activation (Figure 4B). Indeed, the increase in HLA-DR+ T cells in the TILs in the CRC307M model did reach significance when excluding the low T cell chimerism HIS-BRGS mice (Figure 4B). In addition, there were more effector memory CD8+ T cells and fewer TIM-3+ (terminally exhausted) T cells in the combination-treated CRC307P tumors, whereas this difference was not noted in the CRC307M model (Figure 4C,D). In this experiment, no changes in the frequencies of cytotoxic T cells (Granzyme B+ or IFNγ+TNFα+) populations were observed among the combination-treated mice, although higher cytotoxic T cells were observed in untreated (Figure 4E) or treated (data not shown) tumors relative to lymph nodes. Importantly, although there were no differences in frequencies among the T cell populations, higher numbers of cytotoxic T cells were observed in the tumors of treated HIS-CRC307P-BRGS mice, with higher frequencies (Figure 3B) and numbers of human T cells in the tumors. On the other hand, this combination treatment showed no effect on the frequencies of Tregs in either CRC307P lymph organs or tumors, although the CRC307M data showed a trend of reduced Tregs that would need to be validated in another experiment (Figure 4F).

In addition to interrogating the human immune system, immune-related changes on tumor cells were also evaluated using flow cytometry (Figure 5A). By gating on EpCAM+ cells, as described in the protocol, increased expression of both MHC Class I (HLA-ABC) and Class II (HLA-DR) was found on the CRC307P tumor cells excised from combination-treated HIS-BRGS mice (Figure 5B). In the CRC307M model, this same drug treatment induced HLA Class II expression on the tumor cells, although to a lesser degree than in the CRC307P model (Figure 5C). Thus, the combination treatment appears to induce upregulation of MHC class II, independent of T cell infiltration (Figure 5B,C). Notably, MHC expression on the tumor cells was lower than that on human immune cells, a finding consistent with human tumor reports (Figure 5B). Similarly, the combination-treatment resulted in increased PD-L1 expression on the EpCAM+ CRC307P tumor cells (Figure 5D).

Finally, correlations of immune responses with tumor growth were investigated by plotting the SGR of the tumor versus immune parameters. Although the frequency of CD4+ T cells in the tumors of untreated mice showed no correlation with tumor growth, increased CD4+ T cells showed a significant (Figure 6A) correlation with smaller tumor growth, and more specifically the HLA-DR+ activated T cells (Figure 6B), in the combination treated HIS mice.

Figure 1: Schematic illustrating the generation of HIS-BRGS mice and implantation of human CDX or PDX tumors for cancer immunotherapy studies. The timeline is important to ensure the presence of T cells that take months to engraft in the CB-derived HIS mice. Created with Biorender.com. Please click here to view a larger version of this figure.

Figure 2: Analysis of tumor growth. Tumor growth measurements for CRC307P PDX in control (black) or combination treated (red) HIS-BRGS mice quantified by (A) volume over time, (B) weights at the end of the study, or (C) specific growth rate (SGR). (D) SGRs for CRC307M PDX. Data include tumors implanted into both flanks of six vehicle HIS-BRGS mice, seven combination-treated BRGS mice for CRC307P, and only two vehicle and combination-treated mice for CRC307M due to high exclusion rates. Statistical analyses between two independent groups were performed using unpaired, parametric two-group Welch's t-test. **p < 0.01, ****p < 0.0001. Please click here to view a larger version of this figure.

Figure 3: Human immune and T cell chimerism in lymph organs and tumors of CRC PDX tumor-bearing HIS-BRGS mice. (A) Representative flow cytometry analysis of human (hCD45) and mouse (mCD45) hematopoietic and human T (CD3) cells in peripheral blood (PBMC) prior to tumor implantation, and in LNs and tumors (TIL) at the end of the study. (B,C) Equivalent human and T cell chimerism in blood at 14 weeks in HIS-BRGS mice subsequently injected with (B) CRC307P or (C) CRC307M PDXs and untreated or treated with combination immunotherapy (Tx). (D-F) Increased human T cells in lymph organs and tumors (TIL) of combination-treated HIS-BRGS-CRC307P mice but not HIS-BRGS-CRC307M mice. Data show human and T cell frequencies on the Y-axis as a percentage of the parent population in (D) lymph nodes (LNs), (E) spleens (SP), and (F) tumors of individual mice at the end of the study. Statistical analyses between two independent groups were performed using unpaired, parametric two-group Welch's t-test. *p < 0.05, **p < 0.01, ***p < 0.001. Please click here to view a larger version of this figure.

Figure 4: Analysis of T cell activation by flow cytometry in peripheral lymph organs and tumors of CRC PDX bearing HIS-BRGS mice. (A) Representative flow cytometry analysis of T cells in LNs, spleens (SPs), and tumors (TILs) excised from HIS-BRGS mice measuring the following populations: activated T cells (HLA-DR+), naïve T cells (CD45RA+ CCR7+), Tem (CD45RA-, CCR7-), Tregs (CD25+, FoxP3+), and cytotoxic T cells (Granzyme B+, TNFα+, and/or IFNγ+). (B-D) Frequencies of indicated T cell populations in lymph organs or TILs of HIS-CRC307P-BRGS and HIS-CRC307M-BRGS mice: (B) HLA-DR+ activated T-cells, (C) CCR7-CD45RA-Tem CD8+ T cells, (D) inhibitory receptor TIM-3 CD8+ T cells, (E) Granzyme B CD8+ T cells, and (F) CD25+FoxP3+ CD4+ Tregs. For the HIS-CRC307M-BRGS mice in (B), the first data set shows all mice analyzed at harvest, whereas the second data set includes only those with sufficient LN and splenic T cell chimerism. In all graphs, the filled symbols for 307M represent mice with sufficient chimerism, whereas open symbols are excluded HIS mice. Statistical analyses between two independent groups were performed using unpaired, parametric two-group Welch's t-test. *p < 0.05, **p < 0.01, ***p < 0.001. Please click here to view a larger version of this figure.

Figure 5: Flow cytometry analysis of MHC class I, II, and PD-L1 on human tumors in HIS-BRGS mice. (A) Representative flow cytometry dot plots, illustrating gating strategy for measuring expression levels of MHC Class I (HLA-ABC), Class II (HLA-DR), and PD-L1 on epithelial (EpCAM+) human tumors in HIS-BRGS mice. (B-D) Mean fluorescence intensity (MFI) of HLA-ABC (B), HLA-DR (B,C), and PD-L1 (D) on human (hCD45+) and tumor (EpCAM+) cells from dissociated CRC307P (B,D) or CRC307M (C) PDXs excised from HIS-BRGS mice. Statistical analyses between two independent groups were performed using unpaired, parametric two-group Welch's t-test. *p <0.05, **p<0.01, ****p<0.0001. Abbreviation: Tx = Treatment Please click here to view a larger version of this figure.

Figure 6: Analysis of immunocorrelates of response. The SGRs of CRC307P tumor growth in the flanks of untreated (Veh) or combination immunotherapy-treated (Tx) HIS-BRGS mice were plotted versus the frequency of (A) CD4+ or (B) HLA-DR+ T cells as an indicator of immune parameters that correlate with reduced tumor growth (i.e., smaller SGR). A simple linear regression analysis was performed to indicate significant correlations. R² values indicate degree of correlation. *p < 0.05. Please click here to view a larger version of this figure.

Table 1: Surface staining panel worksheet. Please click here to download this Table.

Table 2: Cell staining panels and flow cytometry gating worksheet. Please click here to download this Table.

Supplemental File 1: Recipes for media and solutions used in the study. Please click here to download this File.

Supplemental File 2: Representative flow cytometry gating for each stain. Please click here to download this File.

Over the past 6 years, using our expertise in both immunology and humanized mice, our research team has developed a much needed preclinical model to test immunotherapies on a variety of human tumors^3,7,30,31. This protocol emphasizes the consideration of the variability of the model, with special attention to the immunotherapy-centric human T cell populations. In this protocol, the generation of HIS mice is described, as well as the immune analysis of both their lymph organs and the tumor. Flow cytometry protocols developed for traditional compensation-based cytometers and spectral cytometers have also been included. The ability to interrogate changes in the immune system in large tumor samples, as well as in the periphery, represents a significant advantage to this preclinical model. An additional advantage to this system is that multiple mice, harboring the same unique human tumor, can be allocated into separate cohorts and treated with different combinations of immunotherapy drugs, essentially performing a "mini" drug trial. Combined, these two factors enable investigation into the mechanisms of action and resistance of proposed (combination) immunotherapies. The model has the ability to test drug timing and dosing on a wide variety of human tumors and a multitude of immune-targeted treatments, including targeted therapies, biologics, chemotherapies, radiation, and even cell-based therapies.

Due to the inherent variability in the HIS mice chimerism, a higher number of mice per treatment group is required than a typical syngeneic tumor model. Using these protocols, cohorts of five to seven HIS-BRGS mice per treatment arm are sufficient to obtain statistical differences in immune parameters. This variability in chimerism represents a challenge in this model and must be considered. An experimental example demonstrating both tumor growth and immune response upon a combination treatment to a CRC PDX has been shown here, as well as an example in which the same treatment to a different CRC PDX did not show a similar strong response. This difference in experimental outcome could be due to the different PDX (same patient but metastatic vs. primary) and/or the chimerism. In the second experiment, there were very few T cells in the spleen of the majority of mice in the treatment group, as well as very small LNs; several labs have shown that the largest batch-to-batch variability among HIS-models are the T cell frequencies^20,32,33. We have also shown that human engraftment in the LNs correlates highly with the T cell chimerism³². Across more than 40 experiments, we have noted a requirement of 20% T cells in the spleen for adequate chimerism to detect treatment-related immune changes³. Upon sacrifice, it is important to ensure significant T cell reconstitution, which we define as notable LN development (>1.5 x 10⁶ hCD45 cells) and >20% T cells (of hCD45+ cells) in the spleen. HIS mice that do not meet this requirement are removed from the data set. Chimerism in the blood prior to tumor implant helps predict this parameter; however, there are differences in T cell development in individual mice over time that are best considered at the end of the study. Fortunately, in this model, approximately 95% of HIS-BRGS-mice cohorts develop sufficient T cell chimerism for immunotherapy studies. It should be emphasized that the data analysis at the end of the study should always include interrogation of the chimerism in the peripheral lymph organs and remove HIS mice that do not meet the human and T cell chimerism. In experiments with less than four HIS mice per cohort following this adjustment, the results should be validated with a different HIS-BRGS cohort derived from a distinct CB.

Although our group has focused on the BRGS mouse model recipient, there are numerous models available worldwide. For instance, the NSG (NOD.Cg-Prkdc^scidIl2rg^tm1Wjl/SzJ)recipient is the most widely characterized model and is similar to the NOG (NOD.Cg-Prkdc^scidIl2rg^tm1Sug/Jic) and even the NRG (NOD-Rag1^tm1MomIl2rg^tm1Wjl/SzJ) model, which uses Rag mutations for genetic ablation of mouse T and B cells instead of the SCID mutation²⁰. These models are all on the NOD background strain and thus have the SIRPa^NOD allele, which is important in avoiding phagocytosis of human cells by host macrophages¹³. Reports suggest similar chimerism in these "basic" humanized mouse models as in the BRGS model, with mostly human B cells in the first few months followed by engraftment with human T cells^32,34,35. In these models, the NK and myeloid cells are underrepresented relative to a human^36,37. There are several iterations of "next gen" humanized mouse recipients that can improve lineage-specific development through the introduction of human-specific cytokines. For example, NSG-SGM3 (NOD.Cg-Prkdc^scidIl2rg^tm1Wjl Tg(CMV-IL3,CSF2,KITLG)1Eav/MloySzJ) mice have human IL-3, GM-CSF, and SCF cytokines and improved human myelopoiesis³⁷, or human HLA transgenes for selection on a human HLA (e.g., NSG-HLA-A2 or BRGS-HLA-A2/DR2^38,39). Extensive reviews on this topic^20,21 are recommended to readers. HIS mouse models should be validated in each laboratory with their own source of human stem cells. This validation should include quantifying the human chimerism, including human B, T, NK, and myeloid cells, in the blood over time and in the LNs, spleen, bone marrow, and thymus between 16-20 weeks and again between 24-28 weeks. Special attention should be given to multi-lineage engraftment, including T cell chimerism. Additionally, the growth rates of tumors should be tested in the HIS mice relative to non-humanized mice of the same strain. It is plausible that significant improvements in HIS mouse models could result in rejection of the tumor. As a resource-savings attempt, these growth curves are often done simultaneously with the immune response data analyses. Commercially available HIS mice should include extensive chimerism data from the blood as well as references to chimerism in other organs from other cohorts.

The experimental protocols for analyzing tumor growth and immune responses in HIS mice to human CDXs or PDXs require standard laboratory methodologies and immunological expertise. The generation of HIS mice, including the isolation of the stem cells, injection into immunodeficient mice, and testing of human chimerism in the blood, are all relatively straightforward lab procedures. The logistical issues surrounding this procedure, including breeding highly immunodeficient mouse strains, acquiring a reliable source of human HSCs, etc., are the most cumbersome component of the protocol. Procedurally, more attention must be given to experimental design, including the timing of tumor implantation (typically 19-21 weeks when enough T cells are present), drug dosing, and end-of-study timing. Acquiring tumor measurement data is another routine procedure. However, since the immune data is often more informative than the tumor growth data in this system, flow cytometric analyses of the immune responses can provide pertinent data. The staining of cells for flow cytometry is another straightforward procedure, as is learning how to run a flow cytometer. However, the analysis and interpretation of this data requires unique expertise. While this manuscript serves as a backbone to carry out immuno-oncology studies in HIS mice, each protocol step should be optimized within a new lab. From experience, maintaining the health of the mice, acquiring adequate HSCs from CB units, and timing the tumor injections for sufficient T cells require the most attention and troubleshooting.

The HIS-BRGS mice exhibit extreme immune suppression similar to that observed in many human tumors. This immune suppression is notable, as high expression levels of PD-1 and TIGIT, yet less TIM-3, are observed on the T cells in the peripheral immune organs of the HIS-BRGS mice^3,7. These T cells also show markers of immune activation, including high expression of HLA-DR and low expression of CCR7 and CD45RA, indicative of T effector cells, as shown here in Figure 4 and in references^3,7. This quality is a central paradox to the system; the immunosuppression allows the growth of allogeneic human tumors in the presence of an allogeneic HIS, at rates similar to or even faster than non-humanized BRGS or nude (Nu/J) mice^3,7,23. However, immune stimuli have been shown to be able to overcome this immunosuppression and reject a tumor^7,8. Although significant changes in tumor growth between vehicle and treated mice are not always observed, significant differences in immune phenotypes have been found more often, so analyzing the tumor growth rates correlated to immune phenotypes is suggested. In this analysis, immune parameters were found that significantly correlate with reduced tumor growth³. These data suggest that particular "immunocorrelates" are treatment-related and participate in tumor control. Thus, tumor growth measurements combined with immune phenotypes in the HIS mouse model provide important preclinical data with regard to tumor immune responses to unique human tumors. The challenge for any preclinical model is correlation to clinical results. The HIS mouse oncology field has several examples of clinical correlation: 1) we and others have shown that the human immune infiltration correlates with the tumor type^3,22,23,40; 2) we have shown successful treatment with anti-PD-1 in an ACC Lynch syndrome PDX taken from a patient who subsequently responded to this therapy³¹; and 3) we have shown superior responses to a CRC MSS hi PDX, a cancer with high response rates in the clinic over the notoriously poor-responding CRC MSS PDX⁷. Together with in vitro and other in vivo models, these data provide a nexus for translation into the clinic.