Production of Humanized Mice through Stem Cell Transfer

The development of “humanized” mice has become a prominent tool for translational animal studies of human diseases. Immunodeficient mice can be humanized by injections of human umbilical cord stem cells. The engraftment of these cells and their development into human lymphocytes has been made possible by the development of novel severely immunodeficient mouse strains. Proven protocols for the generation and analysis of humanized mice in the NSG mouse background are presented here. © 2023 The Authors. Current Protocols published by Wiley Periodicals LLC.


INTRODUCTION
Mouse models are widely used in laboratory research and development. The mouse is the preferred experimental animal due to its small size, ease of handling, and short reproductive cycle, as well as the wealth of information that has been generated with mouse models. However, mouse models of human disease pose limitations because of their lack of susceptibility to certain infections and the lack of a human immune system. This impacts the ability to use mice in various fields, for example, the study of human pathogens such as retroviruses. Human immune system (HIS) mouse models are generated by engrafting CD34 + human umbilical stem cells (HUSCs). The NOD.SCID Il2rγ null (NSG) mouse strain has been widely used to perform HUSC engraftment . Immunodeficient NSG mice lack functional mouse T cells, B cells, and NK cells and are deficient in cytokine signaling. After transfer of HUSCs, phenotypically normal human immune cells develop in NSG mice. These cells circulate normally and are susceptible to drugs and infections, but generate only poor adaptive immune responses if any. The Jackson Laboratory has developed an array of NSG mice that are transgenic for a variety of human molecules to improve immune function (https: // www.jax.org/ jax-mice-andservices/ find-and-order-jax-mice/ nsg-portfolio). Their collection includes strains with human cytokines to promote engraftment of human leukocytes (Coughlan et al., 2016), strains that develop functional cytotoxic T cells specific to the transgene human leukocyte antigen class I (HLA-A2) , and strains with knockouts of mouse major histocompatibility complex class II (MHC-II) but inserted with HLA-DQ8 for a human insulin-specific regulatory T cell response (Serr et al., 2016). Basic Protocols 1 and 2 describe the generation of humanized mice by two different methods that are chosen based on the age of the mice and the ease of application. If neonatal mice are humanized within 72 hr of birth, the HUSCs are injected into the liver. If mice are humanized at 4 weeks of age, they are humanized by intravenous tail vein injections. Support Protocols 1 and 2 describe techniques used to handle sensitive HUSCs and monitor leukocyte populations engrafted into humanized mice.
NOTE: All protocols involving live animals must be reviewed and approved by an Institutional Animal Care and Use Committee (IACUC) and must conform to government regulations for the care and use of laboratory animals.

Immunodeficient Mouse Strains
Two mouse strains, NOD.Cg-Prkdc scid Il2rg tm1Wjll (NSG) and BALB/c-Rag2 nul1 IL2rγ null (BRG), are used most often to produce humanized mice. These have either a complete or partial deletion of the IL-2 receptor gamma chain (IL2rγ). The Rag1 or Rag2 genes are inactivated by mutations in NSG mice, whereas both genes are deleted in BRG mice. In both strains, this results in the lack of mature T cells, B cells, or functional NK cells and a deficiency in cytokine signaling (Coughlan et al., 2016;Shultz et al., 2005). Although both strains are proven models for engraftment of HUSCs, one study indicates that NSG mice support HUSC engraftment better than BRG mice (Brehm et al., 2010). NSG and BRG mice are easily obtainable from The Jackson Laboratory and must be bred and maintained in a specific pathogen-free and, if applicable, BSL-2 facility.

Sources of HUSCs
HUSCs are isolated from umbilical cord blood and purified to ≥90% CD34 + cells, a marker for HUSCs. The CD34 + cells are selected using magnetic beads (Miltenyi Clin-iMACS system; Schumm et al., 1999). In addition to umbilical cord blood (Ishikawa et al., 2005), human stem cells can also be purified from various other human tissues, such as fetal liver, bone marrow (Holyoake et al., 1999), and peripheral blood . In our hands, the generation of human lymphocytes in NSG mice was far superior with HUSCs compared to peripheral blood stem cells (unpublished data).
Obtaining human umbilical cord for purification of HUSCs for research requires ethical review approval from an Institutional Review Board (IRB), whereas the purchase of frozen stem cells usually does not require approval. Laboratory-generated HUSCs should be checked by flow cytometry for purity (CD34 expression) and especially the absence of contaminating T cells (CD3, CD4, and CD8 expression), which will lead to a graft-versus-host reaction in mice. Subsequently, small aliquots of purified HUSCs should be tested for adverse reactions (graft-versus-host) in mice. HUSCs from commercial sources are more costly but less labor-intensive to use and usually of very good purity.

HUMAN UMBILICAL STEM CELL ENGRAFTMENT OF NEONATAL IMMUNODEFICIENT MICE
In this protocol, neonatal immunodeficient pups are injected intrahepatically with HUSCs ( Fig. 1; Pearson et al., 2008). Engraftment can also be done intravenously via the intracardiac route or facial vein, intraperitoneally, or by intratibial injection. Prior to engraftment, preconditioning of the pups with low-dose sublethal irradiation is necessary to ensure engraftment, proliferation, and survival of HUSCs.
Once pups are present, watch for cannibalism. Do not waste HUSCs on a litter that will be cannibalized. The presence of a milk spot (stomach filled with milk) indicates that the pups are nursing and are well enough to be injected. The milk spot can be seen through the abdomen skin (Fig. 2).
2. On the day of the engraftment, turn on the X-ray irradiator and initiate the warm-up procedure.
4. Place 24-hr to 48-hr-old pups in a sterile, ventilated cage (typically one litter per cage) in a sterile biosafety hood.
5. Before pups are irradiated and anesthetized, restrain parents and place a small amount of topical, petroleum-based nasal decongestant on their snouts.
This will reduce their ability to detect any scent transferred to the pups during handling, which will increase parent acceptance and reduce potential pup cannibalism.
6. Place the cage containing pups in the RAD+ reflector block of the irradiator chamber (level 1 of the instrument) and irradiate pups for ∼1 min at 1 Gy.  The RS 2000 is a direct replacement for cesium-137 and cobalt-60 irradiators. Each irradiator may vary in dose rate and function. The radiation dose is measured in rad or Gray (100 rad = 1 Gy). Irradiation times correspond with dose rates.
7. Transfer irradiated pups to a weigh boat on ice until they are anesthetized by the cold. Use cessation of movement as a sign of anesthesia (this could take ∼5 min).
Do not place pups directly on ice.
8. Load an insulin syringe with a 26-to 30-G needle with the HUSC suspension.

Restrain pups for intrahepatic injection and dispense 50 μl HUSC suspension into the liver.
A second person is needed to help restrain the pup for intrahepatic injection. For restraint technique and location of liver, see Figure 3.
12. Verify engraftment of HUSCs starting at 10 weeks of age by using flow cytometry to analyze peripheral blood staining for human CD45 + (see Support Protocol 2).

HUMAN UMBILICAL STEM CELL ENGRAFTMENT OF 4-WEEK-OLD IMMUNODEFICIENT MICE
To generate humanized mice at the age of 4 weeks, mice are injected with HUSCs by intravenous tail injection. This method can be used for NSG or NSG.HLA-A2 strains. 2. Count HUSCs and suspend in medium at 3-10 × 10 4 CD34 + cells/50 μl.

Additional Materials
3. Place 4-week-old adult mice in a sterile, ventilated cage (up to five per cage) in a sterile biosafety cabinet.
4. Place the cage in the RAD+ reflector block of the irradiator chamber (level 1 of the instrument) and irradiate mice at 1.5 Gy for ∼2 min.
5. Return irradiated mice to their cages.
6. Prior to engraftment, warm one to several mice by placing a heat lamp within a few inches of the cage (this should take ∼30 min).
Heat will increase the blood flow and make the tail vein more visible for easier injection. If the cage is hot to the touch, move the heat lamp further away.
7. Load an insulin syringe with a 28-G needle with the HUSC suspension.
8. Place the first mouse under the dome on the injection platform and then slide the tail through the slot so that it is facing you. Secure the tail by gently holding it with thumb and index finger, then wipe the tail with an alcohol wipe.
Wiping the tail skin with alcohol will sterilize the surface and improve visibility of the tail vein.
9. Grab the tail with the thumb and index finger from top and bottom and twist it so that one of the four tail veins faces up. 11. Return mouse to its cage and repeat for all remaining mice.
12. After 8 weeks, verify engraftment of HUSCs starting at 10 weeks of age by using flow cytometry to analyze peripheral blood staining for human CD45 + (see Support Protocol 2).

PREPARATION OF HUMAN UMBILICAL STEM CELLS
Human CD34 + hematopoietic stem cells are isolated from human cord blood to alter immunodeficient mice. For sources of HUSCs, see Strategic Planning. This protocol describes how to thaw the cells and prepare them for injection. 2. Wipe the exterior of a vial of frozen cells (typically 10 6 cells) with 70% ethanol and thaw quickly in a 37°C water bath.

Materials
3. Aseptically transfer the entire volume of thawed cells to a 15-ml tube.
4. Rinse vial that the cells were frozen in with 1 ml warm medium and add it dropwise (over ∼1 min) to the 15-ml tube while swirling.
Slow and gentle addition of medium is important to optimize cell recovery.
5. Slowly add warm medium dropwise to the cells, while swirling, until a total volume of 5 ml is reached over a 3-min period.
6. Slowly add medium dropwise up to a total volume of 15 ml over ∼5-10 min, gently swirling after each addition.
8. Remove most of the supernatant by pipet, being careful not to disturb the pellet (leave a few milliliters of supernatant behind).
Save all washes in secondary tubes until the procedure is complete in case cell recovery is needed.
9. Slowly add fresh medium to a volume of 15 ml while gently swirling.
12. Gently resuspend pellet in a known volume of HPGM and count cells.
Determine how many mice will be injected and adjust the volume to use as many of the cells as possible. 14. Place cells in a 37°C incubator until ready to use.

SUBMANDIBULAR BLOOD COLLECTION FROM HUMANIZED MICE AND ANALYSIS OF PERIPHERAL BLOOD VIA FLOW CYTOMETRY
Flow cytometry is an easy method for monitoring human stem cell engraftment in the peripheral blood of humanized mice. Peripheral blood from the facial vein can be collected into a microtainer tube containing EDTA and stained using antibodies specific for human markers. 3. Stain each sample by adding 5 μl of each antibody (typically 0.5 μg per sample). Do not add antibody to unstained controls.

of 13
Current Protocols 4. Cover samples to protect from light and shake at 150 rpm for 30 min at room temperature.
5. While samples are staining, prepare the compensation bead controls: a. Vortex the bottle of beads briefly and then add one drop to a separate flow tube for each fluorophore. b. Add 0.5 μl of each antibody to the respective tubes. Cover and vortex for ∼5-10 s. c. Incubate at 4°C for 30 min. Repeat wash if the pellet remains red.
9. Add 0.5-1 ml flow buffer to each tube and analyze on a flow cytometer.
a. Briefly vortex each sample immediately before analysis. b. Use a flow cytometer that can read four-color cytometry. Collect at least 10,000 events.

Background Information
Humanized mouse models have become increasingly important tools in the study of human diseases. One potential problem in the generation of humanized mice is the development of graft-versus-host disease after transplantation of HUSCs. Signs of graft-versushost disease in mice may include a hunched posture, hair loss, weight loss, and early death. NSG mouse strains have been reported to be the better model for engraftment of HUSCs due to faster expansion of human CD45 + cells, but this may also explain the development of graft-versus-host disease at a faster rate (Ali et al., 2012;Brehm et al., 2010). The absence of T cells in the HUSC preparation is important for avoiding graft-versus-host disease. In addition to the classical T cell-based disease, the development of a macrophage-based granulomatous disease has been described (Huey et al., 2018). Granulomatosis was observed in various organs with both mouse and human macrophages.
Another challenge of the humanized mouse model is the development of human lymphocytes that are not immunocompetent following inoculation of HUSCs (Villaudy et al., 2011).
Part of the problem seems to be that mouse cytokines do not fully support the development of human lymphocytes. Two approaches have been pursued to overcome these hurdles: (1) the use of additional human fetal tissue and (2) the generation of human cytokines in humanized mice. In the first approach, mice are surgically transplanted with human fetal liver and thymus (the BLT mouse), which allows for development of functional human leucocytes (Lan et al., 2006). However, this model requires surgical expertise and the availability of bone marrow, liver, and thymus from a human fetus. Moreover, and the use of fetal tissue for research is not legal in parts of the United States and other countries. Another approach is the generation of a transgenic mouse expressing human MHC and cytokine genes in NSG or BRG mice (Billerbeck et al., 2011;Shultz et al., 2010;Strowig et al., 2011;Willinger et al., 2011), which has resulted in a partial restoration of immune function. A third approach to facilitate the development of humanized mice is to use a viral vector system to deliver human genes into the mouse model prior to injection of HUSCs. NSG mice transduced with adeno-associated

Mouse breeding and care
For establishment of an NSG mouse breeding colony, mice must be kept under specific pathogen-free (SPF) conditions. Mice become sexually active between 6 and 8 weeks of age. Young NSG mice usually breed better than older mice, and performance usually is best if mice are paired at 5-6 weeks of age. Breeding performance and litter size may depend on various factors such as barometric pressure, temperature, humidity, noise, handling of pregnant mice, husbandry enrichments, and the overall health of the breeding pair. In our hands, one-on-one breeding (mating one male and one female) has established stable breeding and reduction of cannibalism. In contrast, trio breeding and continuous breeding have been reported but may not be as successful. The traditional mouse diet contains 6% fat (LabDiet, cat. no. 5K52; Teklad LM-485, cat. no. 7912). Breeding success can be supported by feeding a diet containing higher fat and therefore higher energy content (e.g., Teklad S-2335, cat. no. 7904). Under optimal breeding conditions, an NSG litter should an average of eight pups per litter with a range of two to twelve pups. The majority of NSG breeders stop breeding by 7-8 months of age. To keep the breeding colony active, breeder pairs that have not had a litter in 60 days or consistently produce small litters should be replaced, as should breeder pairs that cannibalize their pups more than once. To reduce cannibalism after handling pups, applying a small amount of Vicks VapoRub to the parents' snouts (National Research Council, 2003) and gently smearing the pup with bedding from the cage will mask human olfactory scents that have been transferred to the pups.
NSG mice are immunodeficient and therefore prone to infections such as Enterococcus spp., Klebsiella oxytoca, Staphylococcus aureus, Pseudomonas, and Corynebacterium bovis (Foreman et al., 2011), which can easily be dealt with by immunocompetent mice. To keep NSG mice healthy and reduce disease outbreaks, proper aseptic handling and housing practices should be performed under strict barrier conditions. Suggested strict bar-rier practices include the use of microisolator cages; autoclaving of all food, water, bedding, and cages prior to entering the room; and manipulating cages inside a biosafety cabinet. All personnel should wear personal protective equipment including disposable gowns, hair bonnets, masks, and disinfected gloves. All cages should be changed at least biweekly, although weekly changes are ideal. In monitoring animal health, the testing of microorganisms typical for immunodeficient animals (see above) must be included.
In recent years, C. bovis has become a pathogen of concern. It is a gram-positive bacterium identifiable by clinical signs of alopecia and dermatitis. It has been demonstrated that C. bovis interferes with human cell engraftment of immunodeficient mice (NSG) whether the mice are symptomatic or asymptomatic (Vedder et al., 2019). Transmission of C. bovis occurs by keratin flake shedding by mice and by fomites that contaminate the facility by personnel. Because it is also transmitted through contaminated cell lines, all cell lines should be tested prior to injection. It is highly recommended to routinely monitor the presence of C. bovis by PCR from feces, maintain sterile housing conditions by vaporized hydrogen peroxide (VHP), and meticulously clean equipment with a disinfectant containing accelerated hydrogen peroxide (aHP) such as Oxiver (Miedel et al., 2018).
Another crucial parameter for the generation of humanized mice is the state of the HUSCs. Human CD34+ cells are isolated from cord blood mononuclear cells via positive immunomagnetic separation (Schumm et al., 1999) and should express CD34 + to ≥90%. HUSCs are sensitive and diminish in their efficacy if not thawed properly (see Support Protocol 1).

Host mouse strains
Many different strains of immunodeficient mice have been developed that are based on SCID mice or the understanding derived from this strain that RAG1 and RAG2 are crucial for B cell and T cell receptor development. Nonobese diabetic/severe combined immunodeficient (NOD/SCID) mice lack functional B cells and T cells and have reduced NK cell activity (Shultz et al., 1995). The creation of NOD.Cg-Prkdc scid Il2rg tm1Wjl /SzJ (NSG) and BALB/c-Rag2 -/-Il2rg -/-(BRG) mice that lack the IL2rγ gene has led to effective hosts for HUSC engraftment. The lack of the IL2rγ chain blocks NK cell development and results in further defects in innate immunity

of 13
Current Protocols (Cao et al., 1995). The irradiation of mice before injection of HUSCs can pose an equipment problem depending on the institutional setup. Recently, a NOD.Cg-Kit W-41J Tyr + Prkdc scid Il2rg tm1Wjl /ThomJ (NB-SGW) mouse was developed to support engraftment of HUSCs without irradiation (Jackson Laboratory, cat. no. 026622). This model is a cross between NOD.Cg-Prkdc scid Il2rg tm1Wjl /SzJ (NSG) and C57BL/6J-Kit W-41J /J (C57BL/6. Kit W41 ). In our hands, the level of human lymphocytes in these mice after reconstitution with HUSCs was comparable to that in NSG mice after irradiation. NB-SGW mice also had 9× higher numbers of human lymphocytes compared to nonirradiated NSG mice at 12 weeks post engraftment (McIntosh et al., 2015). This model provides an excellent alternative to researchers who do not have access to an irradiation source.
Transgenic mice based on the immunodeficient NSG mouse model are a useful tool for investigating human diseases, host immune responses, and the effects of gene knockin and knockout (https:// www.jax.org/ jax-miceand-services/ find-and-order-jax-mice/ nsgportfolio). Immunodeficiency allows engraftment of HUSCs due to deficiencies in functional T, B, and NK cells, which can then be specialized for specific research needs by introduction of human genes. The Jackson Laboratory provides a plethora of transgenic mouse strains that are updated regularly. For example, they produce the NSG.HLA.A2 mouse (strain #014570) that is immunodeficient, expresses HLA class I heavy and light chains, and produces functional CTLs that are HLA.A2 specific. In addition, the transgenic mouse NSG-HLA-DQ8 (strain #026561) expresses HLA-DQ8 and has the knockout allele H2-Ab1. When these mice are humanized, they do not express mouse MHC class II but do produce a human insulin-specific FOXP3+ Treg response to insulin mimetopes (Serr et al., 2016). Another useful option is the NSGS mouse (#:013062), which is immunodeficient and carries human IL-3, GM-CSF, and SCF. These human transgenes aid in the development of myeloid and regulatory T cells (Coughlan et al., 2016).

Host age
An alternative to neonatal engraftment of HUSCs is engraftment in adult mice. When deciding the appropriate age for engraftment, one must consider the time and cost advantages of managing and engrafting mice. Neonatal engraftment can be optimal because pups stay with the parents until weaning, reducing per diem housing costs. The disadvantage of neonatal mice is that they must be injected within 24-72 hr after birth (Brehm et al., 2010) and cannibalism is more widespread than with adults. Nonetheless, we have been successful at humanizing NSG mice at 72 hr after birth. It has been reported that engrafted neonatal NSG mice support human T cell development better than engrafted adults (Brehm et al., 2010), but we have not found a difference in T cell development between mice humanized at birth versus 4 weeks of age. Engraftment of 4-week-old animals may be useful when additional manipulations are necessary. Another consideration may be the maturation of the animals. After infection with HTLV-1, mice humanized at birth succumbed faster to leukemia development than mice humanized at 4 weeks of age (Phelps and Niewiesk, unpublished data).

Alternate engraftment approaches
Engraftment methods differ between neonatal and adult mice. Basic Protocol 1 describes intrahepatic injections (Traggiai et al., 2004) in neonates, but other routes for neonates include intraperitoneal (Gimeno et al., 2004), intravenous via the facial vein (Ishikawa et al., 2002), and intracardiac (Brehm et al., 2010). Humanizing adult immunodeficient mice can be done by intravenous  and intratibial injections.

Neonatal HIS mice
Human peripheral blood lymphocytes (HP-BLs) are detectable in the blood of HUSCengrafted neonatal mice via flow cytometry. Human CD45 + cells are detectable as early as 4 weeks post-inoculation. The earliest time at which a human T cell population is detectable is 10-12 weeks postinoculation. If engraftment was successful, the percentage of human CD45 + cells is 10%-60% of the total white blood cells at this time point. The remaining cells are mouse cells (mostly macrophages and neutrophils) and can be summarily stained with antibodies against mouse CD45 or lineage-specific markers.

Adult HIS mice
For humanization of adult mice, different strains may be used and we have been successful with both NSG and NSG.HLA-A2 mice. Human CD45+ cells are detected by flow cytometry and are routinely measured at 8 weeks

of 13
Current Protocols after humanization. In successfully engrafted mice, human CD45+ cells comprise 10%-55% of the total leukocyte population. There is no significant difference between NSG mice and a transgenic mouse strain (NSG.HLA-A2) in our experiments. T lymphocytes are present at this point but make up less than 10% of the total population. Most leukocytes in NSG mice are human B cells (∼80%), with the remaining population being composed of mouse leukocytes.

Time Considerations
The time required to inject and engraft neonatal NSG mice depends on litter size, stem cell preparation, irradiation, injection, and the time needed for human cells to develop. Breeder mice must be monitored during the gestation period (∼21 days). Before preparation of HUSCs, it is crucial to check the pups for the presence of the milk spot (which indicates that pups are viable) and for the absence of cannibalism. Preparation of injection of HUSCs can be accomplished in ∼ 4 hr. First, preparation of the cells takes ∼2-3 hr. Pups are then exposed to radiation at 1 Gy to aid in HUSC engraftment. The irradiation time will depend on the emission of the radiation source. Most irradiators require <5 min for a 1-Gy exposure. After anesthetization on ice (∼5 min), an experienced researcher can complete intrahepatic injections for a litter of eight pups in ∼8 min. Pups are then returned to their parents and the development of human cells can be verified after 10-12 weeks postinoculation.
For 4-week-old mice, there are a few differences in the amount of time required for humanization. Mice are used at 4 weeks of age for easier injection of the tail vein. Irradiation takes only 2 min, but the animals must be warmed for 30 min before injection. An experienced researcher can complete injections for ten mice in ∼30 min. After humanization, it takes 8 weeks for human leukocytes to develop and become detectable in the mouse.