Update of humanized animal disease models in studying Graft-versus-host disease

ABSTRACT

Graft-versus-host disease (GVHD) is a severe adverse effect that results from bone marrow or peripheral blood cells transplantation and has a high rate of mortality. About 50% of the patients are accompanied with acute Graft-versus-Host Disease (aGVHD) after bone marrow cell transplantation and need systematic treatment. It has an important clinical significance to evaluate the prevention and treatment effects of GVHD. The stable and reliable approaches of humanized animal models are crucial for advancing on the study the biology of GVHD. Relative models transplanting the human immune cells into the mouse body can trigger immunoreaction similar to the humans. As it is a disease triggered by human immune cells, any intervention research prior to clinical treatment has more clinical interrelations compared with the general animal models. In this review, we update the current understanding on humanized animal disease models on studying Graft-versus-host disease and expect to provide more theoretical basis to further study on Graft-versus-host disease.

KEYWORDS:

• Humanized animal models
• Autoimmune diseases
• Graft-versus-host disease
• NOG-SCID mice

Introduction

In these days, cell therapy using immune cell and stem cell that mainly transplants the external cells into the body of patient demonstrates a potential promise to treat and even cure some diseases. However, a series of Graft-versus-Host Disease (GVHD) complications become an obstacle that needs to be overcome. For instance, allogeneic Hematopoietic Stem Cell Transplantation (allo-HSCT) is an effective method to cure the hematological malignancy currently,¹ and it has been favored in the treatment of related diseases in recent years. Nevertheless, the occurrence of the main complications of GVHD seriously reduces the survival rate and life quality of the patients after transplantation.² About 50% of the patients are accompanied with acute Graft-versus-Host Disease (aGVHD) after transplantation and need systematic treatment.³ Particularly, the most severe aGVHD usually goes to the patient who receives an identical zygosity transplantation from unrelated donors.⁴ It is generally recognized that GVHD is an immune attack that targets on the organs of recipients by the T cells from the donors. Subsequently, multi-organ and multi-tissue damages occur, with the mainly involved target organs including immune system, skin, liver and digestive tract.^5-7

As the main complication after allo-HSCT, GVHD is currently believed to occur mainly in three steps: 1) pre-transplant and pretreatment injuries: include the tissue damages caused by total body irradiation (TBI) and large-dose chemotherapy, etc. 2) activation of the Antigen Presenting Cell (APC), induction of the T cells activation and proliferation. 3) damage and apoptosis of target organs.⁸ As the occurrence of GVHD is closely related to the patient’s age, the type of primary disease, the pretreatment plan (including or excluding TBI), GVHD prevention scheme, gender differences of the recipients and donors, the number of the mononuclear cells for transfusion, number of the CD34+ cells, identical degree of Human Leukocyte Antigen (HLA) zygosity and the graft sources (such as bone marrow stem cells and human peripheral blood mononuclear cells (hPBMC). Nonetheless, the most important reason of aGVHD lies in the HLA type differences of donors and recipients.⁹

Based on the clinical studies, the incidence of GVHD has declined over the past few decades, which largely depends on the improvement of HLA zygosity techniques and the optimization of treatment schemes.¹⁰ The optimized scheme includes the optimization usage and dosage of first-line drugs and the use of second-line treatment. Glucocorticoid has long been regarded as the first-line treatment for aGVHD in clinic practice. Therefore, clinicians and investigators spared no effort for optimizing usage and dosage of glucocorticoid treatment for aGVHD. Systemic treatment of methylprednisolone at 2 mg/kg/day or prednisone at 2.0–2.5 mg/kg/day has long been accepted as standard therapy for grade III–IV GVHD, but higher dose of methylprednisolone (10mg/kg/day) does not offer further benefits. Patients with grade I–II GVHD are more responsive to low-dose systemic glucocorticoid and topical steroid therapy. Treatment of methylprednisolone at 1 mg/kg/day showed similar therapeutic effects as 2 mg/kg/day. Although most patients benefit from glucocorticoid treatment, there are around 40% of aGVHD patients who fail to get improvement and need second-line treatment. Fortunately, as understanding for aGVHD pathophysiology is gradually deepen, more and more therapeutic schemes are developed, including monoclonal antibodies (E.g. monoclonal antibodies targeting TNFα, CD2 or IL-2 receptor alpha), mesenchymal stem cells (MSCs). Generally, the new treatments achieved better therapeutic effects for steroid-resistant aGVHD patients. Currently, due to the lack of suitable donors, less than 10% of the patients have access to the HLA-identical sibling donors.¹¹ According to relevant literatures, it is indicated that about 35% to 45% of the patients receive the HLA haploidentical transplantation, and about 60% to 80% of the patients receive the HLA non-identical transplantation currently.¹² Once the patient is diagnosed with aGVHD, hormone therapy is preferred.¹³ Hormone is an effective method for clinical treatment of aGVHD, and the effective rate can reach 25% to 69%. However, once the hormone therapy is ineffective, it usually has a poor prognosis and short-time survival.¹⁴

Currently, there are two main prevention methods for aGVHD clinically, one is to use immunosuppressor for the recipients, the other is to treat with the donor’s hematopoietic stem cell graft and remove the T cells (TCD) from the graft. While the TCD method can soothe the GVHD, it weakens the GVL function and significantly increases the infection and recurrence rates.¹⁵ There is still no effective cure for GVHD. Therefore, study using a stable and reliable clinic approaching aGVHD mice model is extremely important for intensive understand on the pathogenesis, prevention and treatment measures of aGVHD.

Humanized GVHD animal model

From the perspective of pathophysiology, GVHD is an excessive inflammation reaction caused by the donor-source immune competent cells attacking the host cells and organs in vivo.^16-18 GVHD is represented by immunoreaction mainly triggered by T cells and tissue damages caused by clone proliferation of the donor-source T cells after contacting and identifying the antigens of the recipient, or the T cells identifying the MHC of the recipient’s target cells directly or secreting cytokines indirectly. It is a complex pathophysiological process.^19-21 GVHD mainly affects the skin, gastrointestinal tract, liver and bone marrow, because these tissues specifically express the MHC I-type and II-type molecules and contain hematopoiesis-source APCs.²² Currently most of our understanding of GVHD pathogenesis comes from animal models. Therefore, systemic analyzing the GVHD animal model will help us to understand how to prevent and treat GVHD.²³

GVHD was considered to be a graft-versus-host response activated by the immune competent cells in the graft, which was initially recognized by injecting normal mouse’s spleen cells to the mouse after irradiation, while related studies involving the human cells are only conducted in vitro. However, the results of in vitro test are far from those in the in vivo environment, the humanized GVHD animal model represents a new frontline and plays an important role in GVHD studies.

The Severe Combined Immunodeficient (SCID) mice are highly susceptible to be attacked by xenoimmune competent cells and able to develop GVHD after being implanted with xenoimmune competent cells due to its immunocompromised system in vivo.²⁴ C.B-17-scid mice discovered by Bosma et al.,²⁵ lacking functional T and B lymphocytes, provide a first model termed “SCID-hu” by reconstitution with hPBMC or fetal lymphoid tissues in these mice.^24,26–30 Although T and B cells in SCID mice bodies are not able to differentiate into specific functional lymphocytes, the mouse’s endogenous microenvironment for lymphocytes differentiation is normal, and it can still provide the necessary conditions for external lymphocytes differentiation.³¹ To reconstruct the human immune function in SCID mice bodies is a research hotspot of humanized GVHD model.

Current reports mainly use the transplantations of human embryonic tissues such as fetal bone marrow, fetal thymus, fetal liver, fetal lymph node, and the transplantation of the adult’s peripheral blood lymphocyte.³² As the human peripheral blood lymphocyte is more available than the human embryonic tissue and spleen cells, and it is not easy to unify the embryonic ages of different embryonic tissues, so the human PBMC-SCID model established by transplanting the peripheral PBMC to the SCID mice is preferred.³³ Although injecting the human PBMC can induce pathogenesis for mice, we found that to apply the PBMC modulating T cell subgroups with the removal of CD25⁺ cells could obtain more stable xeno-GVHD induction.^34-36

It has been reported in recent years that the human tissues have been transplanted into SCID mice, so as to construct the human-mouse chimeric mice model, which facilitates the observation of the behaviors of human tissue cells in vivo and the response to the therapeutic intervention.³² The xenograft model of human to SCID mice has become the most widely used model in the field of humanized GVHD researches.³⁷ Leung et al.³⁸ applied the human to SCID mice xenograft model, respectively implanted the bone marrow, cord blood and peripheral blood stem cells into the mice bodies and observed the abilities of these three substances for inducing heterogeneous GVHD, which is shown as follows: bone marrow > peripheral blood stem cells > cord blood. The occurrence and severity of GVHD are positively correlated with the number of the T cells implanted. Sandhu et al.³⁹ implanted the human peripheral blood lymphocytes into SCID mice bodies and observed that about 100 percent of the mice died within 4 weeks. It is also found that heterogeneous GVHD is mainly induced by CD4⁺ T cells, anti-heterologous antibodies and lymphocyte factors. Huppes et al.⁴⁰ found that the lesion characteristics of human to SCID mice heterogeneous GVHD are represented by vascular inflammation and many human CD25⁺, human leukocyte antigen HLA-DR⁺ and CD4⁺ lymphocytes presenting tumor-like infiltration in the lymphatic hematopoietic organs of mice. Unlike the allogeneic GVHD recipients, whose bowel, skin and bile duct are destroyed by donor’s lymphocytes. At present, human to SCID mice heterogeneous GVHD model has become an ideal in vivo model for studying the human immune system. GVHD is easily induced in the SCID mice bodies. However, the life time of these immunodeficient mice is very short and they could spontaneously produce functional lymphocytes with age; even in SCID mice, only a low level of human cells was reconstituted because of the remaining host innate immune system.

Nonobese diabetes (NOD)/SCID mice that were modified from SCID mice are better recipients for engraftment of hPBMC than C.B-17-scid mice, probably because of reduced levels of natural killer (NK) cell activity and additional deficiencies in innate immunity.^41,42 The appearance of NOD-SCID mice is just like the normal ones, and it hold the normal weight as well. Nevertheless, the NOD-SCID mice has more defects in innate immunity: all the T and B lymphocyte function tests are negative, lacking of NK cell activity. There are developmental and functional defects in the medullary systems, lacking of complement activity, not able to activate the classic or by pass complement hemolytic activity. These immune deficiencies contribute to the survival of the graft in the NOD-SCID mice bodies.³²

Furthermore, the immunological reconstruction of NOD-SCID mice is more superior.⁴³ After intravenous inoculation of human umbilical cord blood for the NOD/ltsz-SCID mice after irradiation, the level of human cell grafts in the bone marrow is five to ten times higher than that of the C.B-17-SCID mice. Similarly, the implantation of human bone marrow hematopoietic stem cells has been conducted on mice as well. After implantation, human cells can be detected both in bone marrow and peripheral blood of the mice, and high level of human cell grafts can last up to four months in the bone marrow. Injecting human hematopoietic stem cells into NOD/ SCID mice can reconstruct the hematogenous mechanism of the bone marrow.

However, these models still have several problems including the need for sublethal total body irradiation, and the large number of hPBMCs required to induce GVHD, and instability in the timing of onset of GVHD symptoms. In addition, intravenous transplantation of hPBMCs into these animals failed to induce GVHD.⁴⁴ In clinical cases, the cell transfer route is actually restricted through veins but not through the peritoneal cavity in the therapy for bone marrow transplantation. In this sense, the possibility of intravenous transplantation of hPBMCs may be an important issue for xeno-GVHD animal models.

Van Rijn et al.⁴⁵ developed a new model using H-2^d-RAG2^null IL2rγ^null mice lacking T, B, and NK cells. This model has advantages compared with NOD/SCID mice, including no leaky lymphocytes or lymphoid tumor formation, and higher engraftment of human T cells than in previous xeno-GVHD models.

R. Ito et al. thought this model still has disadvantages because a large number of hPBMCs (3.0x10⁷) and total body irradiation are required to induce GVHD. They⁴⁶ then developed a sensitive xeno-GVHD model by transferring PBMCs into NOG mice. After intravenous transplantation of hPBMCs, NOG mice showed early onset of GVHD symptoms and a small number of hPBMCs (2.5x10⁶) was sufficient to induce GVHD, when compared with BALB/cA-RAG2^null IL2rγ^null and NOD/SCID mice. In addition, total body irradiation was not always necessary in the present model. Additionally, they also established a novel xeno-GVHD model for investigating the pathological role of human CD4⁺ or CD8⁺ T Cells. Human CD4⁺ or CD8⁺ T cells were purified from peripheral blood and were transplanted into immunodeficient NOD/Shi-scid IL2rg^null (NOG) mice to establish a novel xeno-GVHD.⁴⁷

NOD-scid IL2rg^null (NOG or NSG) mice are thought the superior immunodeficient mice for receiving a xenotransplant.^48-50 These humanized mice are able to mimic various human diseases, including cancer, infectious, allergic, and autoimmune disorders.^51-56

Although transplanting the human immune cells into the mouse could trigger immunoreaction similar to the humans, these small animals represented by rodents greatly differ from the humans both in biological phenotype and genetics, being not able to fully reflect the real condition in the human body. Therefore, it is important to develop more appropriate animal models that are closer to the real condition of the humans in the future study of xenotransplantation.

The large mammals have the closer genetic relationship with humans, and more similarities with humans are found in genetic structure, protein expression and the internal environment. The experimental results generated by taking the large mammals as the experimental model are much closer to the clinical practice. Pig has many advantages such as the relatively abundant source, anatomical and physiological functions similar to the humans, easy to be genetically modified, so it has been considered as the most suitable source of organs for xenotransplantation. Particularly, the successful application of transgenic technology in pigs represented by the pigs transferring the decay-accelerating factor genes and successful breeding of inbred line pigs will promote the research and application of xenotransplantation. To establish a humanized GVHD model in large mammals is more attractive to the research on GVHD’s occurrence and development.

Zanjani et al.^57-60 observed that the human cells are long-term chimeric in the bone marrow and peripheral blood of the lamb based on xenograft model of human hematopoietic stem cells to fetal lambs, and they also found that implanting the mature human T lymphocytes simultaneously may induce heterogeneous GVHD. Intravenously inject the human bone marrow into the body of newborn pigs, giving or not giving the immunosuppressive agents. After the operation, it is observed that there are long-term low levels of chimeric human cells in the bodies of the pigs.⁶¹

Conclusions

We propose that SCID,NOD-SCID or NOG-SCID mice as satisfying animal strains for induction of GVHD’s xeno model. Immune system of these mice is impaired. Therefore, as the recipients implanted xenogeneic immunoreactive cells, they were vulnerable to the attack of heterogeneous immune active cells and induces GVHD. When hPBMCs were injected into these mice, which could make the mice reconstruct immune system, and the immune system of the mice was very close to the human immune system. It showed various kinds of human Ig, the peripheral blood and the lymphoid organs could detect human lymphocytes, and the human lymphoblastic cells could be used in mice in the body of tetanus toxin. Hepatitis B core antigen forms primary or secondary immune response. Therefore, SCID-hu,NOD-SCID-hu,and NOG-SCID-hu as xeno models can serve as an important tool for studying GVHD.

The animal models that are more similar to the human body system is helpful for one to learn more about the pathogenesis and development diseases, as well as targeting prevention and treatment. The development of humanized animal disease models for studying GVHD has undergone SCID mouse and NOD-SCID mouse to NOG-SCID mouse (Table 1). However, there are still many problems that are difficult to deal so far. The host immune system in patients with GVHD maybe still partially active, unlike in the xeno GVHD model in which severely immunodeficient mice were used. This makes it difficult to investigate the actual mechanisms of how donor T cells are activated or suppressed by host-derived antigen presenting cell, cytokines, or other factors in patients with GVHD. At present, we usually adopt the small animals, such as mice to set up the humanized animal model. These small animals represented by rodents greatly differ from the humans both in biological phenotype and genetics, being not able to fully reflect the real situation in the human body. On the other side, the large mammals have the closer genetic relationship with humans, and more similarities with humans are found in genetic structure, protein expression and the internal environment. The experimental results gained by taking the large mammals as the experimental model are much closer to the clinical practice. However, it is more difficult to build the large animal model with GVHD than the humanized small animal model and few relevant reports are found owing to the followings: 1) llarge animals have no inbred lines and no serious immune deficiency strains similar to SCID mice. Their genetic backgrounds are complex and individual differences vary considerably. 2) most of the immunoreagents widely used in current researches are specific for mice, and relatively few reagents are specifically designed for large animal experiments. Additionally, there are few methods for large animals GVHD detection, and the indicators are unclear, leading to relatively difficult experimental testing. 3) large animals are bigger in size, their own immune system networks are more complex and sound, and they hold stricter requirements on the number and quality of the implanted immune competent cells from the donors. In addition, the feeding cost of large animals is high, the operations on them are complicated, and it is difficult to implement anesthesia on them. Even though there are many problems for us, these animal models could be improved and be widely used as the operating tools of translational researches in the future.

Table 1. Summary of humanized mouse disease models in studying Graft-versus-host] disease.

Animals	l\.1etbo< ls of nulucing GVHD	Stl’engtbs	Weaknesses	Refs
C.B-17-SCID mice	1. mice received 3 Gy total body irradiation prior to reconstitution with human cells.2. Ficoll-Hypaque-separated PBl’v1C frorn healthy donors were injected i.p. into SCIO mice.Each mouse received 20 × 106 PBMC.	1. The first GVHD model;2. The mice showed severe cornbined irnrnunodeficiency symptoms.It can be used for oncology and alIogeneic transplantation,but there \Vere normal NK cells,macrophages and granulocytes.	1. only a low level of human cellswas reconstituted, because of the remaining host innate irnmune system;2. require sublethal total body irradiation;3. a large number of hPBMCs to induce GVHD;4. the timingof onset of GVHD is also unstable.	29
NOD-SCID rnice	1. Mice were irradiated by 2.5 cGy TBI usinga 137Cs source 2–4 h before PBMC injection.2. 20 X 106 CD25-depleted PBl’v1Cwere injected intravenously into mice via tail vein (ref.35);	1. (NOD)/SCID mice are better recipients for engraf tment than C.B-17- SCID mice;2. GVHD model is more stable than C.B-17- SCIO mice.	1. require sublethal total body irradiation2. a Iarge number of hPBl’v1Cs (20 X 106 CD25-depleted PBl’v1C) to induce GVHD.3. instability in the timingof onset of GVHD symptoms.	35
H-2d-RAG2””11 IL2ry””11 m i ce	1. Mice received total body i rradiation with a si ngle dose of 350 cGy {gam m a i rradiation from a l i near accelerator) before i njection of huPBMCs on the sam e day.2 .3x107 PBl’v1C were i nje cted i ntravenousl y i nto m ice	1. no leaky lymphocytes;2. no lymphoid tumor formation;3. higher engraftment of hwnan T cells than NOD/SCID mice and C.B-17- SCID mice GVHD model.	1. a large num ber of hPBMCs (3 × 107 cells) to i nduce GVHD;2. total body i rradiation are requi red to i nduce GVHD;	45
NOG Mice	Purified hum an PBM C, co4• or cos• T cel l swere re-suspended i n PBS and i ntravenousl y i njected (2.5 x 106 cells) i nto the tai l vein of NOG m ice(ref.46, 47);.	1. an extrem ely hi gh engraf tm ent rate of transplanted hum an cellscom pared wi th other im m unodeficient m ice2. could also be im planted in a hi gher proporti on of norm al or canceroushum an cellsand ti ssues;3.NOG m ice showed early onset of GVHD sym ptom s;4.a sm al I num ber of hPBM Cs {2.Sx106) was suffici ent to i nduce GVHD;5.total body i rradiation was not always necessary i n the present m odel.		46,47

Abbreviations

GVHD	=	Graft-versus-host disease
allo-HSCT	=	allogeneic Hematopoietic Stem Cell Transplantation
aGVHD	=	acute Graft-versus-Host Disease
TBI	=	total body irradiation
APC	=	Antigen Presenting Cell
SCID	=	Severe Combined Immunodeficient
hPBMC	=	human peripheral blood mononuclear cells

Disclosure of potential conflicts of interest

The authors confirm that this article content has no conflict of interest.

Additional information

Funding

This review was partially supported by the National Key R&D Program of China (2017YFA0105801), the Natural Science Foundation of Guangdong Province of China (2014A030308005), the Guangdong Introducing Innovative and Entrepreneurial Teams (2016ZT06S252), and the NIH AR059109, AR073409 and STAR grants.