HSCs are the core elements for blood and immune humanization, they are typically derived from fetus, bone marrow and umbilical cord blood. When transplanted into the irradiated immunodeficient mice, these primary human HSCs rapidly developed into T, B and NK cells [94], and could sustain for several months [95, 96]. However, the limited in vitro expansion capacity of HSCs presents a significant obstacle to their broad use in xeno-transplantation. Recently, a breakthrough study has identified chemical agonists and a caprolactam-based polymer that allow for long-term ex vivo expansion of human HSCs. However, despite achieving a 75-fold expansion of total cells and a 55-fold expansion of cord blood CD34${}^{+}$ cells in one month, the expansion efficiency remained limited for adult CD34${}^{+}$ cells form peripheral blood, making it difficult for personalized modeling [97]. In the past decade, PSCs has been proved to be an attractive source for generating various types of blood cells including lymphocytes, myeloid cells and erythrocytes [66]. However, up to date, seldom protocols were able to produce transplantable HSCs from PSCs, but only HSC-like cells with low self-renewal ability and deficient migration and long-term engraftment potential [98, 99]. The most reliable approach was to isolate hematopoietic stem/progenitor cells (HSPCs) from hiPSC derived teratoma parenchyma, which permitted engraftment in secondary mice for reconstructing human immune and hematopoietic system [18], nevertheless pure pluripotent stem cell-derived HSCs without tumorigenesis risk are urgently needed for safety concerns. Transcription factor-mediated specification deserved alternative starting approach to address this issue. Of interest was the report that HOXA9, ERG, and RORA were found to activate self-renewal and multilineage potential in hematopoietic progenitors differentiated from hPSCs. While SOX4 and MYB enabled myeloid and erythroid engraftment. On this basis, multi-lineage progenitor cells were induced from lineage restricted CD34${}^{+}$CD45${}^{+}$ bone marrow precursors. Notably, following transplanted in vivo, the erythroid cells matured and underwent hemoglobin conversion with expression of adult globin, however this strategy still failed to support long-term rebuild of myeloid and erythroid lineages [100]. Later, a modified transcription factor cocktails (ERG, HOXA5, HOXA9, HOXA10, LCOR, RUNX1 and SPI1) has been screened and used to convert hPSC-derived hemogenic endothelium into HSPCs that could differentiate towards multi–lineage in primary and secondary mouse recipients, offering considerable promise for generating humanized animals [66]. To address the complexity and low efficiency for obtain an intermediate hemogenic endothelium state, a single transcription factor MLL-AF4 was transiently expressed to specify hiPSC-derived blood cells into highly engraftable HSPCs. Of note these cells could completely reconstruct the human hematopoietic and immune system in recipient mice without myeloid bias, and sustained long-term engraft capability [67]. However, unlike primary counterpart, hiPSC-derived HSPCs were prone to leukemic transformation during the long-term in vivo engraftment, it may be partly contributed by a decrease of genomic instability of iPSC origin. Besides, the potential safety risks involving in transfected genes cloud not be excluded since most of the screened transcription factor were predominantly found in leukemia, for instance the imbalance of HOX factor prone to lead to acute myeloid leukemia and acute lymphoid leukemia [66, 101]. Thus, safety concerns merit further attentions.

In recent years, chemical compounds have emerged as a potential alternative approach with unique advantages in cell fate conversion and expansion [102]. Aryl hydrocarbon receptor (AhR) antagonists and notch ligand agonists activate cord blood cells expansion in vitro [103], Combinations of UM171, 740Y-P and butanamide, and caprolactam based polymers completely replace exogenous cytokines and albumin to achieve long-term in vitro expansion of human HSCs [97]. While, a cocktail of Bix01294, RG108, VPA and PD0325901 successfully converted human cord blood erythroblasts into induced megakaryocytes, which could produce proplatelets and release functional platelets [104]. More key small molecules and their combinations are under in-depth investigation to promote the production of human HSPCs from hiPSC with higher efficiency, shorten time and lower cost.

Apart from HSCs, numerous terminally differentiated immunocytes have been generated from hiPSCs, and may benefit the studies of specific immune responses from particular immune cell types in humanized models, holding great potential for biomedical applications. hiPSC induced natural killer (NK) cells reportedly showed similar metabolic characteristics and gene expression profiles to adaptive NK cells induced by cytomegalovirus, which owned excellent anti-tumor activity [105], maintain high cytotoxicity and enhance in vivo tumor control in concert with T cells and anti-PD-1 therapy [106], iPSC-NK cells containing chimeric antigen receptors (CAR) provided better options for anti-cancer immunotherapy [107]; hiPSCs induced macrophages displayed similar surface and intracellular markers of human macrophages, as well as functional activities [108], iPSCs-derived CAR-expressing macrophage cells enhanced phagocytosis of tumor cells, and in vivo anticancer cell activity [109], genetically modified iPSC-derived macrophages facilitates drug screening [110]; hiPSCs induced red blood cells were found to express mature adult globin [111], progress in vitro red blood cell generation is expected to address the shortage of blood demand for transfusion application [112]. Moreover, “Designed red cells” with custom phenotypes (Rh invalid, GPB invalid and XK invalid/KEL low) have recently been generated from hiPSCs using CRISPR/Cas9, although the problems of low erythroblast production and poor enucleation rate still existed [113]. Of note, Nakamura et al. [114] transfected and generated a self-renewing and immortalized megakaryocyte line from hiPSC-derived hematopoietic progenitor cells, which permitted massive producing of platelet of clinical quality and quantity. With the development of personalized iPSC bankers, once the differentiated cells (including HSC and their derivations) were used to establish humanized animals, it may bring great opportunity to model hematological disease and cancer for identifying novel diagnostic and targeted therapeutic markers in individual patient.

To generate humanized liver, various human hepatic sources have been tested. Primary human hepatocytes (PHH) have long been considered as the most preferred source for repopulation in the host liver in view of efficiency and function, over 90% repopulation rate has been reported in humanized mouse in the past decade [115]; while in rats, over 30% was achieved by our group very recently [37]. In view of the shortage of PHH, alternative hepatic sources such as fetal liver cells and their derived hepatic stem cells (isolated according to specific surface markers such as) had been explored, these cells permitted quite robust expansion in vivo, over 70% repopulation could be achieved in a relatively short time around 2 months, however the human mature hepatic markers expression were extremely lower, imply the immature of the engrafted cells [16, 116]. Recently, the explore of cell signal modulators, particularly TGFb and GSK3 inhibitors, had paved the way to convert primary hepatocyte, which is normally considered not expandable in vitro, into expandable progenitor state [117, 118]. Of note, in one study, it was reported that when transplanted in vivo, these cells could achieve repopulation level comparable with primary hepatocyte [116]. However, to our practical experience, the expanded cells quickly loss the maturation and transplantation potentials after 3–5 passages. Moreover, the development of differentiation and reprogramming strategy have enabled generation of a large number of donor hepatocytes from pluripotent stem cells, and meanwhile permitted personalized liver disease modeling and drug screening, deserved great promises. Numerous reports have demonstrated that hepatocyte-like cells (HLCs) directedly differentiated from hiPSC could repopulate the damaged mouse liver [119]. Meanwhile hiHeps transdifferentiated from fibroblasts hold similar potentials [92]. A comparation between these two strategies from the same donor identified differences in gene expression for phase II drug metabolism and lipid accumulation, Importantly, both cell types markedly elevated hepatic function following transplantation in vivo, suggesting plasticity advantage for generating functional liver [119]. However, much lower efficiency of repopulation rate (less than 5%) and human albumin secretion (around 100 ng/mL) remained major limitations. Similarly, hepatic stem cells were induced with refined transcription factors combinations. Importantly, they held expansion capability and meantime owned dual-differentiation capability towards both hepatocytes and cholangiocytes, which were supposed to meet the diversity of cellular components in the humanized liver [120]. In our recent study, iPSC-derived hepatic progenitor cells were induced with combination of CHIR99021, A83-101 and FGF2, and expanded in vitro over 20 passages for billions-fold increase, without significant decrease of maturation potential [121], holding great promise for generating functional humanized liver in larger animals beyond rodents.

Although primary human cells, including HSCs and hepatocytes remain the first choice for building effective humanized animals, the shortage of sources, ethical issues, and differences in the genetic background in different sources make them difficult to use in a stable and large scale. Alternatively, iPSC source is considered nearly infinite, with highly homogeneity in the same batch. Furthermore, constant breakthroughs have been made in using in human iPSCs derived cells other than liver and hematopoietic system, such as neurons [122], beta cells [123], alveolar type 2 cells [124], intestinal cell [125], and keratinocytes [79, 126]. Undoubtedly this progress will pave the way to generate more specific and complex humanized models in the near future.

The idea of constructing a transient human immune system in mice dates back to the 1970s. Early experiments sparked interest in modifying the mouse immune system to accommodate cells and tissues from the human immune system. Human peripheral blood mononuclear cells (PBMCs), or human lymphocytes has been isolated from lymph nodes/spleen and injected into the immunodeficient mice, However, these early attempts were limited by the short lifespan of human lymphocytes in the mouse peripheral blood [148]. Subsequently, HSCs transplantation was considered as more effective approach to realized humanization. However, initial attempts to engraft HSCs into irradiated CB17-SCID mice were limited by the high activity of host NK cells and spontaneously generated B and T cells, leading to low levels of human cell engraftment and a risk of graft versus host disease (GVHD) [127, 149, 150]. Advances in mouse strain mutations, including the development of the IL2$\gamma$ mutation, have enabled superior human hematopoietic and immune cell engraftment [44]. Multilineage differentiation of HSCs was observed in various studies, and engrafted CD4${}^{+}$T cells showed memory phenotype [16, 113, 116]. However, the engrafted human B cells were typically immature [113], and T cell, NK cell and myelomonocytic lineage also displayed some functional impairment due to blocked lymphoid differentiation [16, 116]. Moreover, due to the lack of HLA on the thymus, robust human T cell-mediated immune response was failed to establish [151, 152]. Besides, residual activity of host NK cells as well as other innate immune system functions interfered with human HSC engraftment, leading to a tendency to develop spontaneous thymic lymphoma [153].

The transplantation of human fetal liver and thymus fragments has been a promising approach to create the BLT (bone marrow, liver, thymus) model, which enables the development of human hematopoiesis and functional human immune system in mice [15]. The BLT models provided a human thymic microenvironment that allowed the sustained development of human hematopoiesis and functional human immune system, supporting the development of human T cells and their selection on human MHC molecules, and eventually mediate strong immune responses in vivo [15, 154]. However, bone marrow microenvironment required for innate immune cell development was still lack, the incidence of GVHD was higher than CD34${}^{+}$ model, and some subpopulations of abnormal recombination were found in the BLT model [16].

To address these limitations, researchers have explored alternative approaches, such as hu-SRC and hu-BLT mice, which do not reject human ESCs transplantation but suffer from inadequate reconstitution of the innate immune system and the development of a wasting disease-like syndrome that renders them inadequate for long-term transplantation [128]. Recent studies have shown that hiPSC-thymus organoid can support the de novo generation of a diverse population of functional human T cells, potentially overcoming the limitations for immune cell maturation after transplantation [128]. Additionally, in the latest study of human hepatocyte transplantation, it was revealed that xenografts recruited macrophages and neutrophils that cleared themselves. Therefore, the elimination or modulation of macrophage activities was suggested to support the long-term repopulation [38]. Future studies should focus on the mechanisms of immune recognition of xenografts, such as the SIRP$\alpha$-CD47 axis, and on the generation of immunodeficient mice genetically expressing HLA molecules.

In recent years, research on human iPSCs and transdifferentiated HSCs have attracted much attention. iPSC-HSCs were confirmed to migrate from teratomas into the mice bone marrow (BM) after transplantation. and sustain multi-lineage and long-term reconstructive ability when re-transplanted into secondary mice [18]. However, iPSC-HSCs avoid of teratoma phase have not yet differentiated successfully. Although the efficiency of direct reprogramming method of HSCs increases rapidly, it is still unable to generate functional and transplantable HSCs. Transcription factor combinations SOX4, MYB, HOXA9, ERG, and RORA have been identified to specify myeloid-restricted precursors into short-term transplantable pluripotent progenitor cells with low engraftment level and a strong tendency towards bone marrow differentiation, while absent of adult hemoglobin expression [100]. On the other hand, seven factors (ERG, HOXA5, HOXA9, HOXA10, LCOR, RUNX1 and SPI1) reportedly convert hematopoietic endothelial cells into long-term transplantable HSPCs that could develop into functional red blood cells, B cells and T cells in vivo, although the rate of red lineage denuclearization was still very limited [66]. Furthermore, BM differentiation bias remains a major obstacle to the derivation of functional HSPCs from hiPSCs. A single transcription factor MLL-AF was found to induce iPSC as a source of highly transplantable HSPC without lineage bias, allowing for the complete reconstruction of the human hematopoietic system, However, since the genome was unstable compared to the primary HSPC, the risks of leukemia transformation increased [67].

However, there are still significant challenges ahead, such as how to completely replace mouse MHCs with diverse HLA molecules and express human-derived growth factors and cytokines [92]. Rebuilding red blood cells and granulocytes, as well as prolonging the maintenance duration, are also yet to be addressed [93, 94]. Further investigation of the hematopoietic developmental signals and the identification of critical gene network regulators may help overcome these limitations.

The development of immune humanized models has significantly advanced our understanding of immune system development and related diseases [155, 156]. Combined transplantation of HSC with islet cells facilitated the modeling of autoimmune diseases type 1 diabetes (T1D) [157]; meanwhile, the human immunity provided an unprecedented platform for antiviral research, including the transmission and pathogenesis of pathogen, such as human immunodeficiency virus and human herpesvirus 4 [158, 159]. Additionally, novel tumor immunotherapy strategies are also being evaluated in immunized humanized mouse models [158, 160, 161]. These models have greatly improved our understanding of immune processes at the cellular and molecular levels, accelerating the current development of translational medicine. So far, neither BM or umbilical cord blood has yet achieved the immune functions as well as BLT in humanized mice [118], partly due to the deficiency of thymic microenvironment essential for innate immune development. iPSCs are considered as attractive and promising source to address these issues, although direct differentiation of functional HSCs was not succeeded yet [18, 129]. How to completely replace mouse MHCs with diverse HLA molecules and express human derived growth factors and cytokines remains large challenges ahead [162], while rebuilding red blood cells and granulocytes, as well as prolonging the maintenance duration are yet unsettled [163]. Further investigation of the hematopoietic developmental signals and the identification of the critical gene network regulators may help overcome these limitations.

By introducing human liver genes into the mouse genome, researchers have generated a series of liver gene-humanized mice, including mice expressing human exogenous receptors, drug-metabolizing enzymes and transporters, as well as combinatorial models [164]. However, transgenic mice do not possess a complete “humanized metabolic network” within the liver and was not sufficient for studying human liver metabolism. Animal liver with human hepatocyte replacement was considered much similar to human liver with respects to liver zonation, and phase I, II and III enzymes expression, thus particularly suited for studying drug metabolism and pharmacokinetics, drug-drug interactions and hepatotoxicity [165]. The chimeric liver is a mixture of endogenous mouse hepatocytes and human hepatocytes, and the level of human liver characteristics were greatly depended on the human hepatocyte repopulation ratio [79]. Currently, PHHs are the most widely used source for liver humanization. In FRG mice model, PHH could stably achieve an impressive 80% repopulation rate [37], ALB TRECK/SCID mice also allowed efficient repopulation of human primary fetal liver cells and hepatic stem cells, with a near 100% in some cases. While adult human hepatocytes failed to survive in this mouse model [16]. Although humanized liver models have been successfully generated in mice, practical applications, particularly in the pharmaceutical industry, require larger animals such as rats. Very recently, our team made a remarkable breakthrough in generating humanized liver in a rat model using similar gene knockout strategy as for FRG mice. Unlike the robust repopulation in mice (usually took 3–4 month to achieve over 70%), liver humanization in rat took around 7 months to reach a maximum level of around 30%, suggesting that generating humanized liver in larger animals other than mice is challenging. Since most applications, particularly drug testing requires humanization level over 30% in animal liver, transplantation of much larger amount of human hepatocyte was required to achieve efficient liver humanization [37]. While, PHHs have been shown to be the best source for in vivo repopulation and functionality, their limited availability, genetic and phenotypic diversity, and rapid differentiation in culture greatly limit their applications. The discovery of PSCs and their potential to differentiate into hepatocytes has drawn much attention as an alternative. Recent research has focused on increasing the repopulation efficiency and enhancing the hepatic functions to be closer as PHHs. Co-culture of mesenchymal and endothelial cells has been shown to significantly promote differentiation and engraftment of hiPSC-derived endodermal progenitor cells in vivo, achieving 70% repopulation rate in IL2rg${}^{\mathrm{-}\mathit{}\mathrm{/}\mathrm{-}}$ rats with damaged liver, although the maintenance period was limited to 22 days [166]. Notably, short-term repopulation of hiPSC-HLCs have been achieved in nonhuman primate liver fibrosis model [91], implying the possibility to generate humanized animals in big animals. In addition to repopulation efficiency and functionality [167], yet some critical factors such as intestinal obstruction, portal vein thrombosis, and intestinal ischemia have been found to be potential causes of death due to poor blood flow of the grafts [168]. With the development of iPSC-based organoid technologies, hiPSC-derived three-dimension hepatic grafts with vascularized structures and enhanced functionality have been established. Unlike single cell transplantation, they showed no translocation to the other organs, no sign of massive necrosis or portal vein thrombus [166]. Thus, organoid transplantation appealed to be a novel way to achieve more effective and safer humanization. Liver humanized animals are valuable models for studying viral infections such as Hepatitis B Virus (HBV) which cannot naturally infect most animals, including mice. These animal models allow for the study of viral infections, host-virus interactions, and testing of new antiviral drugs. Currently, Alb-uPA/SCID, FRG and TK-NOG mice are extensively reported for HBV infection modeling potentials, nevertheless, absence of human immunity makes these humanized mice unsuitable for studying virus-induced immune responses [169]. Integration of humanized immunity may well lead to more accurate modeling of viral infections. Additionally, these models have been widely used in drug researches due to their superior advantages overwhelming current preclinical models [170] (Fig. 1), since they were capable of studying human-specific metabolites that do not exist in other species and revealing human-specific drug responses. Particularly, iPSC-derived humanized liver is believed to better mimic the patient specific liver metabolism and drug response, may serve as ideal platform for personalized disease modeling and drug testing. However, it is still difficult to control the degree of chimerism, and the remaining mouse hepatocytes may recover unexpectedly, or have higher affinity for certain drugs than their human counterparts [171]. Meanwhile, human specific extrahepatic drug metabolism and interactions cannot be identified in these humanized mice due to lack of humanization in other metabolic organs like intestine [172], more comprehensive and systematic humanization was required to achieve the human metabolism modeling.

Integration of human immunity may potentially lead to more accurate modeling of viral infections. Moreover, these models are also beneficial in drug research as they can study human-specific metabolites that are not present in other species, providing insights into human-specific drug responses. Particularly, a promising technique for personalized disease modeling and drug testing is the use of iPSC-derived humanized liver models. These models are thought to better replicate patient-specific liver metabolism and drug response. However, controlling the degree of humanization is challenging. Additionally, remaining mouse hepatocytes in the model may unexpectedly recover or have a higher affinity for certain drugs than their human counterparts [102]. Moreover, without humanization in other metabolic organs like the intestine, it is impossible to identify human-specific extrahepatic drug metabolism and interactions in these humanized mice [103]. Therefore, a more comprehensive and systematic approach to humanization is necessary to achieve better human metabolism modeling.

Mice with humanized skin are commonly used to study transplantation rejection and infectious diseases, by surgically grafting human skin onto mice, researchers can visualize and assess skin diseases. Currently these mice have been widely used to investigate autoimmune and inflammatory skin diseases such as pemphigus vulgaris, GVHD and genetic skin diseases. The human atopic dermatitis (AD) mouse model and humanized alopecia areata (AA) model are both induced lesions in mice transplanted with human skin. However, due to the differences in skin structure and immune function between mice and humans, and the different pathogenesis of various AD subtypes, this model can only simulate acute AD not other complex AD dermatitis [130, 131, 132, 173]. A systemic lupus erythematosus model was successfully generated after transplanting human PBMC from active patients into the dorsal intradermal region of BALB/c nude mice, followed by daily ultraviolet radiation B, resulting in persistent and stable patchy erythema with scaling in the implanted area [174]. However, the expression patterns of adhesion molecules in grafted skin differed from those in original grafts, which significantly affected the homing and recirculation of human lymphocytes in the graft, directly affecting the observations of immunoreaction [175]. Therefore, the construction of humanized animals for skin and immunity is highly in line with the research needs.

Human iPSC-derived skin cells are also increasingly used in the construction of humanized animals. Melanocyte derived from hiPSCs permitted to induce melanosis in nude mice skin where melanocytes aggregated [133]. The hiPSC-derived dermal papilla could partially mimic the role in hair follicle development, and when combined with keratinocytes, it could simulate the three-dimensional assembly of nature hair follicle structure in vitro [176]. An in vitro 3D iPSC-skin organ system with mature hair follicle and sebaceous gland was transplanted into the wound of mouse skin, and it displayed the most skin functions [177]. However, the construction of complete skin organs from iPSCs remains a major obstacle. So far, only melanocyte, keratinocyte, and 3D skin organ system have been successfully induced from iPSC, while it is currently unavailable to simulate complete skin functions. Additionally, different parts of the skin have different thicknesses, with or without hairs, and accessory organs such as sweat glands are different, so establishing partial and localized skin humanization may not thoroughly reflect all aspects of human skin.

In a study, human PSC-derived pancreatic endoderm cells developed into islet-like structures that produced human insulin and maintained normal glycemia when transplanted. Immunotherapies targeting T cell costimulatory pathways attenuated the rejection of xenograft, suggesting that PSC-derived islets may serve as a potential substitute for T1D with proper immunological interventions [122].

To facilitate diabetes drug screening and disease modeling, researchers have been developing functional pancreatic $\beta$ cells from adult stem cells, ESCs and iPSCs. In a study, human PSCs-derived pancreatic endoderm developed into islet-like structures that produced human insulin and maintained normal glycemia when transplanted. implying the potential that under proper immunological interventions PSC-derived islets may serve as potential substitute for T1D [134].

To determine the in vivo differentiation potential and functionality, $\beta$ cells derived from PSCs were transplanted in situ into the pancreas of newborn mice, which survived and expressed $\beta$ cell transcriptional factors and markers associated with functional maturation [178]. Additionally, hiPSC-derived mature human islet-like organoids containing endocrine-like cell types exerted hypoglycemic function and rapidly re-established glucose homeostasis in diabetic NOD/SCID mice and non-human primates with diabetes [179, 180]. These functional and transplantable sources make the fundamental basis for generating humanized islets.

In clinical experiment, T1D patients indeed restore glucose homeostasis following islet transplantation, but most patients lose islet function due to allograft rejection Human islet xenografts were found to have a 60% occurrence of rejection, which affected the long-term observation of graft outcome [135]. The construction of human islets with human immune system in mice is an ideal model to study the mechanism of xenograft rejection and to develop therapeutic or preventive strategy. NSG mouse model reconstituted by CD34${}^{+}$stem cell and regulatory T cell was utilized as an excellent platform for assessing the natural immune response to islet allografts in humans [181]. Although the technology for constructing humanized pancreatic islets seems relatively mature, immune rejection remains the main obstacle to the long-term existence and functions of humanized pancreatic islets. The breakthrough of pancreatic humanized mice majorly depended on progress in establishing immune tolerance or immune-humanization.

Although the technology for constructing humanized pancreatic islets seems relatively mature, immune rejection remains the main obstacle to the long-term existence and functions of humanized pancreatic islets. The breakthrough in pancreatic humanized mice mainly depends on progress in establishing immune tolerance or immune-humanization.

The newly developed tools such as brain organoids, while a promising tool for studying the human brain, lack the full complexity of in vivo systems, including microenvironments, neuronal circuits, vascular circulation, and immune systems. As a result, they cannot fully capture the complexity of brain development, function, and dysfunction, which limits their use in precise disease modeling or drug screening applications [182, 183]. Humanized animal of brain may better assist in preclinical drug testing of brain-like organs from patient under physiological conditions.

While humanized animal models of the brain have successfully shown the integration of human cells and functional neural networks, questions remain about their ability to accurately simulate aging and dementia in the brain. Nonetheless, the integration of human brain organoids in immunodeficient animals, such as intracerebral integration into NOD-SCID mice, has been achieved, where neurons gradually mature, synapse occur, and functional neural networks form. Additionally, transplantation of induced pluripotent stem cell-derived dopaminergic progenitor cells in a monkey model of Parkinson’s disease expressed markers of dopaminergic neurons in the midbrain, and human iPSC-derived neural precursor cells implanted into the cortex of newborn mice showed potential. Encouragingly, engrafted human microglia into the central nervous system through in vivo differentiation in mice brain responded dynamically to local and peripheral injury.

Since primary brain cells are not readily available, currently induction from hPSCs is the main strategy to obtain neurons, astrocytes and oligodendrocytes for neurodegenerative and myelin diseases studies. In a crab-eating monkey model with Parkinson’s disease (PD), induced by the neurotoxin MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine), hiPSC-derived dopaminergic progenitor cells expressed markers of dopaminergic neuron in the midbrain after transplanted in vivo [184]. Moreover, hiPSC-derived neural precursor cells could implant into the cortex of newborn mice [136]. Functional and healthy midbrain dopamine neurons cells from iPSC can effectively restore motor function and re-operate the host brain after transplanting into the rodent model of PD [19]. Followed immune humanization by iPSC-derived hematopoietic progenitor cell transplantation, human microglia could be engrafted into central nervous system though in vivo differentiation in mice brain, and responded dynamically to local and peripheral injury [137]. Encouragingly, up to 80% microglial chimeras were achieved under optimized differentiation and xenotransplantation methods [185]. Nonetheless, the integration of human brain organoids in immunodeficient animals, such as intracerebral integration has been achieved in NOD-SCID mice, where neurons gradually mature, synapse occur, and functional neural networks formed [138]. However, neurons derived from PSCs have significant differences from primary cells, despite sharing the same genotypes and exhibiting similar electrophysiological properties [186]. Neurons derived from fetal iPSC could not simulate the neurodegenerative diseases in the elderly, while the majority of iPSC-neurons generated by direct differentiation were prone to be fetal or neonatal cell-like [187]. These differences arise from variations in epigenetic profiles, which result in significant differences in gene expression. The modification of the epigenetic characteristics of reprogrammed cells is an important obstacle to realizing human neurodegeneration, not to mention the importance of disease epigenetics for disease modeling [188, 189]. It is still doubtful whether iPSC-derived neurons and other central nervous cells can accurately simulate aging and dementia in the brain. Major concerns raised how these differences would affect when hiPSC-derived neurons are used as predictors of human neural responses upon chemical or pathological challenges [190].

While mixed cortical neurons from iPSC have been successfully established, the identification of specific cortical subtypes remains a challenging task [191]. The generation of these subtypes from hiPSCs with correct transcriptome and epigenomic characteristics is crucial to improving the validity and accuracy of disease modeling. However, nervous system is a complex and interconnected integrator, in this regard, animals brain replaced with human brain organoid provide the platform for evaluating the differentiation, axonal growth, and synaptic integration of neural organs from patient specific iPSCs under physiological conditions in vivo, promoting research on the pathogenesis of neurodevelopment, neuropsychology, and neurodegenerative diseases.

Human lung, intestinal and cancer cells have also been successfully transplanted into mice to generate humanized models. Wahl et al. [192] transplanted up to 40 cell types (including non-hematopoietic cells) from human lung tissue into immunocompromised mice to form highly vascularized lung grafts. In this BLT-based humanized mice, the lung grafts and autologous human hematopoietic cells mediated robust humoral and T-cell responses, making it suitable model for studying pathogenic infections in vivo [192]. Human colonic and intestinal epithelial organoids have also been used to reconstruct intestinal epithelial cells in mice, providing a technical basis for organoid transplantation in case of colonic injury [193]. Additionally, transplantation of patient-derived tumor tissue or cells has been widely used to develop individualized xenograft models for studying cancer biology, metastasis, and new therapies.

With the rapid development of transplantation sources and humanization strategies, gene editing techniques have further advanced the creation of humanized models, overcoming limitations inherent in conventional models. For instance, transgenic mice expressing human MHC class I and HLA II were reportedly compatible with human T cells engraftment [194, 195], while mice expressing human p53, a tumor suppressor gene, are ideal for cancer cell engraftment [196]. In mice expressing urokinase fibrinogen activator, which inhibits mouse hepatocyte function, drug metabolism occurs exclusively through engrafted human hepatocytes, which may facilitate the monitoring of human-specific drug pharmacokinetics [197].