Development of Humanized Mouse Models for Studying Human NK Cells in Health and Disease

Abstract

Humanized mice, which we define as immunodeficient mice that have been reconstituted with a human immune system, represent promising preclinical models for translational research and precision medicine as they allow modeling and therapy of human diseases in vivo. The first generation of humanized mice showed insufficient development, diversity and function of human immune cells, in particular human natural killer (NK) cells and type 1 innate lymphoid cells (ILC1). This limited the applicability of humanized mice for studying ILC1 and NK cells in the context of human cancers and immunotherapeutic manipulation. However, since 2014, several next-generation humanized mouse models have been developed that express human IL-15 either as a transgene or knock-in (NOG-IL15, NSG-IL15, NSG-IL7-IL15, SRG-15) or show improved development of human myeloid cells, which express human IL-15 and thereby promote human NK cell development (NSG-SGM3, MISTRG, BRGSF). Here we compare the various next generation humanized mouse models and describe the methodological procedures for creating mice with a functioning human immune system and how they can be used to study and manipulate human NK cells in health and disease.

1. Introduction

Natural killer (NK) cells are cytotoxic lymphocytes that are able to kill virus-infected cells and tumor cells by lysis or by inducing apoptosis. Together with type 1 innate lymphoid cells (ILC1), NK cells belong to group 1 ILCs¹. Although long been considered as a homogeneous population of innate lymphocytes, recent studies revealed subpopulations among NK cells that exhibit diverse phenotypic and functional characteristics, in part influenced by factors such as tissue localization and viral infections². Recently, NK cells and ILC1 have also received much attention regarding the role they play in anti-tumor immunity and cancer immunotherapy^3,4.

The cytokine interleukin 15 (IL-15) has been shown to be necessary for the proper development, maturation and function of NK cells, tissue-resident NK cells and ILC1, CD8αα intraepithelial lymphocytes and memory CD8⁺ T cells⁵. IL-15 has also been shown to be essential for functional anti-tumor responses of NK and T cells in cancer immunotherapy. Yet, there is poor interspecies cross-reactivity of IL-15, with only 65% of amino acids being identical between humans and mice⁶.

Since 2014, several next-generation humanized mouse models have been generated, which showed improved development and/or function of human NK cells (Table 1). These humanized mice expressed human IL-15 either as a transgene or knock-in (NOG-IL15, NSG-IL15, NSG-IL7-IL15, SRG-15) or showed improved development of human myeloid cells, which expressed human IL-15 and thereby promoted human NK cell development (NSG-SGM3, MISTRG, BRGSF). These next-generation humanized mice can therefore be used as preclinical in vivo models to study human NK cells in health and disease. More details about these humanized mice, in particular their applicability for studying human ILC1 and NK cell subsets (circulating and/or tissue-resident) and their function (tumor cell killing, ADCC) in vivo, is listed in Table 1.

Table 1:

Humanized mouse models to study human NK cells

Name	Full Name	Description / Applicability	References
NSG	NOD SCID Il2rg−/−	Human NK cell subsets develop but these NK cells need pre-stimulation with IL-15 to kill K562 or exert ADCC.	8
NOG-IL2	NOD SCID Il2rgnull hIL2 tg	Improved development and function of human NK cell subsets (compared to NSG); NK cells killed K562 and exerted ADCC of CCR4+ Hodgkin’s lymphoma cells in vivo.	9
NOG-IL15	NOD SCID Il2rgnull hIL15 tg	Mice were not engrafted with CD34+ cells. Instead, NK cells were isolated from human PBMC and transferred into NOG-IL15 mice. Transferred NK cells killed K562 in vivo. In vitro-expanded NK cells exerted ADCC of gastric carcinoma cells (NCI-N87).	10
NSG-IL15	NOD SCID Il2rg−/− hIL15 tg	Mouse strain available at The Jackson Laboratory (JAX): https://www.jax.org/strain/030890; no published data available yet.	-
NSG-IL7-IL15	NOD SCID Il2rg−/− hIL7 KI hIL15 KI	Improved development of human NK cell subsets, including tissue-resident CXCR6+ NK cells (compared to NSG); improved NK cell cytotoxicity towards K562 (in vitro)	11
SRG-15	Balb/c x 129 Rag2−/− Il2rg−/− hSIRPA KI hIL15 KI	Improved development of human ILC1 and NK cell subsets, including tissue-resident ILC1 and NK cells; mass cytometry revealed similar phenotypic and functional diversity between NK cells in SRG-15 and humans; NK cells killed K562 and exerted ADCC of CD20+ Burkitt’s lymphoma cells in vivo.	7
NSG-SGM3	NOD SCID Il2rg−/− hIL3/hGMCSF tg hSCF tg	Improved development of human myeloid cells (compared to NSG), which produce hIL-15 and thereby improve development of human NK cells; no data available about NK cell-mediated killing of tumor cells.	12,13
MISTRG	Balb/c x129 Rag2−/− Il2rg−/− hMCSF KI hIL3/hGMCSF KI hSIRPA KI hTHPO KI	Improved development of functional human myeloid cells (compared to NSG), which produce hIL-15 and thereby improve development of functional human NK cells; NK cells killed an MHC class I-deficient B cell line LCL721.221 in vivo.	14
BRGSF	Balb/c Rag2−/− Il2rg−/− SIRPANOD Flk2−/−	Flt3L treatment necessary (6 injections with 5 μg human Flt3L-Fc) to improve development of human myeloid cells, which produce hIL-15 and thereby improve development of human NK cells and ILC1; poly(I:C) i.p. induces IFN-γ and CD107a in NK cells; no data available about NK cell-mediated killing of tumor cells.	15

Abbreviations: NOG: nonobese diabetic (NOD), severe combined immunodeficient (SCID), interleukin 2 receptor gamma (Il2rg) null; NSG: NOD, SCID, Il2rg knockout; SRG-15: Rag2 knockout, Il2rg knockout, human Sirpa knock-in, human Il15 knock-in; MISTRG: Rag2 knockout, Il2rg knockout, human Sirpa knock-in, human Thpo knock-in, human Il3/Csf2 knock-in, human Csf1 knock-in; BRGSF: Rag2 knockout, Il2rg knockout, Sirpa^NOD, Flk2 knockout.

In this chapter we focus on protocols in SRG-15 humanized mice. In our study we intercrossed RG mice with S^h/hRG-15^h/h mice to obtain experimental mice that have a heterogeneous expression of SIRPA and IL-15 (S^h/mRG-15^h/m; h=human, m=mouse; see Note 1)⁷. Knock-in replacement of the mouse Il15 coding sequence by the human IL15 coding sequence had the advantage of proper expression of physiological levels of IL-15 in a tissue- and cell-specific manner⁷. SRG-15 mice reconstituted with human CD34⁺ HSPCs promoted efficient development of circulating and tissue-resident NK cell subsets. Mass cytometric analysis further confirmed phenotypic and functional heterogeneity similar to NK cells in human peripheral blood (Herndler-Brandstetter et al., Figure 4)⁷. SRG-15 humanized mice also enabled NK cell-mediated killing of MHC class I-deficient tumor cells (K562) as well as NK cell-mediated antibody-dependent cellular cytotoxicity (ADCC) of Burkitt’s lymphoma cells in vivo⁷.

Next-generation humanized mice such as SRG-15 can therefore facilitate translational research by enabling the study of human ILC1 and NK cells and by facilitating the development / testing of novel human ILC1/NK cell-based therapeutic approaches that target infections and malignancies.

2. Materials

2.1. Isolation of human CD34⁺ hematopoietic stem and progenitor cells (HSPCs)

1. 50 mL conical centrifuge tubes.

2. Ficoll-Paque PLUS density gradient media.

3. 10 mL serological pipette.

4. 3 mL Pasteur Pipette.

5. EasySep™ Human Cord Blood CD34 Positive Selection Kit II” (STEMCELL Technologies, #17896) or EasySep™ “Human CD34 Positive Selection Kit II” (STEMCELL Technologies, #17856).

6. Cold FACS buffer (1–2% FBS, 1mM EDTA, 0.1% sodium azide).

7. Fluorochrome-conjugated anti-human CD34 antibody, clone 561 (BioLegend), 7-AAD Viability Staining Solution (BioLegend).

8. Flow cytometer (e.g. LSR Fortessa)

9. Cold freezing medium (90% fetal bovine serum [FBS], 10% dimethyl sulfoxide [DMSO]).

10. Pre-cooled 1.5–2 mL cryovials.

11. Pre-cooled cell freezing container (Mr. Frosty™ filled with isopropanol or isopropanol-free Corning™ CoolCell™; all Thermo Fisher Scientific). Change the isopropanol in Mr. Frosty™ freezing containers after five freeze-thaw cycles.

2.2. Intrahepatic injection of human CD34⁺ HSPCs

1. 10 cm² petri dish.

2. Sterile phosphate-buffered saline (PBS), 70% ethanol, sterile Eppendorf tubes.

3. X-RAD 320 or another X-ray irradiator.

4. 100 μL Hamilton syringe with a 26G Hamilton needle.

2.3. Analysis of engraftment

1. EDTA-coated tubes or 5 mL round-bottom flow cytometry tubes with heparin (1,000 USP units per mL; Sigma-Aldrich).

2. Red blood cell (RBC) lysis buffer (BioLegend).

3. Cold FACS buffer (1–2% FBS, 1mM EDTA, 0.1% sodium azide).

4. Fluorochrome-conjugated antibodies as listed in Table 3.

Table 3:

Analysis of human NK cell differentiation and function in the bone marrow of humanized mice

Panel	Antibody	Clone	Vendor
Overall engraftment	mCD45	30-F11	BioLegend
	hCD45	HI30	BioLegend
	hCD3	SK7 or UCHT1	BioLegend
	hCD33	WM53	BioLegend
	hCD19	HIB19	BioLegend
	hNKp46	9E2	BioLegend
	hCD56	HCD56	BioLegend
NK cell development	hCD45	HI30	BioLegend
	hCD34	561	BioLegend
	hCD10	HI10a	BioLegend
	hCD45RA	HI100	BioLegend
	hNKp46	9E2	BioLegend
	hCD56	HCD56	BioLegend
	hCD16	3G8	BioLegend
	hCD94	HP-3D9	BioLegend
	hCD117	104D2	BioLegend
	hCD3	SK7 or UCHT1	BioLegend
NK cell maturation	hCD45	HI30	BioLegend
	hNKp46	9E2	BioLegend
	hCD56	HCD56	BioLegend
	hCD16	3G8	BioLegend
	hCD158	HP-MA4	BioLegend
	hCD158b	Dx27	BioLegend
	hCD158e1	DX9	BioLegend
	hCD3	SK7 or UCHT1	BioLegend
	hCXCR6	K041E5	BioLegend

To analyze human NK cell differentiation in the bone marrow of humanized mice, the following markers are used: Pro-NK cells: CD34⁺ CD10⁺ CD45RA⁺ CD117⁻; Pre-NK cells: CD34⁻ CD10⁻ CD45RA⁺ CD117⁺; immature NK cells: CD56⁺ CD117⁺ CD94⁻ CD16⁻ CD10⁻ CD34⁻; mature NK cells: CD94⁺ CD56^bright CD16⁻ (CD56^bright NK cell subset) and CD94⁺ CD56^dim CD16⁺ (CD56^dim NK cell subset).

2.4. Tumor xenografts

1. Isoflurane vaporizer with isoflurane and oxygen.

2. Small electric shaver.

3. PBS, 70% ethanol, cold FACS buffer.

4. Disposable, 1 mL 30G insulin syringe.

5. Raji cells (CCL-86; ATCC).

6. Rituxan® (Rituximab; Genentech/Roche).

7. Forceps, surgical scissors, disposable scalpel.

8. 70 μm cell strainer.

9. 10% FBS / RPMI 1640 supplemented with collagenase D (Sigma-Aldrich).

3. Methods

3.1. Isolation of human CD34⁺ HSPCs

The isolation of human cord blood CD34⁺ hematopoietic stem and progenitor cells (HSPCs) is performed using the EasySep™ “Human Cord Blood CD34 Positive Selection Kit II” (STEMCELL Technologies, #17896) according to the manufacturer’s instructions. The isolation of human CD34⁺ cells from other samples, such as fresh or frozen G-CSF mobilized peripheral blood or bone marrow mononuclear cells, or from frozen cord blood mononuclear cells, is performed using the EasySep™ “Human CD34 Positive Selection Kit II” (STEMCELL Technologies, #17856). Alternatively, the human CD34 MicroBead Kit, an MS/LS column and a MACS Separator (all Miltenyi Biotec) can also be used to isolate human CD34⁺ cells. The human CD34⁺ cell isolation protocol is described in detail below.

1. Transfer cord blood into 50 mL conical centrifuge tubes, add RosetteSep™ Cocktail (5μL/mL of sample), mix and incubate for 20 minutes at room temperature.

2. Add Ficoll-Paque PLUS to a new 50 mL conical centrifuge tube.

3. Slowly layer the sample on top of the density gradient medium to a 1:1 volume ratio using a 10 mL serological pipette.

4. Centrifuge at 500 × g for 30 minutes at room temperature and with the brakes OFF.

5. Harvest the mononuclear cells at the interface using a 3 mL Pasteur pipette (or a 10 mL serological pipette) as depicted in the Ficoll-Paque PLUS manufacturer’s instructions.

6. Proceed to next steps for human CD34 positive selection according to the manufacturer’s protocol (see the two different protocols mentioned above).

7. Check purity and viability of isolated human CD34⁺ HSPCs by flow cytometry. Take a small aliquot and add 50 μL cold FACS buffer (1–2% FBS, 1mM EDTA, 0.1% sodium azide).

8. Add fluorochrome-conjugated anti-human CD34, clone 561 and 7-AAD Viability Staining Solution (all BioLegend) and stain cells for 20–30 min on ice and protected from light.

9. Add 1 mL cold FACS buffer, spin cells at 300 × g for 10 min and discard supernatant.

10. Resuspend cell pellet in 150–200 μL cold FACS buffer and analyze on a flow cytometer (e.g. LSR Fortessa).

11. Proceed to transcutaneous injection of human CD34⁺ cells into newborn mouse liver.

12. Alternatively, cryopreserve purified human CD34⁺ cells. Spin CD34⁺ cells at 300 × g for 10 min and discard supernatant.

13. Resuspend cells at 1 × 10⁶ cells/mL in cold freezing medium (90% FBS, 10% DMSO). Dispense cell suspension aliquots (max. 1 mL per tube) into pre-cooled 1.5–2 mL cryovials and place the cryovials inside a pre-cooled cell freezing container (Mr. Frosty™ filled with isopropanol or isopropanol-free Corning™ CoolCell™, all Thermo Fisher Scientific). Place the freezing container in a −80°C freezer overnight and transfer to liquid nitrogen the following day.

3.2. Intrahepatic injection of human CD34⁺ HSPCs

Newborn mice (2–5 days old) are used for engraftment. Transplantation is performed by transcutaneous injection of human CD34⁺ HSPCs into newborn mouse liver as summarized below:

1. Place newborn mice in a 10 cm² petri dish for sublethal X-ray irradiation with an X-RAD 320 or another irradiator. Irradiation doses for the different humanized mouse strains are shown in Table 2. Although optional, irradiation is also recommended for the MISTRG strain (see Note 2).

2. Resuspend human CD34⁺ cells at desired concentration in sterile PBS in a sterile Eppendorf tube and maintain cells on ice. Injection volume is 20 μL per mouse.

3. Intrahepatic injection of 20 μL human CD34⁺ cell suspension into newborn mice is performed using a 100 μL Hamilton syringe with a 26G Hamilton needle. Disinfect the needle with 70% ethanol and rinse with sterile PBS prior to injection.

4. Intrahepatic injection is performed either by direct transcutaneous injection into the newborn mouse liver or by transcutaneous injection into the liver via the sternal notch as described by Song et al., Figure 1¹⁶. The needle tip is placed into the dark red liver (below the rib cage and above the stomach filled with milk) and the 20 μL human CD34⁺ cell suspension is carefully injected. Wait a few seconds before withdrawing the needle to reduce bleeding and loss of CD34⁺ cells.

5. Place pups back in cage with their mothers. Post-injection survival rate should be higher than 90%.

Table 2:

Engraftment details for different humanized mice to study human NK cells

Strains	IR dose (cGy)	Number of hCD34+ cells	Time to reconstitution	Functional NK cell development
NSG	80	100k	≥12 weeks	No
NOG-IL2	250	50k	4 weeks	Yes *
NOG-IL15	Unknown	Unknown	Unknown	Unknown
NSG-IL15	80	100k	≥12 weeks	Unknown
NSG-IL7-IL15	150	2.5k – 26k	≥12 weeks	Yes
SRG	360	100k	≥12 weeks	No
SRG-15	360	100k	≥12 weeks	Yes *
NSG-SGM3	80	50k	7–8 weeks	Unknown
MISTRG	80 or none	10–50k	7–8 weeks	Yes *
BRGSF	300	200k	8–9 weeks	Yes

^*Studies with these three next-generation humanized mouse stains demonstrated NK cell-mediated killing of tumor cells in vivo (MHC class I-deficient tumor cells and/or ADCC).

3.3. Analysis of engraftment

Analysis of human immune cell engraftment is performed between 7–14 weeks post injection. The time to optimal reconstitution of a human immune system varies among humanized mouse strains (Table 2). Blood can be obtained via retro-orbital plexus, facial veins or mandibular veins, and tail veins.

1. Collect 50–100 μL of blood into EDTA-coated tubes or 5 mL round-bottom flow cytometry tubes pre-filled with 10–20 μL heparin (1,000 USP units per mL).

2. Resuspend blood in 2–3 mL RBC lysis buffer, vortex thoroughly and incubate at room temperature for 5 minutes. Vortex again, spin at 300 × g for 10 min and discard supernatant.

3. Repeat step 2.

4. Resuspend cell pellet containing white blood cells in 50 μL cold FACS buffer.

5. Add fluorochrome-conjugated anti-human (h) and anti-mouse (m) antibodies to determine overall human immune cell engraftment, including human NK cell reconstitution. Incubate for 20–30 min on ice in the dark. Three different staining panels are provided in Table 3 (overall engraftment, NK cell development, NK cell maturation).

6. Resuspend cell pellet in 150–200 μL cold FACS buffer and analyze on a flow cytometer (e.g. LSR Fortessa).

3.4. Tumor xenografts

NK cells can kill MHC class I-deficient tumor cells (e.g. K562 tumor cells) but they can also kill tumor cells, which have been labeled with a tumor-opsonizing antibody via antibody-dependent cellular cytotoxicity (ADCC). The latter one has been utilized for cancer immunotherapy. Mice reconstituted with a human immune system can be used to study and manipulate immune responses to tumors. Here we describe a protocol to study ADCC using a human CD20-expressing Burkitt’s lymphoma xenograft model and Rituximab, a clinically approved anti-human CD20 monoclonal antibody.

1. Resuspend Raji cells (Burkitt’s lymphoma cells; CCL-86; ATCC) at desired concentration in sterile PBS and maintain cells on ice.

2. Place the humanized mouse (7–14 weeks post engraftment with hCD34⁺ cells) into an inhalational anesthesia chamber with a properly calibrated vaporizer for ≥1 minute. Adjust the oxygen flowmeter to 0.8 to 1.5 L/min. Adjust the isoflurane vaporizer to 3% to 5%.

3. For subcutaneous injection, shave the implantation site (right flank) of the mouse. To identify the optimal tumor injection site on the right flank of the mouse, see Yao et al., Figure 2C¹⁷.

4. Clean the injection site with 70% ethanol. Grasp and lift the skin on the right flank with forceps and insert a 30G insulin syringe (1 mL; disposable; use of an insulin syringe leads to minimal loss of fluid) between the skin and the underlying muscles, and gently inject 5 × 10⁶ Raji cells in 100 μL PBS.

5. Monitor tumor development for up to 14 days. Tumor is usually visible 7 days post injection.

6. Inject PBS or 4 mg/kg of rituximab every third day for up to five injections.

7. Measure tumor volume every 1–3 days for up to 2 weeks by caliper measurement according to the following formula: Tumor volume (mm³) = 0.5 × (length × width²).

8. At the endpoint, euthanize the mouse and surgically remove the tumor xenograft underneath the skin using forceps and surgical scissors. Measure the tumor weight.

9. To analyze tumor-infiltrating immune cells, cut the tumor xenograft into small pieces and digest them using 10% FBS/RPMI 1640 supplemented with 1 mg/mL collagenase D in a shaker for 45 min at 37 °C.

10. Pour the tumor pieces / cells into a 70 μm cell strainer and thoroughly squash the tumor pieces and filter the cell suspension through the 70 μm cell strainer.

11. Spin cells at 300 × g for 10 min and discard supernatant.

12. Resuspend cell pellet in 50 μL cold FACS buffer and add fluorochrome-conjugated antibodies. Use the overall engraftment staining panel listed in Table 3.

13. Incubate for 20–30 min on ice in the dark.

14. Add 1 mL cold FACS buffer, spin cells at 300 × g for 10 min and discard supernatant.

15. Resuspend cell pellet in 150–200 μL cold FACS buffer and analyze on a flow cytometer (e.g. LSR Fortessa).