Significance of humanized mouse models for evaluating humoral immune response against cancer vaccines

Personalized Medicine Universe xxx (2018) 1e6

Contents lists available at ScienceDirect

Personalized Medicine Universe journal homepage: www.elsevier.com/locate/pmu

Review

Significance of humanized mouse models for evaluating humoral immune response against cancer vaccines Yoshie Kametani a, b, *, Asuka Miyamoto a, c, Toshiro Seki d, Ryoji Ito e, Sonoko Habu f, Yutaka Tokuda c a

Department of Molecular Life Science, Division of Basic Medical Science, Tokai University School of Medicine, Isehara, Kanagawa, Japan Institute of Advanced Biosciences, Tokai University, Hiratsuka, Kanagawa, Japan Department of Breast and Endocrine Surgery, Tokai University School of Medicine, Isehara, Kanagawa, Japan d Division of Nephrology, Endocrinology and Metabolism, Department of Internal Medicine, Tokai University School of Medicine, Isehara, Kanagawa, Japan e Central Institute for Experimental Animals, Kawasaki, Kanagawa, Japan f Department of Immunology, Juntendo University School of Medicine, Tokyo, Japan b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 February 2018 Received in revised form 4 April 2018 Accepted 5 April 2018

Because accumulation of mutations produces unique neoantigen patterns in cancer cells, it has been hypothesized that these patterns are recognized by the patients' immune system. Therefore, the development of cancer vaccines has been challenging owing to the possibility of an anti-cancer effect induced by the immune system against such neoantigens. However, it is difficult to develop effective vaccines because of the variety of mutations induced in cancer cells and human leukocyte antigen (HLA), which is predicted to present the neoantigen-derived peptides. Moreover, the activation of cancerspecific cytotoxic cells is inhibited by cancer cell immunoediting in each patient. Although cellular immunity can be analyzed ex vivo, there are no definite methods to evaluate humoral immunity. A humanized mouse model has been developed and used for evaluating the multipotency of hematopoietic stem cells. Presently, significant improvements in the humanized mouse model approach have partially recapitulated human humoral immunity in vivo, and human antibody production can be induced in the mouse model. These mice can be used to produce new cancer vaccines and to establish polypharmacy protocols in a preclinical model. In this review, we discuss the preclinical evaluation of the cancer vaccines using humanized mice transferred with patient-derived peripheral blood mononuclear cells (PBMCs) to advance a personalized medicine approach. Copyright © 2018, International Society of Personalized Medicine. Published by Elsevier B.V. All rights reserved.

Keywords: Cancer vaccine Humanized mouse B cell function Immune checkpoint

1. Introduction It is well-known that a high rate of mutations leads to the accumulation of neoantigens in cancer cells. Therefore, enhancement of the immune function against such neoantigens is believed to induce anti-cancer effects. Massive effort has been made towards the development of anti-cancer vaccines [1]. However, the mutations observed in cancer cells are highly varied, which make the development of pan-specific vaccines difficult. Moreover, human leukocyte antigen (HLA) molecules also vary greatly and do not easily predict the efficacy and adverse effects of vaccines.

* Corresponding author. Department of Molecular Life Science, Division of Basic Medical Science, Tokai University School of Medicine, Isehara, Kanagawa, Japan. E-mail address: [email protected] (Y. Kametani).

Presentation of a specific peptide on the HLA molecules has been predicted by several algorithms, but their accuracy is inadequate [2,3]. Because neoantigens tend to have limited and unique sequences, the specific peptides also tend to be of a limited length. Therefore, HLA molecules, which are predicted to present neoantigens, are also very strictly determined. This is one of the most difficult issues to address for advancing personalized medicine for patients with cancer. To verify the overall vaccine effect of such peptides, humanized mice equipped with the human immune system have gradually attracted considerable interest. Humanized mice have been developed to measure the multipotency of hematopoietic stem cells (HSCs). Until now, various immunodeficient mouse strains have been developed and used for examining HSC multipotency [4,5]. However, the development and differentiation of human T and B cells that enable the induction of

https://doi.org/10.1016/j.pmu.2018.04.002 2186-4950/Copyright © 2018, International Society of Personalized Medicine. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: Kametani Y, et al., Significance of humanized mouse models for evaluating humoral immune response against cancer vaccines, Personalized Medicine Universe (2018), https://doi.org/10.1016/j.pmu.2018.04.002

2

Y. Kametani et al. / Personalized Medicine Universe xxx (2018) 1e6

complete humoral immunity have not been achieved in these mouse strains. Therefore, the humanized mouse models have not adequately analyzed human B cell functions. Two well-known immunodeficient mouse strains, NOD.Cg-Prkdcscid Il2rgtm1Sug/Jic (NOG) and NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG), have been used for the development of human T and B cells via HSC transplantation [6,7]. However, even in mice, humoral immunity is severely diminished. Because these mice harbor a limited number of human myeloid cells, it is considered that the malfunction of humoral immunity is due to the developmental delay in human myeloid cells. Thus, transgenic mouse strains expressing human cytokine genes that enhance myeloid cell development were created, based on the NOG and NSG mice strains, and the production of antigenspecific IgG antibodies was observed along with other lymphoid tissues [8]. However, these mouse strains must mate with several other mice strains, and the selection of mice with three independent human genes must be continued. Despite the effort to maintain these strains, few antibodies are produced in these mice. Moreover, humanized mice use human HSCs derived from cord blood, indicating that the mice cannot mimic the immune environment of the patient, although such mimicry is essential for personalized diagnosis and medicine. If we use mobilized HSCs, it is highly invasive for the patients and is impractical. Following the first report by Williams, human peripheral blood mononuclear cells (PBMCs) have been transplanted in immunodeficient mouse strains [9]. However, because these immunodeficient recipient mice developed graft versus host disease (GVHD), the mouse model was not a suitable model for the analysis of human immune reactions during a long duration of vaccine treatment. Improvements of the recipient mouse strains using various transgenes related to the immune system are still continuing. These mouse strains are expected to develop normal human immune cells without the development of GVHD, in addition to antigen-specific IgG production. If such mouse strains are established, they will become promising tools for development of new cancer vaccines and protocols of vaccination and polypharmacy and verification of large volume of data needed to apply personalized medicine. Simultaneously, if the mouse strains can develop normal human cellular immunity, they may have a larger impact in this field. In this review, we will discuss the usefulness of current humanized mouse models to evaluate personalized peptide vaccination protocols. 2. Limitation of vaccine selection after neoantigen screening As mentioned above, the high rate of mutations accumulates neoantigens in cancer cells [1,10]. Generally, serological analysis of recombinant cDNA expression libraries (SEREX) is performed forwide-scale screening and detection of cancer antigens in the sera of the patients [11]. The immune reaction, which targets the antigens selected by SEREX, may induce highly effective anti-cancer effects if the reaction functions effectively. In fact, several mouse models have been reported to successfully produce such characteristics. However, it is also well-known that cellular immunity is significantly suppressed in tumor-bearing individuals. There are many reports to correlate malfunction and cancer immunoediting, involving Treg differentiation, expression of the ligands of immune checkpoint molecules that induce T cell exhaustion, and HLA expression downregulation [12,13]. However, several successful antigens, such as MAGE-A1, NY-ESO-1, and SSX-2, have been identified [14e17]. If these peptides are presented on class I HLA, they may be effective in invoking an immune response in the patient [18]. Some peptide vaccines are not effective in tumorbearing individuals but are effective in inhibiting recurrence. For example, E75 and GP2 are breast cancer vaccines, the effects of

which have not been identified in patients with cancer. However, the inhibition of recurrence has been observed in patients treated with peptides and granulocyte-macrophage colony-stimulating factor (GM-CSF) adjuvant [19,20]. Although the mechanism of action has not been determined, their efficacy is recognized in human erbB-2 receptor 2 (Her2) 1þ and Her2 2þ patients compared to that in Her2 3þ patients. Moreover, as previously described for general neoantigens, these peptide vaccines have limited HLA alleles for effective presentation. Generally, peptide vaccines also need class I and class II HLA epitopes to activate both cytotoxic T cells (Tc) and helper T cells (Th). The reason is because Th cells supply cytokines essential for the expansion and survival of Tc cells [21]. Therefore, if we aim to establish a peptide vaccine with optimal effectiveness, the peptide sequence and HLA type of the patients become more restricted. However, prediction of the presentation ability of class I and class II HLAs using various algorithms is not always effective. Because most of the self-reactive T cells are deleted in the thymus by negative selection, neoantigen-specific T cells may be deleted in the thymus if the neoantigens are the only modified molecules and possess highly limited unique sequences. We have been developing a candidate peptide, named CH401MAP, as a breast cancer vaccine [22]. CH401MAP is composed of 20 amino acid peptides of the HER2 molecule. CH401MAP peptides involve the epitope of a monoclonal antibody (mAb), CH401, which was developed by Ishida et al. at Sapporo Medical University; this mAb induced apoptosis of HER2-positive cancer cells [23]. The epitope is identified to be different from that of Trastuzumab, a well-known and effective molecularly targeted mAb drug. Compared with the epitopes of other already available HER2 peptide vaccines, the CH401MAP epitope has the highest affinity for HLA-A*24:02, an HLA class I allele present at the highest ratios in Japanese individuals. The peptide was predicted to be presented by other HLA isotypes. Eventually, the peptide has been predicted by several algorithms to be presented by most of the class I and class II HLA molecules of Japanese breast cancer patients. We also showed that PBMCs from the breast cancer patients can be stimulated by the peptides and secrete interleukin-2 (IL-2) [24]. These results suggest that CH401MAP also has the potential to activate cellular immunity in breast cancer patients. Contrary to cellular immunity, there have been no successful reports on the induction apoptosis in cancer cells in response to a humoral immune reaction after cancer vaccination. Although active immunization with vaccines, which is simple, convenient, and yields high quality of life (QOL), has been effective for many infectious diseases, it is intriguing that humoral immunity in cancer patients does not induce apoptosis in cancer cells. In any event, evidence suggests suppression of humoral immunity in cancer patients. To evaluate the effectiveness of the vaccines, it must be confirmed whether B cell clones expressing IgG molecules specifically reactive to the neoantigens are present among the B cells of the patient. Till date, however, methods have not been developed, other than directly injecting the antigen into the patients, to examine whether a humoral response is elicited after vaccination. No protocols exist for detecting the effector function related to humoral immunity, such as the activity of Th cells and production of antibodies against the peptide vaccine. It is one of the most serious problems to overcome. Moreover, adverse effects due to vaccination are another serious problem. It is difficult to predict the adverse effects induced by the treatment in individual patients. Human papilloma virus vaccine, which has been developed for the prevention of cervical cancer and has been shown to be effective, occasionally leads to the development of severe adverse effects in the recipient [25]. The mechanism

Please cite this article in press as: Kametani Y, et al., Significance of humanized mouse models for evaluating humoral immune response against cancer vaccines, Personalized Medicine Universe (2018), https://doi.org/10.1016/j.pmu.2018.04.002

Y. Kametani et al. / Personalized Medicine Universe xxx (2018) 1e6

is not completely understood yet, although the immune response is related to the symptoms. 3. Evaluation of humoral immunity induced against peptide vaccines as anti-cancer drugs Although peptide vaccines present many limitations, killer activity was induced by some peptide vaccine treatments to induce significant anti-cancer effect [26]. However, very few patients responded to the vaccines. Because many excellent reviews have been published, we will not discuss cellular immunity with respect to killer T cell induction and the killing activity in detail [27e29]. Moreover, previous reports on cancer immunity mention the detection of anti-cancer antibodies in tumor-bearing patients using techniques similar to SEREX, which has been used for detecting neoantigens [11]. Nevertheless, patients who produce anti-cancer antibodies cannot induce apoptosis of cancer cells, indicating that the antibodies detected in the patients' sera are not sufficient to reject the growing cancer mass. The evidence is, at a glance, contradictory to the high efficacy of passively transfused humanized antibodies, such as Trastuzumab and Rituximab. Therefore, the presence of a high antibody titer against cancer antigens should be sufficient to induce an anti-cancer effect in the immune system of cancer patients, but, curiously, the patients cannot reject the cancer cells. The antibodies produced were not sufficient, both quantitatively and qualitatively, suggesting a suppressive effect induced by the patients' humoral immunity. One of the potential explanations may be that the epitope of the antibodies produced by cancer patients is different from that of anti-cancer drug mAbs, and anticancer effect cannot be induced by the non-effective, epitopebearing antibodies produced in the patients. Hence, vaccines for inducing humoral immunity against cancer neoantigens have not yet been established. Accordingly, cancer-bearing patients may need to be continuously administered with mAbs carrying the effective epitope instead of inducing antibody secretion by active immunization in vivo. In contrast, B cells inhibit the activation of T cells, which decreases the killer activity of Tc cells and, thus, the anti-cancer effect [30]. Furthermore, B cells are increased in the PBMCs of cancer patients [31], and the B cell ratio is significantly higher in the PBMCs of breast cancer patients (treatment naïve patients and recurrent patients were not included) than in that in PBMCs of healthy donors [24]. The reason why vaccination does not elicit an immune response to secrete enough amount of anti-tumor antibodies, even if B cells are abundant in the PBMCs of the patients, is unknown. Recent reports have suggested a relation between cancer and regulatory B cells (Bregs); thus, immune suppression by B cells may become an important aspect of cancer immunoediting [32,33]. It is relatively easy to identify mouse Bregs by detecting the cell surface markers, but human Bregs are difficult to identify in fresh PBMCs because human Bregs have various phenotypes and their identification requires the secretion of cytokines. Bregs are found in tumor tissues, indicating the infiltration of these cells associated with poor prognosis and induction of inflammation [34e36]. These findings indicate that Bregs play a role in tumor progression. Until now, it is unclear whether the capacity of conventional B cells to secrete antibodies is inhibited by Bregs of the patient and, thus, needs to be clarified. If Bregs suppress antibody secretion by conventional B cells, and the failure of vaccination is due to Breg function, Breg control may become another target of cancer immunity. If we do not use in vivo animal models, accumulating integrated clinical evidence and predicting personalized medicine protocols needs a substantial amount of examination and complex diagnostic

3

skills. The performance level may decrease if all the patients need such diagnosis protocols. If humanized mouse models are introduced in the validation of personalized medicine, they may become powerful tools aimed at insuring patients' safety and cost reduction. 4. History of humanized mouse as a monitoring model of human immunity Humanized mice have been developed to measure the multipotency of human HSCs and progenitors. In parallel with the improvement of transplantation techniques, severely immunodeficient mouse strains have been developed [37e40]. After enormous research and development efforts, the transplantation of human HSCs into severely immunodeficient mice finally lead to the generation of human T, B, and myeloid cells, which are localized in the periphery of the mice [41e43]. These mice models can be used to analyze human hematopoietic cells, including leukemic stem cells [44]. However, human antibody production has also been attempted using these mouse models transferred with various types of human cells (Fig. 1) [45]. This approach is based on the success of Trastuzumab and Rituximab as anti-cancer drugs and is another aspect of the research using humanized mice models. NOG and/or NSG strains, established at the Central Institute for Experimental Animals or at the Jackson Laboratory, are severely immunodeficient mouse strains. Both of these mice strains have a deficiency of IL-2rgc [39,46,47]. NOG mice possess truncated IL2rgc, and NSG mice have a complete deletion of this gene. Using these strains, it is possible to develop human T and B cells from human HSCs in a xenogenic environment. However, most of the human B cells differentiated in the mice expressed CD5, a marker of B1 cells, but the specific IgG antibody was not produced (Table 1) [48e52]. We also reconstituted human immunity in NOG mice

HSC of HD/PaƟents Vaccine

Response of immune system by vaccine

Immature/aberrant immunity

Lymphoid Tissues of Newborn

Vaccine

Response of immune system by vaccine

GVHD

PBMC of HD/PaƟents Vaccine

Response of immune system by vaccine

GVHD

Fig. 1. Characteristics of humanized mouse model. Humanized mouse can be developed using one of the three major protocols. The first involves transplantation of HSCs (Upper panel). The mouse does not exhibit GVHD, but the collection of HSCs requires an invasive procedure. Moreover, the human immune system that develops in the mouse is immature or aberrant. The second is transplantation of human lymphoid tissues, such as fetal liver and thymus (Middle panel). This model requires human fetal tissues, which creates ethical problems. Moreover, the mouse transplanted with human tissues eventually exhibits GVHD. The third is the transplantation of human PBMCs (Lower panel). This model has the advantage of reconstructing the patients' own immune system in the mouse, but the mouse tends to exhibit GVHD.

Please cite this article in press as: Kametani Y, et al., Significance of humanized mouse models for evaluating humoral immune response against cancer vaccines, Personalized Medicine Universe (2018), https://doi.org/10.1016/j.pmu.2018.04.002

4

Y. Kametani et al. / Personalized Medicine Universe xxx (2018) 1e6

Table 1 The humanized SCID-derived immunodeficient mouse and its antibody response. Mouse strain

Transplanted tissues

Antigen

Isotype

Reference

SCID-Hu

SCID

Aaberge IS; pneumococcal vaccine

IgG

McCune JM 1988, Aaberge IS et al., 1992

Hu-HSC

NOG NOG NOG-HLA-DR4/IAbKO NSG NSG-HIS-CD4/B

human fetal liver and thymic fragments under kidney capsule HSC (CB/MPB/BM) CD34 þ HSC human CD34 þ HSC human CD34 þ HSC human CD34 þ HSC

IgM Igs IgM, IgG IgM, IgG IgG

Matsumura T et al., 2003, Kametani Y et al., 2006 Yajima M et al., 2008, Watanabe Y et al., 2009 Suzuki M et al., 2012 Watanabe S et al., 2007, Singh M et al., 2012 Shurz LD et al., 2010, Huang J et al., 2015

Hu-PBL

SCID NOG-IL-4-Tg NOD-SCID

DNP-KLH/CH401MAP/TSST-1 OVA OVA OVA, HIV Plasmodium falciparum circumsporozoite (PfCS) protein xenograft KLH/CH401MAP HIV-1, WNV envelope protein

IgM, IgG IgG IgM, IgG

Williams S et al., 1992 Kametani Y et al., 2017 Biswas S et al., 2011, Immunology

Aaberge IS; pneumococcal vaccine Dengue virus infection, Zika virus

IgG IgM

McCune JM et al., 1988, Aaberge IS et al., 1992 Jaiswal S et al., 2015, Schmitt K et al., 2018

BLT

SCID NSG

human PBMC human PBMC human fetal liver and thymic fragments under kidney capsule with autologous CD34 þ HSC human fetal liver and thymic fragments under kidney capsule with autologous CD34 þ HSC

transplanted with HSC and immunized with CH401MAP, keyhole limpet hemocyanin (KLH), or toxic shock syndrome toxin-1 (TSST1) emulsified in Freund's complete adjuvant and measured the specific antibody titer by ELISA. As a result, although antigenspecific IgM and non-specific IgG were produced, antigen-specific IgG was not produced in the mice (Table 1) [53]. These mice did not develop a germinal center, which has a structure composed of T, B, and follicular dendritic cells and plays a crucial role in specific IgG antibody production. Although HLA class I transgenic mice evoke partially improved human humoral immune responses, the responses cannot be completely mimicked [8,54,55]. Currently, various transgenic mouse strains expressing human cytokines and surface antigens along with more severely immunodeficient mouse strains are being developed. Although most of the strains support the differentiation of hematopoietic cells from human HSCs, establishing a normal human immune environment in a mouse model still needs further improvement. Meanwhile, another humanized mouse model, so-called BLT mice, has been reported. In this mouse model, immunodeficient mice are co-transplanted with human fetal liver and thymus tissues along with autologous CD34þ HSCs. BLT mice are a modification of the SCID-Hu mice developed by MacCune [52,56,57]. In these mice, antigen-specific antibody production was partially achieved, and experiments on infection with bacteria or viruses have been conducted [58]. Severely immunodeficient NSG mice are used to establish NSGeBLT mice [59]. Human SCF, GM-CSF, and IL-3 genes were transduced into NSG mice, and an improved BLT mouse strain was established. Human HSCs, fetal liver, and fetal thymus were transplanted, and the dengue and/or Zika virus were infected into the mice. The mice expressed a higher immune response than that of conventional NSG mice, although GVHD could not be avoided (Table 1) [60,61]. However, there is a serious ethical problem because human fetal tissues were used to develop the BLT mice. Apart from this problem, various strains of severely immunodeficient mice and human lymphoid tissues and/or hematopoietic cells were combined to establish useful humanized mouse models. Among them, as described below, reconstitution of human immunity with human PBMCs has been well-studied and reported. 5. Development of humoral immunity in humanized mouse As described above, it is not easy to reconstitute human humoral immunity in severely immunodeficient mice. Moreover, to reflect the immune environment of patients, it is not practical to use HSCbased humanized mice. The transfer of PBMCs into severely immunodeficient mice can reflect the immune system of the

patients [9]. There are reports on detection of specific IgG antibodies (Table 1). However, these mice develop severe GVHD, which suppresses normal B cell function [62]. In mice transplanted with human PBMCs, Tc and Th cells were activated to induce GVHD [63]. Under these conditions, transient production of human antibody was observed, but long-term production of antigen-specific antibodies was not achieved. Therefore, we developed a novel model of PBMC-transferred NOG mice without GVHD, in which antigenspecific IgG antibodies are produced. We introduced human IL-4 under the control of the cytomegalovirus (CMV) promoter into NOG mice, which systemically expressed human IL-4. After transplantation of human PBMCs into NOGeIL-4eTg mice, GVHD was significantly delayed compared to that in NOG mice, and human B cells remained in the lymphoid tissues over four weeks. These human B cells were categorized as plasmablasts by surface marker analysis. CH401MAP and KLH-specific IgG antibodies were successfully produced in these mice after immunization, but germinal center formation was not observed [64]. Although evidence suggests the absence of affinity maturation in these mice, the amount of specific IgG antibody could be quantified. This mouse model is promising to evaluate B cell differentiation ability, antibody production, and cytokine profile of an individual patient. 6. Evaluation of therapy using immune checkpoint antibodies in the humanized mouse model Of late, so-called immune checkpoint antibodies, such as antiprogrammed cell death-1 (PD-1), anti-programmed cell death ligand-1 (PD-L1), and anti-cytotoxic T lymphocyte-associated protein 4 (CTLA-4) antibodies as well as antibodies against members of the same family are expected to become significant anti-cancer drugs [65e67]. These antibodies neutralize signals and induce exhaustion and apoptosis of long-term activated T cells [68]. Immune checkpoint molecules are activation antigens expressed on the surface of immune cells and their ligands. For example, to be activated normally, naïve T cells express CD28, which interacts with B7 molecules (CD80 and CD86) expressed on antigen-presenting cells. Following their activation, T cells express CTLA-4 molecules, which have a higher affinity for B7 molecules (CD80 and CD86). This CTLA-4/B7 interaction suppresses further activation of T cells, preventing their hyperactivation. However, a part of activated T cells develops into effecter T cells that express PD-1 molecules later than CTLA-4 expression. These activated T cells are exhausted by interacting with PD-L1 molecules, which are the ligands of PD-1 molecules, and these T cells undergo apoptosis to prevent their infinite activation. These molecules form a family named immune

Please cite this article in press as: Kametani Y, et al., Significance of humanized mouse models for evaluating humoral immune response against cancer vaccines, Personalized Medicine Universe (2018), https://doi.org/10.1016/j.pmu.2018.04.002

Y. Kametani et al. / Personalized Medicine Universe xxx (2018) 1e6

checkpoint molecules. Immune checkpoint antibodies are prepared to neutralize the interaction of these activation antigens with their ligands. These antibodies are believed to maintain T cells in their activated state by avoiding their exhaustion, which is induced by immune checkpoint molecules. Because tumor-specific T cell exhaustion decreases the anti-cancer effect of activated T cells in a cancer-bearing state, immune checkpoint antibodies, which evade the exhaustion of T cells, have attracted much attention as promising anti-cancer drugs. The function of immune checkpoint antibodies is different from that of conventional anti-tumor antigen antibodies, such as Trastuzumab and rituximab, which directly react with tumor antigens expressed on the surface of cancer cells. The evidence gained from the use of these anti-tumor antigen antibodies confirmed that malignant tumors are closely associated with the “suppression of the production of enough specific antibodies against tumor antigens” because these antibodies are considered to activate additional humoral response. Meanwhile, immune checkpoint antibodies prevent the exhaustion of Tc and Th cells to maintain the cytotoxic and humoral immune responses, respectively. Although anti-tumor antigen antibodies induce anti-cancer responses at a high rate, immune checkpoint antibodies induce only 20% of anticancer responses. Moreover, the latter has the risk of inducing severe adverse effects. To understand the underlying cause of cases with and/or without response, researchers have continued their experimental work. Human immunity in humanized mice is not completely recapitulated, but it may be useful for verifying the efficacy and adverse effects of the antibody drugs. For example, the accumulation of Opdivo in the salivary and lachrymal glands was observed in a humanized mouse model [69]. Inflammatory adverse effects were observed in the mucosa of the gastrointestinal system and pituitary gland in human patients [70e72]. We may be able to reconstitute such inflammation in our NOG or transgenic NOG model with further development of the mouse model. Using MHC class I/II double knockout NOG, Akiyama et al. succeeded in reconstituting the patients' immune environment and the effect of Opdivo injection [69]. Humanized mouse models transplanted with human PBMCs will support the evaluation of a protocol of molecular targeted drugs and polypharmacy. 7. Conclusion In this review, we discussed the humanized mouse model, which is useful for the evaluation of humoral immunity and personalized cancer vaccine treatment. Although neoantigens distinguish between the different characteristics of each cancer cell type, vaccines are difficult to develop owing to the diversity of antigens and HLA types. Moreover, the strategy to evaluate humoral immunity induced by vaccines against neoantigens has not been established. Meanwhile, a humanized mouse system has been developed to investigate the human immune response in vivo. Recently, some strains of immunodeficient mice have been shown to be useful for evaluating humoral immune response. If the immune cells of patients are obtained by a non-invasive protocol and reconstituted in humanized mice, they may be used for the evaluation of potential drugs. In the near future, preclinical data for the drug treatment obtained from humanized mice are expected to be successfully applied for personalized therapy of cancer patients. References [1] Efremova M, Finotello F, Rieder D, et al. Neoantigens generated by individual mutations and their role in cancer immunity and immunotherapy. Front Immunol 2017;8:1679.

5

[2] Parker K, Bednarek M, Coligan J. Scheme for ranking potential HLA-A2 binding peptides based on independent binding of individual peptide side-chains. J Immunol 1994;152:163e75. [3] Rammensee H, Bachmann J, Emmerich NP, et al. SYFPEITHI: database for MHC ligands and peptide motifs. Immunogenetics 1999;50:213e9. [4] Peschle C, Botta R, Müller R, et al. Purification and functional assay of pluripotent hematopoietic stem cells. Rev Clin Exp Hematol 2001;5:3e14. [5] Ando K. Human CD34-hematopoietic stem cells: basic features and clinical relevance. Int J Hematol 2002;75:370e5. [6] Ito M, Hiramatsu H, Kobayashi K, et al. NOD/SCID/gamma(c)(null) mouse: an excellent recipient mouse model for engagement of human cells. Blood 2002;100:3175e82. [7] Shultz LD, Lang PA, Christianson SW, et al. NOD/LtSz-Rag1null mice: an immunodeficient and radioresistant model for engraftment of human hematolymphoid cells, HIV infection, and adoptive transfer of NOD mouse diabetogenic T cells. J Immunol 2000;164:2496e507. [8] Jangalwe S, Shultz LD, Mathew A, et al. Improved B cell development in humanized NOD-scid IL2Rgnull mice transgenically expressing human stem cell factor, granulocyte-macrophage colony-stimulating factor and interleukin-3. Immun Inflamm Dis 2016;4:427e40. [9] Williams SS, Umemoto T, Kida H, et al. Engraftment of human peripheral blood leukocytes into severe combined immunodeficient mice results in the long term and dynamic production of human xenoreactive antibodies. J Immunol 1992;149:2830e6. [10] Schietinger A, Philip M, Schreiber H. Specificity in cancer immunotherapy. Semin Immunol 2017;20:276e85. [11] Zhou S, Yi T, Zhang B, et al. Mapping the high throughput SEREX technology screening for novel tumor antigens. Comb Chem High Throughput Screen 2012;15:202e15. [12] Butt AQ, Mills KH. Immunosuppressive networks and checkpoints controlling antitumor immunity and their blockade in the development of cancer immunotherapeutics and vaccines. Oncogene 2014;33:4623e31. [13] Takeuchi Y, Nishikawa H. Roles of regulatory T cells in cancer immunity. Int Immunol 2016;28:401e9. [14] Chaux P, Luiten R, Demotte N, et al. Identification of five MAGE-A1 epitopes recognized by cytolytic T lymphocytes obtained by in vitro stimulation with dendritic cells transduced with MAGE-A1. J Immunol 1999;163:2928e36. [15] Traversari C, van der BP, Luescher IF, et al. A nonapeptide encoded by human gene MAGE-1 is recognized on HLA-A1 by cytolytic T lymphocytes directed against tumor antigen MZ2-E. J Exp Med 1992;176:1453e7. [16] Jager E, Chen YT, Drijfhout JW, et al. Simultaneous humoral and cellular immune response against cancer-testis antigen NY-ESO-1: definition of human histocompatibility leukocyte antigen (HLA)-A2-binding peptide epitopes. J Exp Med 1998;187:265. [17] Ayyoub M, Stevanovic S, Sahin U, et al. Proteasome-assisted identification of a SSX-2-derived epitope recognized by tumor-reactive CTL infiltrating metastatic melanoma. J Immunol 2002;168:1717e22. [18] Fritsch EF, Rajasagi M, Ott PA, et al. HLA-binding properties of tumor neoepitopes in humans. Cancer Immunol Res 2014;2:522e9. [19] Clifton GT, Gall V, Peoples GE, et al. Clinical development of the E75 vaccine in breast cancer. Breast Care (Basel) 2016;11:116e21. [20] Mittendorf EA, Ardavanis A, Litton JK, et al. Primary analysis of a prospective, randomized, single-blinded phase II trial evaluating the HER2 peptide GP2 vaccine in breast cancer patients to prevent recurrence. Oncotarget 2016;7: 66192e201. [21] Cheever MA, Greenberg PD, Fefer A. Potential for specific cancer therapy with immune T lymphocytes. J Biol Response Mod 1984;3:113e27. [22] Miyako H, Kametani Y, Katano I, et al. Antitumor effect of new HER2 peptide vaccination based on B cell epitope. Anticancer Res 2011;31:361e3368. [23] Ishida T, Tsujisaki M, Hinoda Y, et al. Establishment and characterization of mouse-human chimeric monoclonal antibody to erbB-2 product. Jpn J Cancer Res 1994;85:172e8. [24] Tsuda B, Miyamoto A, Yokoyama K, et al. B-cell populations are expanded in breast cancer patients compared with healthy controls. Breast Cancer 2017 in press (Epub ahead of print). [25] de Vos van Steenwijk PJ, van Poelgeest MI, Ramwadhdoebe TH, et al. The longterm immune response after HPV16 peptide vaccination in women with lowgrade pre-malignant disorders of the uterine cervix: a placebo-controlled phase II study. Cancer Immunol Immunother 2014;63:147e60. [26] Tagliamonte M, Petrizzo A, Tornesello M, et al. Antigen-specific vaccines for cancer treatment. Hum Vaccin Immunother 2014;10:3332e46. [27] Yamada A, Sasada T, Noguchi M, et al. Next-generation peptide vaccines for advanced cancer. Cancer Sci 2013;104:15e21. [28] Reinherz EL, Keskin DB, Reinhold B. Forward vaccinology: CTL targeting based upon physical detection of HLA-bound peptides. Front Immunol 2014;5:418. [29] Lizee G, Overwijk WW, Radvanyi L, et al. Harnessing the power of the immune system to target cancer. Annu Rev Med 2013;64:71e90. [30] Affara NI, Ruffell B, Medler TR, et al. B cells regulate macrophage phenotype and response to chemotherapy in squamous carcinomas. Cancer Cell 2014;25: 809e21.
O, Novotn
[31] Babusíkova a L, Korec S, et al. Change in the proportion of T and B lymphocytes in human malignant neoplasia in relation to the clinical stage. Neoplasma 1975;22:413e21. [32] Schwartz M, Zhang Y, Rosenblatt J. B cell regulation of the anti-tumor response and role in carcinogenesis. J Immunother Cancer 2016;4:40.

Please cite this article in press as: Kametani Y, et al., Significance of humanized mouse models for evaluating humoral immune response against cancer vaccines, Personalized Medicine Universe (2018), https://doi.org/10.1016/j.pmu.2018.04.002

6

Y. Kametani et al. / Personalized Medicine Universe xxx (2018) 1e6

[33] Weingartner E, Golding A. Direct control of B cells by Tregs: an opportunity for long-term modulation of the humoral response. Cell Immunol 2017;318:8e16. [34] Zhou X, Su YX, Lao XM, et al. CD19(þ)IL-10(þ) regulatory B cells affect survival of tongue squamous cell carcinoma patients and induce resting CD4(þ) T cells to CD4(þ)Foxp3(þ) regulatory T cells. Oral Oncol 2016;53:27e35. €ßer HA, Gryschok L, et al. Characterization [35] Shimabukuro-Vornhagen A, Schlo of tumor-associated B-cell subsets in patients with colorectal cancer. Oncotarget 2014;5:4651e64. [36] Andreu P, Johansson M, Affara NI, et al. FcRgamma activation regulates inflammation-associated squamous carcinogenesis. Cancer Cell 2010;17: 121e34. [37] Ito M, Kobayashi K, Nakahata T. NOD/Shi-scid IL2rgamma(null) (NOG) mice more appropriate for humanized mouse models. Curr Top Microbiol Immunol 2008;324:53e76. [38] Yajima M, Imadome K, Nakagawa A, et al. A new humanized mouse model of Epstein-Barr virus infection that reproduces persistent infection, lymphoproliferative disorder, and cell-mediated and humoral immune responses. J Infect Dis 2008;198:673e82. [39] Brehm MA, Wiles MV, Greiner DL, et al. Generation of improved humanized mouse models for human infectious diseases. J Immunol Methods 2014;410: 3e17. [40] Hasgur S, Aryee KE, Shultz LD, et al. Generation of immunodeficient mice bearing human immune systems by the engraftment of hematopoietic stem cells. Methods Mol Biol 2016;1438:67e78. [41] Willinger T, Rongvaux A, Takizawa H, et al. Human IL-3/GM-CSF knock-in mice support human alveolar macrophage development and human immune responses in the lung. Proc Natl Acad Sci USA 2011;108:2390e5. [42] Singh M, Singh P, Gaudray G, et al. An improved protocol for efficient engraftment in NOD/LTSZ-SCIDIL-2Rgnull mice allows HIV replication and development of anti-HIV immune responses. PLoS One 2012;7:e38491. [43] Ito R, Takahashi T, Katano I, et al. Establishment of a human allergy model using human IL-3/GM-CSF-transgenic NOG mice. J Immunol 2013;191: 2890e9. [44] Goyama S, Wunderlich M, Mulloy J. Xenograft models for normal and malignant stem cells. Blood 2015;125:2630e40. [45] Villaudy J, Schotte R, Legrand N, et al. Critical assessment of human antibody generation in humanized mouse models. J Immunol Methods 2014;410: 18e27. [46] Ito R, Shiina M, Saito Y, et al. Antigen-specific antibody production of human B cells in NOG mice reconstituted with the human immune system. Curr Top Microbiol Immunol 2008;324:95e107. [47] Ito R, Takahashi T, Katano I, et al. Current advances in humanized mouse models. Cell Mol Immunol 2012;9:208e14. [48] Matumura T, Kametani Y, Ando K, et al. Functional CD5þ B cells develop predominantly in the spleen of NOD/SCID/gcnull (NOG) mice transplanted either with human umbilical cord blood, bone marrow, or mobilized peripheral blood CD34þ cells. Exp Hematol 2003;31:789e97. [49] Watanabe S, Terashima K, Ohta S, et al. Hematopoietic stem cell-engrafted NOD/SCID/IL2Rgamma null mice develop human lymphoid systems and induce long-lasting HIV-1 infection with specific humoral immune responses. Blood 2007;109:212e8. [50] Lepus CM, Gibson TF, Gerber SA, et al. Comparison of human fetal liver, umbilical cord blood, and adult blood hematopoietic stem cell engraftment in NOD-scid/gammac-/-, Balb/c-Rag1-/-gammac-/-, and C.B-17-scid/bg immunodeficient mice. Hum Immunol 2009;70:790e802. [51] Watanabe Y, Takahashi T, Okajima A, et al. The analysis of the functions of human B and T cells in humanized NOD/shi-scid/gammac(null) (NOG) mice (hu-HSC NOG mice). Int Immunol 2009;21:843e58. [52] Biswas S, Chang H, Sarkis P, et al. Humoral immune responses in humanized BLT mice immunized with West Nile virus and HIV-1 envelope

[53]

[54]

[55]

[56] [57]

[58]

[59]

[60]

[61]

[62]

[63] [64]

[65]

[66] [67]

[68]

[69]

[70]

[71] [72]

proteins are largely mediated via human CD5þ B cells. Immunology 2011;134:419e33. Kametani Y, Shiina M, Katano I, et al. Development of human-human hybridoma from anti-Her-2 peptide-producing B cells in immunized NOG mouse. Exp Hematol 2006;34:1240e8. Suzuki M, Takahashi T, Katano I, et al. Induction of human humoral immune responses in a novel HLA-DR-expressing transgenic NOD/Shi-scid/gcnull mouse. Int Immunol 2012;24:243e52. Huang J, Li X, Coelho-dos-Reis JG, et al. Human immune system mice immunized with Plasmodium falciparum circumsporozoite protein induce protective human humoral immunity against malaria. J Immunol Methods 2015;427:42e50. McCune JM. Development and applications of the SCID-hu mouse model. Semin Immunol 1996;8:187e96. McCune JM, Namikawa R, Kaneshima H, et al. The SCID-hu mouse: murine model for the analysis of human hematolymphoid differentiation and function. Science 1988;241:1632e9. Aaberge IS, Michaelsen TE, Rolstad AK, et al. SCID-Hu mice immunized with a pneumococcal vaccine produce specific human antibodies and show increased resistance to infection. Infect Immun 1992;60:4146e53. Joo SY, Chung YS, Choi B, et al. Systemic human T cell developmental processes in humanized mice cotransplanted with human fetal thymus/liver tissue and hematopoietic stem cells. Transplantation 2012;94:1095e102. Schmitt K, Charlins P, Veselinovic M, et al. Zika viral infection and neutralizing human antibody response in a BLT humanized mouse model. Virology 2018;515:235e42. Jaiswal S, Smith K, Ramirez A, et al. Dengue virus infection induces broadly cross-reactive human IgM antibodies that recognize intact virions in humanized BLT-NSG mice. Exp Biol Med (Maywood) 2015;240:67e78. Fujii H, Luo ZJ, Kim HJ, et al. Humanized chronic graft-versus-host disease in NOD-SCID il2rg-/- (NSG) mice with G-CSF-mobilized peripheral blood mononuclear cells following cyclophosphamide and total body irradiation. PLoS One 2015;10:e0133216. Ito R, Katano I, Kawai K, et al. Highly sensitive model for xenogenic GVHD using severe immunodeficient NOG mice. Transplantation 2009;87:1654e8. Kametani Y, Katano I, Miyamoto A, et al. NOG-hIL-4-Tg, a new humanized mouse model for producing tumor antigen-specific IgG antibody by peptide vaccination. PLoS One 2017;12:e0179239. Gong J, Chehrazi-Raffle A, Reddi S, et al. Development of PD-1 and PD-L1 inhibitors as a form of cancer immunotherapy: a comprehensive review of registration trials and future considerations. J Immunother Cancer 2018;6:8. Postow MA, Sidlow R, Hellmann MD. Immune-related adverse events associated with immune checkpoint blockade. N Engl J Med 2018;378:158e68. Khoja L, Day D, Wei-Wu Chen T, et al. Tumour- and class-specific patterns of immune-related adverse events of immune checkpoint inhibitors: a systematic review. Ann Oncol 2017;28:2377e85. Sharpe AH, Wherry EJ, Ahmed R, et al. The function of programmed cell death 1 and its ligands in regulating autoimmunity and infection. Nat Immunol 2007;8:239e45. Ashizawa T, Iizuka A, Nonomura C, et al. Antitumor effect of programmed death-1 (PD-1) blockade in humanized the NOG-MHC double knockout mouse. Clin Cancer Res 2017;23:149e58. Abdel-Rahman O, Helbling D, Schmidt J, et al. Treatment-related death in cancer patients treated with immune checkpoint inhibitors: a systematic review and meta-analysis. Clin Oncol (R Coll Radiol) 2017;29:218e30. Arora N, Gupta A, Singh P. Biological agents in gastrointestinal cancers: adverse effects and their management. J Gastrointest Oncol 2017;8:485e98. Torino F, Barnabei A, Paragliola RM, et al. Endocrine side-effects of anti-cancer drugs: mAbs and pituitary dysfunction: clinical evidence and pathogenic hypotheses. Eur J Endocrinol 2013;169:R153e64.

Please cite this article in press as: Kametani Y, et al., Significance of humanized mouse models for evaluating humoral immune response against cancer vaccines, Personalized Medicine Universe (2018), https://doi.org/10.1016/j.pmu.2018.04.002