key: cord-1052485-4rjr7dha
authors: Laudanski, Krzysztof
title: Humanized Mice as a Tool to Study Sepsis—More Than Meets the Eye
date: 2021-02-27
journal: Int J Mol Sci
DOI: 10.3390/ijms22052403
sha: 50cab32774e3522706f98fe901e92ce70680557c
doc_id: 1052485
cord_uid: 4rjr7dha

(1) Background. Repetitive animal studies that have disappointed upon translation into clinical therapies have led to an increased appreciation of humanized mice as a remedy to the shortcomings of rodent-based models. However, their limitations have to be understood in depth. (2) Methods. This is a narrative, comprehensive review of humanized mice and sepsis literature to understand the model’s benefits and shortcomings. (3) Results: Studies involving humanized models of sepsis include bacterial, viral, and protozoan etiology. Humanized mice provided several unique insights into the etiology and natural history of sepsis and are particularly useful in studying Ebola, and certain viral and protozoan infections. However, studies are relatively sparse and based on several different models of sepsis and humanized animals. (4) Conclusions. The utilization of humanized mice as a model for sepsis presents complex limitations that, once surpassed, hold some potential for the advancement of sepsis etiology and treatment.

Sepsis, septic shock, septicemia, or the more colloquial blood poisoning, is a prevalent, mortal, and formidable challenge in medicine [1] . In a recent consensus statement, sepsis was defined as a "life-threatening organ dysfunction caused by the dysregulated host response to infection," but even this definition is challenged [2, 3] . So far, the definition of sepsis predominantly relies on physicians' judgment compounded with vague and ambiguous symptomology. The pathological processes associated with sepsis are most likely very heterogeneous, however this is rarely appreciated in clinical settings [4] . Sepsis is often linked to bacterial triggers, however any pathogen can elicit the immune system pathological response [5] . Sepsis has a tremendous impact in developed, developing, and austere countries. In the United States alone, the incidence of sepsis encompasses 6% of all hospital admissions. The mortality varies between 20% and 40% [1, 4, 6] . The costs related to sepsis treatment and subsequent disability of survivors is 20 billion dollars per annum in the United States alone, with profound and difficult-to-quantify individual and societal consequences [6] [7] [8] . Only early engagement of antibiotic was demonstrated to improve sepsis outcome [4, 6] . Significant efforts were directed towards finding effective remedies for sepsis, but, almost universally, they failed in terms of the translation of discoveries in animal models to clinical practice [9, 10] .

Sepsis is initiated by binding of the pathogen-associated molecular pattern (PAMP) to toll-like receptors (TLR) followed by the activation of innate immunity [4, 6] . This early response activates the endothelium, complements coagulation, lactoferrin, transferrin, NK cells, phagocytes, and neutrophils, as well as a variety of other mechanisms. Concomitantly, neuronal inflammatory reflexes augment immune system responses [11] . Cytokine release induces several symptoms including fever (IL-6, prostaglandins), switch to catabolism (TNFα, IL-6), tachycardia and tachypnea. Vasodilation and increased vascular permeability leads to tissue swelling, hypoxia, and intravascular volume depletion [12] . Ensuing hypotension is compounded by septic cardiomyopathy [13] . Oxygen delivery to the organs may be further impaired by microcirculation defects, increased endothelial permeability, and mitochondrial abnormalities [6, 12, [14] [15] [16] . The collateral damage due to innate immunity activation is mediated by the nonspecific nature of endogenous bacteriocidic agents, innate effects of IL-1β, TNFα, granzymes, or prostaglandin metabolism products. Activation of the coagulation cascade may result in disseminated intravascular coagulation [17] . As the innate system response peaks, monocytes, dendritic cells, and, to some degree B cells, capture and process antigens [4] . Subsequently, specific T and B clones undergo a massive buildup, directing the immune response precisely against the invading pathogen and limiting collateral damage. Upon eradication of the pathogen, a de-activation of the immune system has to occur. This part of the process is called compensatory antiinflammatory response syndrome (CARS), but the name implies that the immune system is actively suppressed [4, 18] . In fact, during CARS the immune system is being de-activated, not suppressed [4] .

This intricate process can be derailed in several ways, leading to different sepsis presentations [6, 9, 18] . More importantly, the complex nature of sepsis is species-specific, putting the utility of the rodent models of sepsis to question [19] . Here, we provide a narrative review of research involving humanized mice with special emphasis on their benefits and shortcomings.

There are three main approaches to model sepsis on rodents ( Figure 1 ). The most straightforward approach to mimic sepsis is the toxemia model. Lipopolysaccharide (LPS), or another PAMP, is injected into animals to trigger an immune response [20] [21] [22] [23] [24] . This allows for high standardization and reproducibility, but the model is so distant from clinical relevance that it should not be referred to as a septic model. However, it is useful to study essential immune responses allowing for precise manipulation of the kind and dose of pathogen used. Type of infected compartment of infection and host susceptibility are essential factors determining animal responses in the toxemia model [25, 26] .

Host disruption models of sepsis approximate clinical sepsis with greater accuracy. They include cecal ligation and puncture (CLP), cecal ligation and incision (CLI), and colon ascendants stent peritonitis (CASP) [27] [28] [29] . All these models disrupt the gut barrier, allowing colon flora to induce sepsis-like immune responses [30] . Their advantages and shortcomings are summarized in Figure 1 .

The CLP model including humanized mice has been utilized frequently [31] [32] [33] . One study compared both LPS-endotoxemia and the CLP model [33] . The bone marrow progenitor changes were distinct in these two models, suggesting that CLP and endotoxemia models are not equivalent. Besides problems with model equivalency and variable mortality, CLP and CLI models are susceptible to surgical conditions, making it challenging to compare studies published in different labs. The fecal slurry model is the only model where there is no disruption of the gut barrier [34] . The fecal matter is given directly to the peritoneum. Variability in this model stems from different qualitative and quantitative differences in fecal matter. Advantages are the lack of the surgical injury to animal. The authors did not find a single case of use of this model in humanized animals. 

Several reasons underlying unsuccessful translation of the animal studies in therapies were articulated [27, 35, 36] . Mice and rats have a significantly higher resistance to sepsis as compared to humans. Certain animals are inbred and usually at the same age, resulting in uniform response with loss of sepsis presentation heterogeneity. The interactions between key internal organs (neuronal, immune systems, others) may be different in rodents, further limiting their equivalency [11] . Pre-existing co-morbidities are infrequently present in animal models before inducing sepsis, but they are important factors in human sepsis [35, 37] . After the induction of sepsis, a minority of researchers treat animals in a way resembling the clinical treatment provided to septic patients, because most 

Several reasons underlying unsuccessful translation of the animal studies in therapies were articulated [27, 35, 36] . Mice and rats have a significantly higher resistance to sepsis as compared to humans. Certain animals are inbred and usually at the same age, resulting in uniform response with loss of sepsis presentation heterogeneity. The interactions between key internal organs (neuronal, immune systems, others) may be different in rodents, further limiting their equivalency [11] . Pre-existing co-morbidities are infrequently present in animal models before inducing sepsis, but they are important factors in human sepsis [35, 37] . After the induction of sepsis, a minority of researchers treat animals in a way resembling the clinical treatment provided to septic patients, because most studies are short term [10, 28, 30, 38] . Additionally, several manifestations of organ failures have no clear equivalences in rodents and humans (e.g., mental status). Neither are immune system components identical in both species, with the spleen being a prime example of an organ with a distinct role in the human vs mice immune system [39, 40] . Several of these concerns were raised in "Of mice, not a man", a landmark opinion piece upsetting long established beliefs, paradigms and culture in sepsis research [35] . Subsequent discussion about a relevance of animal models to human sepsis remains rather animated, but it is difficult to deny that mice immune systems have several distinctive features [1, 4, 35, 36, 38] .

Humanized mice are xenotransplant animals with a human immune system developed through the expansion of grafted human cells [41] . Subsequent models of immune system reconstitution approximate the human immune system with increased accuracy while depressing the native immune system more completely.

The creation of the humanized mice starts from a mouse strain with acquired immunodeficiency [41] . The most common type used is SCID [41] [42] [43] [44] [45] . Other strains frequently used are IL-2 rgnull , NOD.Cg-Prkdc scid Il2 rgtm1Wjl (NSG), NODShi.Cg-Prkdc scid Il2rg tm1Sug (NOG), and C;129S4-Rag2 tm1Flv Il2rg tm1Flv (BRG) mice [46] . All these strains suffer significant immune-system deficits, allowing for more efficient grafting and reducing the emergence of graft vs host disease [45, 47, 48] . However, animals have to be irradiated before grafting to suppress the native immune system.

Once the mice are deprived of their native immune system, the implantation of the human immune system can occur. Depending on the type of transplanted tissue, three major types of humanized mice exist (Table 1) . Finally, humanized animals may be genetically modified to provide them with the endogenous ability to produce human cytokines (GM-CSF, Flt-L) to support the growth and survival of the implanted human immune system [49] .

The septic shock and humanized mice narrative began from a groundbreaking study by Hotchkiss et al., utilizing CLP to induce sepsis and euthanizing animals after 24 h [50] . They demonstrated that the production of cytokines in response to infectious insult was significantly increased in animals 20 weeks after grafting. Most importantly, humanized mice had a profound apoptotic response, a commonplace phenomenon in the aftermath of sepsis [4] . Additionally, delayed hypersensitivity skin reaction was diminished. However, animals were euthanized 24 h after sepsis induction, limiting the study to the acute response phase. Mortality was unknown among the cohort of CLP induced animals. Animals did not receive fluid resuscitation or antibiotics. Despite Hotchkiss's pioneering research, it is still surprising that it took over 15 years for the first study of septic shock in humanized mice to appear since the first introduction of this model in 1992 [26] . Further studies confirmed the secretion of TNFα, IL-6, and IL-10 in response to TLR-stimulation by PAMP in humanized mice [21, 51] . Concomitantly, the release of high mobility group box 1 (HMGB1) in sepsis was demonstrated and established a viable therapeutic option six years before the first similar human study [52, 53] . Inflammasome complexes (NFL3) were demonstrated in humanized mice, but in a model of autoimmune arthritis [54] .

Several other bacterial diseases were demonstrated in humanized mice. Typhoid infection in the SCID humanized mice replicated the entire disease process [55] . Skin infection with Staphylococcus in humanized mice showed a clinical resemblance, however animals required a higher dose of inoculated bacteria while subsequent lesions were more pronounced than wild-type animals [56] . Humanized mice were instrumental in discovering a novel way of interaction of Staphylococcus Enterotoxin B (SEB) with iNKT or MAIT cells [20, 22] . The same group translated these results to mouse models, establishing several important links [23, 57] . Knop et al. utilized humanized mice to simulate Gram-positive bacterial infection instead of SEB toxemia. They demonstrated several common responses typical of sepsis as in classic study by Hotchkiss's group [50, 58] . Most importantly, they compared the responses from humanized mice to wild type. Humanized mice had higher weight loss over five days after CLP induction, validating their translational relevance, because cachexia is sign of sepsis. Above all, they showed emergence of T cell anergy and Fas expression as compared to the wild type. These are human-specific responses to the septic challenge [4] . In a separate study, BLT-type humanized mice were used as a model for sepsis with a 28 day euthanasia window. Mice were treated with antibiotics and fluid resuscitation, however the mortality was a significant problem and required modification of the protocol [31, 32] . The study demonstrated a change in human monocyte and dendritic cells' reactiveness in animals surviving sepsis after 28 days from the original insult. M-CSF production secondary to PU.1 epigenetic dysregulation was one of the main mechanisms driving the abnormal fate of MO [31] . This finding is well-aligned with the belief that epigenetic changes sustained sepsis-induced changes long after the septic episode was resolved [59] . Transplant of stem cells partially restored the pre-septic function of the immune system, but the underlying mechanism was somewhat unclear. Bone marrow dysfunction was frequently suggested as a primary driver in immune suppression in sepsis [4, 60] . Skirecki et al. showed early impairment of bone marrow progenitors in a humanized mouse model of CLP sepsis [33] . However, the relatively short time between re-grafting and sacrificing animals could preclude full restoration of the immune system.

Humanized mice were utilized to model neonatal sepsis in two studies from the same research group to demonstrate potential effect of prostaglandin and steroid modulators [25, 26] . Pervasively, authors argued that imperfection of the human immune system in humanized mice equates to the "immature" immune system seen in neonates. However, neonates' immune systems are relatively well developed, and neonatal sepsis has several distinctive features [61] . Endogenous immune system dysfunctions in humanized mice result from imperfection in several immunological compartments, co-existence with the mouse immune system, and scarcity of human growth factors, not neonatal-like immaturity [62, 63] . It is far-reaching to equate the imperfection of the immune system performance related to the xenotransplant nature of the model as an indication that this resembles neonatal immune system features.

The emergence of anergy or increased prevalence of monocyte suppressive cells were observed during septic shock in humanized mice and wild animals [20, 57] . Both mechanisms are part of de-escalation mechanisms, and their emergence is a complex process during the evolution of the response to septic shock [4] . Emergence of unfavorable T cell anergy was signified by increasing the expression of TIM-3, LAG-3, PD-1L, and PD-1 in humanized mice [64] .

Modeling HIV infection is one of the success stories for humanized mice, which has provided a sustainable model to study infection and a platform for testing several HIV/AIDS-related phenomenon [65] [66] [67] [68] [69] . Zika infection and related viremia have also been well studied in humanized models [70] . Dengue infection is almost universally complicated by hemorrhagic fever and death, and humanized mice are a perfect model to study this, considering the high mortality of these diseases [71, 72] . The pivotal role of NK and IFNγ in protecting against acute viremia in hemorrhagic fever was discovered using this model [73] .

Dobner et al. showed that NSG humanized mice could reproduce acute and chronic adenoviral infection [74] . Most importantly, the persistence of chronic infection was signified by established immunity and expression of viral genes within bone marrow only. These findings aligned exceptionally well with clinical findings. The authors did not test the animals with sepsis models afterward to establish if the viral infection would re-activate, because this has been observed in some survivors of sepsis.

Similarly, Epstein-Barr virus (EBV) infection was successfully established in humanized mice [75, 76] . In a more recent study, the authors suggested that EBV infection leads to hemophagocytic lymphohistiocytosis (HLH), but they failed to provide clinical data supporting such an assumption. HLH is increasingly appreciated as the illness producing a sepsis-like syndrome, and introduction of a humanized model could be instrumental in establishing effective therapies [77] .

There is no report to date investigating the application of humanized mice to study SARS-CoV-2, except one review [5, 78] .

The immune system response to protozoa is often overlooked, even though malaria is one of the world's most prevalent conditions. Humanized mice can sustain the whole life cycle of Plasmodium falciparum if both human liver tissue and erythroid compartment have been provided [79] [80] [81] [82] [83] . Immune responses to malaria are different from sepsis in several aspects, but the dysregulated immunological response is commonplace among these two conditions [83] . Humanized mice allowed for the discovery of the exosomes in malaria, delivering proof of the principle regarding new diagnostic techniques [84] . New therapeutic compounds and vaccine testing were conducted using humanized mice as well [85, 86] .

The impetus for the development of humanized mice was to understand sepsis pathophysiology, to accelerate the development of new therapies, to minimize exposure of human subjects to potentially ineffective or even harmful therapies, and to decrease commercialization costs for promising therapies [4, 19, 25, 35, 44, 47] . The overall effort to produce, breed, and maintain these hybrid animals has been lower than conducting research with human subjects, but substantially higher than using a rodent sepsis model. The question arises if the animal models are worth the additional cost and effort.

Our review of the literature showed that humanized mice contributed to understanding sepsis. The highlights of utilizing humanize mice models are concentrated: on describing the mechanism of apoptosis, interactions of staphylococcal superantigens with the immune system, development of vasculitis in hemorrhagic fever and meningitis, and short-and long-term reprogramming of the leukocyte via epigenetic mechanisms [32, 57, 58, 70, 74, 79, 85, 87] . Humanized mice enabled demonstration of the causative effect and producing experiments that could be replicated using virtually identical animals, re-created from the same stem cell batch.

On a few occasions, humanized animals were employed to test certain compounds' effectiveness to modify sepsis's natural history. Ernst et al. published two papers showing the effect of betamethasone and indomethacin in neonatal sepsis modeled by "immaturity" of the humanized mice immunity [25, 26] . A third paper complemented these studies, showing a similar effect when a pathogen from Gram-positive bacteria stimulated neonatal cord mononuclear leukocytes [25, 88] . They observed several parallel responses in both experiments, suggesting that betamethasone may be sufficient to moderate cytokine release in neonates compared to betamethasone with indomethacin [88] . The most significant contribution of the studies conducted by Ernst et al. is that they could prevent unnecessary testing on human subjects. Ye et al. tested HMGB1 siRNA using a short AchR binding peptide to mitigate cytokine storm in the CLP model of sepsis induced in humanized mice (BLT model) [52, 89] . Humanized mice were used as a testbed for modulating the expression of PD-1, one of the most promising therapeutic agents, verifying several observations from human and animal models [64, 87] . The C5aR antagonist was tested in a humanized model of Staphylococcal skin infection [56] . The drug reduced inflammation but resulted in immunosuppression that had an overall detrimental effect on survival. Utilization of the humanized models in chronic viral infection and malaria was credited with reduced harm, accelerated testing, and reduced costs [86, 90] . Finally, autogeneic transplant of stem cells to the humanized BLT mice previously exposed to CLP sepsis seemed to correct peripheral and bone marrow immunological level properties, as was theorized previously [31, 33, 60] . Of note, transfer of mesenchymal cell was shown to reverse at least some detrimental effects of sepsis [91] . Consequently, testing the drug in humanized mice eliminated the risk in human subjects.

Some limitations of humanized mice in the study of sepsis are related to animals' use in general. There are some excellent reviews comparing human sepsis to murine models [35, 36] . Here, we focus on limitations specific to humanized mice.

One of the primary deficiencies of humanized mice is a dynamic process of reconstitution of the human immune system in terms of the number of competent cells and composition of the immune system. Absence of one leukocyte population may hamper immune responses in general [92] . At the same time, the emerging human immune system must acquire a tolerance to the residual host's immune system, dampening the graft versus the host disease, thus allowing native immune system to survive. Concomitantly, the transplanted immune system is immersed in mouse-originated growth [40] . Consequently, the reconstitution of human immune systems in xenotransplants is not absolute, but rather an ever-changing work in progress [42, 62, 63, 92] . The development of more sophisticated humanized mice models (HIS, BLT) significantly reduced this problem, but did not eliminate it [38, 41, 47, 66, 70] .

The second confounder of humanized mice is the fact that two immune systems co-exist. The positivity for CD45, a marker of successful grafting, varies between 20% and 80% [44, 47, 71] . Conversely, between 20% to 80% of the cells are mouse in origin. It is unclear what the remaining cells are functionally, or how they affect overall performance septic immune response. The level of grafting is very infrequently reported by researchers.

Upon successful grafting, two processes emerge simultaneously: (1) reconstitution of the components of the human immune system; and (2) resurgence of the native mouse immune system. Animals who were grafted with human stem cells needed time to reconstitute their human immune system, but at the same time, the strongly suppressed mouse immune system will attempt to reconstitute itself [21, 42] . As the implanted stem cells age, the human immune system matures with different leukocyte compartments emerging at different times, but the risk of the resurgence of native immune systems is also higher in older animals. In more sophisticated models of humanized mice, an emergence of lymphoma due to the dysfunction of the native immune system was cited, but the introduction of double ablation methods alleviated this problem. Then, another study emerged suggesting higher mortality in humanized mice secondary to the inhibitory effect of mouse cells on human grafted leukocytes [51] . These interactions are difficult to measure, and account for, during cross-study comparisons. Both immune systems interact with each other, however there is almost no literature deep-diving into the importance of the process, especially in the context of time.

The innate nature of the laboratory animals does not mimic the human population's heterogeneity, or the clinical reality experienced by patients. Humanized mice are even more inbred because their hosts are highly preselected. It is somewhat unclear if implantation of a series of homogenized animals with stem cells obtained from the same donor should be counted as separate animals or "clones". This may present the future researcher with a thought-provoking dilemma. Should they use multiple stem cells to create several progenies? Should they treat animals grafted with the same stem cells longitudinally? Furthermore, there is randomness in how stem cells settle and develop in various leukocyte lineages, even in inbred animals [21, 33, 50, 54, 56, 61] . However, such heterogeneity still does not reach the level of human heterogeneity.

Not a single manuscript involving humanized animals with pre-existing co-morbidity of some kind was identified by the author. Animals are kept in artificial, sterile, highly controlled, socially isolating conditions. Except for one study, no therapeutic intervention typically applied during sepsis was employed to mimic current states of disease.

Another question concerns the age of animals, especially if research focuses on age extremes. However, the question arises as to how we should account for the humanized mouse's age. Is their age at the time of transplant the actual age of the mice or the stem cells' age? In the study by Lange, animals used were of two "age" groups. Animals were subjected to resting 8-13 weeks versus 15-22 weeks after transplantation [21] . He suggested that cells after transplantation undergo maturation reflecting the aging of the immune system. Lange et al. arrived at a conclusion that seemed to juxtapose to findings presented by Ernst et al. [21, 25, 26] . The latter suggested that deficiencies of the humanized mice immune systems reflect neonatal immature immune systems [25, 88] . Neither group of researchers provided sufficient evidence of the accurate and verifiable age of the humanized immune system. Furthermore, only 20% of cells were positive for the human CD45 antigen, suggesting that grafting was not completed, and some native immune systems survived. A more reasonable approach was shown by Andre, who grafted one group of mouse neonatal umbilical cord-blood derived cells, while another group was created used adult peripheral stem cells [62] . This demonstrated the emergence of impaired CD80 + CD86 + cells and the reduced T cell ability to stimulate after the induction of sepsis in aged animals. The question of age is important because it influences several factors in the immune system response to pathogens, particularly the production of IL-6, a critical molecule in septic response [21] . These data suggests that the age of the grafted cells is a dominant factor in the aging of humanized mice, but it is unknown whether humanized leukocytes aged at the same speed as their "regular" counterparts.

Mouse host repopulation of the immune system typically occurs in the presence of human growth cytokines and hormones, but the bone marrow support system is of mouse origin [40, 48, 91, 93] . Furthermore, the mesenchymal system is damaged by radiation used to suppress the mouse immune system before grafting. This creates a substandard cytokine environment for human stem cells due to the production of mice cytokines, a process which is further weakened by radiation injury. Newer models of humanized mice partially mitigate this problem by attempting complete reconstitution of the supportive environment in bone marrow, but entirely successful reconstitutions are not possible at this time, even with modern techniques supplementing human growth factors and cytokines [19, 38] . There are several techniques to provide human cytokines to boost growth of the transplanted human immune system. Some of them rely on grafting additional human tissue such as the liver to study malaria, or blood vessels to investigate sepsis-related vasculitis [71, 82, 84, [94] [95] [96] . Some studies supplemented human growth factors exogenously with the recovery of individual leukocyte compartments. Another approach is to genetically modify animals to provide more human cytokines and growth factors [97, 98] . However, the introduction of several genes may skew the leukocyte population and potentially render the model questionable [97] . It is uncertain how these modifications affect the quality of the immune system and influence the results of the studies.

Crosstalk between different organs during sepsis is a critical element modulating the natural history of sepsis. The immunological nervous reflex is a significant contributor to the de-activation and moderation of the immune system [11] . Subcortical structures of the central nervous system seemed to be involved in several immunological functions and sepsis [4, 10, 11] . However, it is unclear how much impairment is present in the humanized mice due to their xenotransplant nature, considering that mice nervous systems will interact with human cells. Additionally, specific organs have a slightly different function between humans and mice. The spleen is an immunologically significant organ in rodents, but its function is somewhat less important in humans. Nevertheless, several researchers report the spleen-based leukocyte alterations in post-septic humanized mice, while examining lymph nodes could be more reflective of changes in human [92] . Furthermore, humanized mice may exhibit impaired interactions between the immune system and organs because several human factors (cytokines) interact with mouse hosts (hormones, organs), even though some other mediators are potentially identical (prostaglandins) [92] .

The final limitations to using humanized mice in models of sepsis are the cost and availability of mice. Humanized mice are expensive to procure and to maintain. Recovery of the immune system after grafting the animals with stem cells takes several weeks. During that time, animals must be housed in a sterile facility due to their impaired immunity. Additional expenses at this stage are the grafting and cytokine boosts. Animal attrition may result in significant expenses. Therefore, an essential question concerning researchers is whether all these expenses justify a better approximation of human sepsis. A subsequent concern is the availability of mice; they must be procured locally or obtained from a commercial vendor. Obtaining animals from a local facility has several advantages if there is evidence of provider expertise in manufacturing these animals. Considering significant labor requirements and the length (~3 months) of the process necessary to generate humanized mice, assistance from a local core facility specializing in working with these models is helpful at this stage and should be sought.

In 1992, one of the first publications mentioning humanized mice as the model emerged [42] . Since then, humanized mice have been frequently considered, praised, and applauded in numerous reviews as the most promising tools [25, 38, 43, 44, 47, 78, 93] . After all, the logic behind their application is appealing and seemingly sound.

This high level of optimism should be contrasted with the relatively small contribution of humanized mice to the general knowledge of sepsis. Since 1992, there have been fewer than 50 original papers published about sepsis and septic shock including humanized mice. This number is substantially low compared to the higher volume of publications about sepsis in general (n = 168,000) or utilizing mice to investigate sepsis (n = 13,000). The vast majority of the researchers published one manuscript involving humanized mice without any apparent sequels. The authors often quote high monetary and labor requirements, significant deficit in basic science, and a lack of clear advantages in accelerating drug development as significant reasons to abandon humanized models.

Any scientific model's development should always be linked to the clinical problem it is trying to address in all its complexity. Application of humanized mice to study sepsis encounters an unclear definition of the disease in question, and high heterogeneity [2, 5, 10, 13, 16, 61] . Furthermore, several confounders related to existence of the human immune system co-existing in mouse hosts must be considered [19, 48] . The progress in engineering humanized mice may yield unique experimental conditions considering the dynamism of grafts interacting with host.

One of the reasons for implementing humanized mice is a more effective translation of the therapies into clinical reality. There is an increasing implementation of humanized mice into pre-clinical trials [41, 53, 54, 64, [84] [85] [86] 90] . There is also the increasing emphasis on the long-term outcomes of septic patients [1, 10] . Unfortunately, most studies euthanized the animals within 24-48 h [20,23,53,92] . The most comprehensive study was performed over 28 days, which mimics the clinical estimation of sepsis mortality [31] . As we know, at the onset of sepsis, the human immune system becomes dysregulated and damages organ function directly or indirectly. Attention to organ failure should be included in research protocols. Some have suggested developing humanized mice with two organs: the immune system and lungs, to mimic certain conditions [78] . Others have suggested supplementing animals with human probiota [64] .

In conclusion, humanized models are valuable tools, but they are still in their infancy. Several confounders related to the xenotransplant nature of these animals renders any research to be scrutinized and related to other investigations.

We selected 136 publications from PubMed using keywords such as humanize mice, sepsis, and septic shock. This database was manually reviewed, and 8 reviews were eliminated, while 86 publications were found not relating to humanized mice in terms of stem cell transplantation. We eliminated any manuscript referencing humanized mice in the expression of other human tissues from immune ones. Subsequently, 54 publications are discussed. The representative samples of most critical publication are summarized in Table 2 . NOD.Cg-Prkdc/scidIL2rgamma (NSG) mice with the human cord blood CD34(+) CLP Sepsis induced a generalized up-regulation of both human and murine plasma cytokines (TNFalpha, IL-6, IL-10, IL-8/KC, MCP-1); it was additionally aggravated in P-DIE vs P-SUR. Human cytokines were strongly overridden by the murine ones (approx. ratio 1:9) but human TNFα was 7-fold higher than mouse TNFα

The effect of host environment on human leukocytes was attributed to increased mortality 

Surviving sepsis campaign international guidelines for the management of septic shock and sepsis-associated organ dysfunction in children

Evolving sepsis definitions and their impact on Acute Medical Units

Developing a New Definition and Assessing New Clinical Criteria for Septic Shock: For the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3)

Sepsis and septic shock

The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3)

Understanding Long-Term Outcomes Following Sepsis: Implications and Challenges

Long-Term Quality of Life Among Survivors of Severe Sepsis: Analyses of Two International Trials

Sepsis-induced immune dysfunction: Can immune therapies reduce mortality?

Surviving sepsis campaign: Research priorities for sepsis and septic shock

Neural reflex regulation of systemic inflammation: Potential new targets for sepsis therapy

The effects of sepsis on endothelium and clinical implications

Sepsis-Induced Cardiomyopathy: A Comprehensive Review

A unified theory of sepsis-induced acute kidney injury: Inflammation, microcirculatory dysfunction, bioenergetics, and the tubular cell adaptation to injury

Widespread Down-Regulation of Cardiac Mitochondrial and Sarcomeric Genes in Patients with Sepsis*

Septic shock: A microcirculation disease

Examination of biomarker expressions in sepsis-related DIC patients

Sepsis Pathophysiology, Chronic Critical Illness, and Persistent Inflammation-Immunosuppression and Catabolism Syndrome

Of Mice, Dirty Mice, and Men: Using Mice to Understand Human Immunology

MAIT cells launch a rapid, robust and distinct hyperinflammatory response to bacterial superantigens and quickly acquire an anergic phenotype that impedes their cognate antimicrobial function: Defining a novel mechanism of superantigen-induced immunopathology and immunosuppression

Age-dependent cellular reactions of the human immune system of humanized NOD scid gamma mice on LPS stimulus

Invariant Natural Killer T Cells Are Pathogenic in the HLA-DR4-Transgenic Humanized Mouse Model of Toxic Shock Syndrome and Can Be Targeted to Reduce Morbidity

Rapid and Rigorous IL-17A Production by a Distinct Subpopulation of Effector Memory T Lymphocytes Constitutes a Novel Mechanism of Toxic Shock Syndrome Immunopathology

Comparing hemodynamics, blood gas analyses and proinflammatory cytokines in endotoxemic and severely septic rats

Humanized Mice, a New Model to Study the Influence of Drug Treatment on Neonatal Sepsis

A novel model to study neonatal Escherichia coli sepsis and the effect of treatment on the human immune system using humanized mice

Cecal Ligation and Incision: An Acute Onset Model of Severe Sepsis in Rats

Cecal ligation and puncture: The gold standard model for polymicrobial sepsis?

Colon Ascendens Stent Peritonitis (CASP)-A Standardized Model for Polymicrobial Abdominal Sepsis

Long-term Monocyte Dysfunction after Sepsis in Humanized Mice Is Related to Persisted Activation of Macrophage-Colony Stimulation Factor (M-CSF) and Demethylation of PU.1, and It Can Be Reversed by Blocking M-CSF In Vitro or by Transplanting Naïve Autologous Stem Cells In Vivo

The clinical and immunological performance of 28 days survival model of cecal ligation and puncture in humanized mice

Early severe impairment of hematopoietic stem and progenitor cells from the bone marrow caused by CLP sepsis and endotoxemia in a humanized mice model

Temporal changes in tissue cardiorespiratory function during faecal peritonitis

Of mice and not men: Differences between mouse and human immunology

Abandon the Mouse Research Ship? Not Just Yet! Shock

Treatment of suspected sepsis and septic shock in children with chronic disease seen in the pediatric emergency department

The future of murine sepsis and trauma research models

Normal Structure, Function, and Histology of the Spleen

Immune Regulation by Mesenchymal Stem Cells Derived from Adult Spleen and Thymus

Innovations, challenges, and minimal information for standardization of humanized mice

Cytokine stimulation of multilineage hematopoiesis from immature human cells engrafted in SCID mice

Humanized mice for the study of infectious diseases

Humanized mice in translational biomedical research

A severe combined immunodeficiency mutation in the mouse

The fate and lifespan of human monocyte subsets in steady state and systemic inflammation

Humanized mice for immune system investigation: Progress, promise and challenges

Humanized hemato-lymphoid system mice

Introduction of Human Flt3-L and GM-CSF into Humanized Mice Enhances the Reconstitution and Maturation of Myeloid Dendritic Cells and the Development of Foxp3(+)CD4(+) T Cells

Sepsis-induced human lymphocyte apoptosis and cytokine production in "humanized" mice

The Fluctuations of Leukocytes and Circulating Cytokines in Septic Humanized Mice Vary with Outcome

Human macrophage and dendritic cell-specific silencing of high-mobility group protein B1 ameliorates sepsis in a humanized mouse model

Therapeutic targeting of HMGB1 during experimental sepsis modulates the inflammatory cytokine profile to one associated with improved clinical outcomes

An NLRP3 Mutation Causes Arthropathy and Osteoporosis in Humanized Mice

Humanized nonobese diabetic-scid IL2rγ mice are susceptible to lethal Salmonella Typhi infection

Increased Susceptibility of Humanized NSG Mice to Panton-Valentine Leukocidin and Staphylococcus aureus Skin Infection

Swift Intrahepatic Accumulation of Granulocytic Myeloid-Derived Suppressor Cells in a Humanized Mouse Model of Toxic Shock Syndrome

Staphylococcus aureusInfection in Humanized Mice: A New Model to Study Pathogenicity Associated with Human Immune Response

Epigenetic modifications of the immune system in health and disease

Adoptive transfer of naïve dendritic cells in resolving post-sepsis long-term immunosuppression

Neonatal sepsis

Monocytes derived from humanized neonatal NOD/SCID/IL2R mice are phenotypically immature and exhibit functional impairments

Long-term human CD34 + stem cell-engrafted nonobese diabetic/SCID/IL-2R gamma( null ) mice show impaired CD8 + T cell maintenance and a functional arrest of immature NK cells

Therapeutic Effect of an Anti-Human Programmed Death-Ligand 1 (PD-L1) Nanobody on Polymicrobial Sepsis in Humanized Mice

Humanized mice dually challenged with R5 and X4 HIV-1 show preferential R5 viremia and restricted X4 infection of CCR5 + CD4 + T cells. Microbes Infect

RNAi-Mediated CCR5 Knockdown Provides HIV-1 Resistance to Memory T Cells in Humanized BLT Mice

HIV Replication Is Not Controlled by CD8+ T Cells during the Acute Phase of the Infection in Humanized Mice

Plasmacytoid Dendritic Cells Suppress HIV-1 Replication but Contribute to HIV-1 Induced Immunopathogenesis in Humanized Mice

Coexistence of potent HIV-1 broadly neutralizing antibodies and antibody-sensitive viruses in a viremic controller

Zika viral infection and neutralizing human antibody response in a BLT humanized mouse model

Humanized Mice Show Clinical Signs of Dengue Fever according to Infecting Virus Genotype

Dengue virus infection and immune response in humanized RAG2( −/− )gamma(c) ( −/− ) (RAG-hu) mice

Dengue Virus-Infected Dendritic Cells, but Not Monocytes, Activate Natural Killer Cells through a Contact-Dependent Mechanism Involving Adhesion Molecules

Humanized Mice Reproduce Acute and Persistent Human Adenovirus Infection

Human plasmacytoid dendritic cells regulate immune responses to Epstein-Barr virus (EBV) infection and delay EBV-related mortality in humanized NOD-SCID mice

Humanized mice mount specific adaptive and innate immune responses to EBV and TSST-1

Hemophagocytic lymphohistiocytosis: A potentially underrecognized association with systemic inflammatory response syndrome, severe sepsis, and septic shock in adults

Of Mice and Men: The Coronavirus MHV and Mouse Models as a Translational Approach to Understand SARS-CoV-2. Viruses

A humanized mouse model for sequestration of Plasmodium falciparum sexual stages and in vivo evaluation of gametocytidal drugs

Plasmodium falciparum genetic crosses in a humanized mouse model

Plasmodium falciparum full life cycle and Plasmodium ovale liver stages in humanized mice

Humanized HLA-DR4.RagKO.IL2RγcKO.NOD (DRAG) mice sustain the complex vertebrate life cycle of Plasmodium falciparum malaria

Protection against malaria after immunization by chloroquine prophylaxis and sporozoites is mediated by preerythrocytic immunity

Characterization of Plasmodium vivax Proteins in Plasma-Derived Exosomes from Malaria-Infected Liver-Chimeric Humanized Mice

Humanized DRAGA mice immunized with Plasmodium falciparum sporozoites and chloroquine elicit protective pre-erythrocytic immunity

Antimalarial efficacy of MMV390048, an inhibitor of Plasmodium phosphatidylinositol 4-kinase

Gene array analysis of PD-1H overexpressing monocytes reveals a pro-inflammatory profile

The effect of Indomethacin and Betamethasone on the cytokine response of human neonatal mononuclear cells to gram-positive bacteria

Targeted delivery of HMGB1 siRNA to macrophages and dendritic cells for sepsis treatment in humanized mouse model (P6223)

a clinical development candidate for the treatment of severe malaria

Concise Review: Mesenchymal Stem (Stromal) Cells: Biology and Preclinical Evidence for Therapeutic Potential for Organ Dysfunction Following Trauma or Sepsis

Deficient cytokine response of human allergen-specific T lymphocytes from humanized SCID mice and reconstitution by professional antigen-presenting cells

Myeloid Engraftment in Humanized Mice: Impact of Granulocyte-Colony Stimulating Factor Treatment and Transgenic Mouse Strain

Dominant Cross-Reactive B Cell Response during Secondary Acute Dengue Virus Infection in Humans

Animal Models of Dengue Virus Infection

Adhesion of Neisseria meningitidis to Dermal Vessels Leads to Local Vascular Damage and Purpura in a Humanized Mouse Model

Development of human CD4 FoxP3 regulatory T cells in human stem cell factor-, granulocyte-macrophage colony-stimulating factor-, and interleukin-3-expressing NOD-SCID IL2Rγ humanized mice

Development and function of human innate immune cells in a humanized mouse model

Author would like to thank Lucia DiNapoli for critical review of this work.

The authors declare no conflict of interest.