SEARCH

SEARCH BY CITATION

Keywords:

  • graft-versus-host disease (GVHD);
  • immunodeficiency-primary;
  • MHC class II

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

Haematopoietic humanization of mice is used frequently to study the human immune system and its reaction upon experimental intervention. Immunocompromised non-obese diabetic (NOD)-Rag1–/– mice, additionally deficient for the common gamma chain of cytokine receptors (γc) (NOD-Rag1–/– γc–/– mice), lack B, T and natural killer (NK) cells and allow for efficient human peripheral mononuclear cell (PBMC) engraftment. However, a major experimental drawback for studies using these mice is the rapid onset of graft-versus-host disease (GVHD). In order to elucidate the contribution of the xenogenic murine major histocompatibility complex (MHC) class II in this context, we generated immunodeficient mice expressing human MHC class II [human leucocyte antigen (HLA)-DQ8] on a mouse class II-deficient background (Aβ–/–). We studied repopulation and onset of GVHD in these mouse strains following transplantation of DQ8 haplotype-matched human PBMCs. The presence of HLA class II promoted the repopulation rates significantly in these mice. Virtually all the engrafted cells were CD3+ T cells. The presence of HLA class II did not advance B cell engraftment, such that humoral immune responses were undetectable. However, the overall survival of DQ8-expressing mice was prolonged significantly compared to mice expressing mouse MHC class II molecules, and correlated with an increased time span until onset of GVHD. Our data thus demonstrate that this new mouse strain is useful to study GVHD, and the prolonged animal survival and engraftment rates make it superior for experimental intervention following PBMC engraftment.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

Mice functionally engrafted with human haematopoietic cells may represent a valuable preclinical tool for basic and applied research of the human immune system. Engraftment efficiencies of human cells, however, depend strongly upon the immunodeficiency status of the recipient mouse strain and its ability to foster the human donor cell survival and expansion. Early attempts to generate ‘humanizable’ immunodeficient mouse strains were based on mice with severe combined immunodeficiency (SCID) [1-3]. In these mice a mutation in the catalytic subunit of the DNA-dependent protein kinase (PRKDC) abrogates efficient V(D)J coding-joint formation, thus leading to T and B cell deficiency [4-6]. Similarly, Rag1- or Rag2-deficient mice lack T and B cells due to their inability to initiate V(D)J recombination [7, 8]. In contrast to T and B cells, natural killer (NK) cells are not affected in all these mice [9] and are thought to be responsible for frequent human haematopoietic cell transplant rejection due to the lack of mouse major histocompatibility complex (MHC) class I molecules on the transplanted human donor cells. The latter makes human donor cells susceptible to mouse NK cell recognition by the ‘missing self’ recognition mode [10]. Indeed, an improvement for xenogenic graft acceptance was achieved when these mice were bred to lack NK cells, most prominently by the introduction of common gamma chain of cytokine receptor (γc)-deficient alleles. This alteration resulted in high engraftment rates of human cells [11-15].

In parallel, mutations affecting T and B cells were transferred onto the non-obese diabetic background (NOD [16]), also resulting in improved human donor cell engraftment [17, 18]. This is due possibly to a lower level of NK cell activity in NOD mice [19]. Also, γc mutant allele(s) were bred onto the NOD background, finally resulting in NOD-SCID γc–/– [NOD/Shi-scid/interleukin (IL)-2Rγnull (NOG), NOD-SCID-γ (NSG)] and NOD-Rag1–/– γc–/– (NRG) mice, that allow for very high engraftment rates of human cells [18, 20, 21].

Apart from more elaborate techniques, involving simultaneous transplantation of human fetal thymus and liver tissues along with haematopoietic stem cells [bone marrow liver thymic (BLT) mice [22]]], humanization of NOG/NSG or NRG mice is performed generally using either haematopoietic stem cells or using mature peripheral blood cells (PBMCs). With respect to the latter, the transfer of human PBMCs (huPBMCs) into NOD-SCID, NOG/NSG or NRG mice triggers graft versus-host disease (GVHD) [23]. This disease is mediated by donor-derived human immune cells responding to xenogenic host antigens. In the clinic, GVHD is a frequently observed complication upon allogeneic stem cell transplantation. Thus, in principle, PBMC-humanized mice are an excellent model with which to evaluate therapeutic strategies to interfere with GVHD development. Unfortunately, however, while the PBMC transfer leads to high lymphocyte engraftment rates, the time-frame for experimental intervention and analysis is somewhat limited, as the xenogenic GVHD progresses rapidly. This complication caused the avoidance of this model to study the human immune system and its interaction with human pathogens such as Epstein–Barr virus (EBV) or human immunodeficiency virus (HIV) [24]. An extension of the time until acute GVHD occurs would therefore improve this animal model and would make it applicable for studies to manipulate GVHD or even allow host/pathogen interaction studies.

The principal host components responsible for the triggering of GVHD are the xenogenic mouse MHC class I and class II molecules. Studies with NSG mice lacking MHC class I (β2mnull) or MHC class II (Aβnull) showed that the deletion of MHC class II delayed disease progression significantly compared to NSG mice, but did not abrogate it. In contrast, MHC class I-deficient NSG mice were relatively resistant to GVHD development [25]. These data indicate that the recognition of murine MHC class I, presumably by CD8+ donor cells, constitutes the dominant effector pathway for GVHD; however, by recognition of murine MHC class II, CD4+ donor T cells appear to contribute significantly to mounting the xenogeneic GVHD.

In this study, we present newly generated mouse strains on the NRG background in which expression of murine MHC class II was abrogated and exchanged for the human HLA class II antigen DQ8 (NRG Aβ–/–DQ8 mice). This was achieved by intercrossing NRG with NOD.DQ8/Ab0 mice [26] that carry an Aβ-deficient allele [27] and that are transgenic for the human HLA class II molecule DQ8 [28]. Engraftment of the resulting mice with DQ8 haplotype-matched human donor PBMCs reduced host-directed xenogenic incompatibility and thus decreased GVHD development. Of note, this was observed despite the fact that CD8+ T cells would still react towards xenogenic MHC class I.

A major drawback of NOG/NSG or NRG mice is that adaptive immune responses are hardly inducible [18]. In haematopoietic stem cell-reconstituted mice expressing HLA class I, some of the mice showed HLA-A2-restricted CD8+ T cell responses upon infection with pathogens [29, 30]. Furthermore, it was reported that haematopoietic stem cell-reconstituted NRG mice expressing HLA class II (HLA-DR4) develop functional human T and B cells and the latter could even undergo immunoglobulin class switch recombination [31]. We speculated that DQ8 expression could also allow for the generation of serum immunoglobulins following PBMC reconstitution; we were therefore interested in testing the NRG Aβ–/–DQ8 mice concerning the onset of GVHD and their ability to engraft a functional human immune system with respect to T/B cell collaboration.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

Animals

Mice were kept in individually ventilated cages under barrier conditions on commercial mouse chow and water at the Paul-Ehrlich-Institut. For our experiments we used NRG (NOD.Cg-Rag1tm1Mom Il2rgtm1Wjl/SzJ) as a control and NRG Aβ–/–DQ8tg [NOD-Rag1tm1MomIl2rgtm1WjlH2-Ab1tm1DoiTg (HLA-DQA1, HLA-DQB1)1Dv] mice. They were established from breeders obtained from the Jackson Laboratory (Bar Harbor, ME, USA). The HLA transgene carries DQA*0301 and DQB*0302 alleles (see [28]; there termed NOD.DQ8). Experiments commenced when mice were aged 6–8 weeks without preconditioning. Mice were monitored daily for the onset of GVHD using body weight and visual examination parameters (based on hunched posture, ruffled hair, reduced mobility). Unless mentioned, experiments were conducted at least three times, resulting in a similar outcome. Euthanasia was performed when mice lost more than 20% of initial body weight. Experiments were performed in accordance with legal requirements.

Collection and transplantation of huPBMCs

Residual buffy coats from whole blood donations of healthy volunteers were obtained from the German Red Cross Blood donor Service Baden-Wuertemberg-Hessen, Frankfurt. PBMC were purified from buffy coats by Ficoll-Hypaque density centrifugation and suspended in phosphate-buffered saline (PBS) for intravenous (i.v.) injection of 5 × 107 cells/mouse.

HLA-DQ8 genotyping of PBMC donors

Donor DNA was extracted from blood using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany) and used for genotyping. HLA-DQ8-positive individuals were identified by polymerase chain reaction (PCR) using the Olerup SSP HLA-DQB1*03 Kit (Olerup, Vienna, Austria).

Antibodies and flow cytometry

All antibodies were obtained from BD Biosciences (Heidelberg, Germany): anti-human (huCD45)-phycoerythrin (PE) (clone H3.7), anti-huCD3-allophycocyanin (APC) (clone H5.2), anti-huCD4-APC-cyanin-7 (Cy7) (clone H13.2), anti-huCD8-PE-Cy7 (clone H11.1), anti-huCD19-PE-Cy5 (clone H4.5), anti-huCD56-PE-CY5 (clone H4.4), anti-huCD5-APC (clone H5.4), anti-huCD14-Pacific Blue (clone H12.1) and anti-mouse CD45-fluorescein isothiocyanate (FITC) (clone 30F11). Blood drawn from the retro-orbital sinus (20 μl) was collected into ethylenediamine tetraacetic acid (EDTA)-coated tubes (BD Biosciences). Blood was incubated for 20 min at room temperature (RT) with anti-CD16/32 antibody to block non-specific Fc-receptor-mediated binding. Antibodies were incubated for 15 min at 4°C at the appropriate dilution as determined by previous titration. Erythrocytes were lysed with fluorescence activated cell sorter (FACS) lysing solution (BD Biosciences), according to the manufacturer's protocol. Then, cells were washed with FACS buffer and fixed with 1% paraformaldehyde (Fluka Chemica, Taufkirchen, Germany) in PBS. At least 10 000 events were acquired with an LSRII instrument (BD Biosciences) and analysed using FACS Diva Software. In addition to the human markers, for all analyses anti-mouse CD45 staining was included to allow for the exclusion of all murine haematopoietic cells. Human PBMCs from buffy coats were isolated as described and used as positive staining control. Matching isotype control antibodies were used as negative controls.

Histopathological analyses/immunohistochemistry

Tissues were recovered from mice at necropsy, fixed in 4% formalin and processed for (immuno-)histology. Briefly, organs were embedded in paraffin, cut into 2 μm sections, deparaffinized and then stained with either haematoxylin (Merck, Darmstadt, Germany) or anti-CD8 (GeneTex, Eching, Germany) and TrueBlue (KPL, Wedel, Germany). Sections were analysed using an Axiophot microscope (Zeiss, Göttingen, Germany, ×10 magnification) and Axiovision software for analysis.

Statistical analysis

All statistical analyses were performed using Prism GraphPad software (San Diego, CA, USA). Analysis of variance (anova) test for the area under the curve in Fig. 1 was performed with sas®/stat software (version 9.3, SAS System for Windows). Student's t-test was used for statistical analyses unless noted otherwise. In general, means were used and statistical deviations are presented as standard deviation unless noted otherwise. A P-value < 0·05 was deemed statistically significant.

figure

Figure 1. Repopulation kinetic of human (hu)CD45+ cells in the peripheral blood of human peripheral blood mononuclear cells (huPBMC)-DQ8 repopulated mice. NOD-Rag1–/– γc–/– (NRG) and NRG Aβ–/–DQ8tg mice were transplanted adoptively with 5 × 107 PBMC from DQ8+ donors (huPBMC-DQ8) intravenously. Engraftment of each animal was evaluated throughout the entire duration of the experiment by analysing cells from the peripheral blood by flow cytometry. Bars represent the mean engraftment of all live mice in the group. Groups were compared by means of an analysis of variance (anova) for the area under the curve (AUC), weighted for the observation period (day 5 until euthanasia of the mouse) with a significance of P-value: 0·0014. Some mice of the NRG Aβ–/– DQtg group survived until day 40, which may influence the difference between the groups. To account for this effect, an additional analysis containing the data until day 21 only was performed and gave consistent results that were statistically significant (P-value: 0·0294).

Download figure to PowerPoint

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

Humanized NRG Aβ–/–DQ8 transgenic mice show enhanced repopulation efficiency when reconstituted with matching human PBMCs

The effect of HLA class II on the engraftment efficiency of haplotype-matched human PBMCs in recipient mice lacking T, B and NK cells was studied by comparing the engraftment of human CD45+ lymphocytes in NRG Aβ–/–DQ8 recipient mice to that of conventional NRG mice, the latter expressing mouse MHC class II. Repopulation was monitored following the adaptive transfer of 5 × 107 DQ8-positive huPBMCs (huPBMC-DQ8) i.v. This dose was chosen to ensure high repopulation efficiencies of NRG mice [25]. Human lymphocytes were monitored in the peripheral blood as human CD45+ cells (Fig. 1). Similar to published data, the percentage of human cells increased quickly within the first 9–12 days following huPBMC-DQ8 injection [25]. NRG mice possessed engraftment rates of up to 55% human CD45 cells, whereas NRG Aβ–/–DQ8tg mice showed higher engraftment rates of up to 80% human CD45+ cells. Interestingly, the repopulation kinetics, rather than the repopulation efficiency, between the two mouse strains did not differ. NRG Aβ–/–DQ8tg mice showed an enhanced number of human CD45+ cells compared to NRG mice (Fig. 1, days 16–21). This observation was significant (P = 0·0294) when tested by anova until day 21 after transfer of PBMCs, when NRG mice had to be euthanized due to GVHD severity (cp. Fig. 4). It appears that NRG Aβ–/–DQ8tg mice tolerated huPBMCs-DQ8 better than did NRG mice.

Human T cells are expanded selectively in humanized mice

Next, the distribution of human haematopoietic cell subtypes in the repopulated mice was analysed and compared to that in the donor inoculum. The analysis shown in Fig. 2 was performed 5 days after repopulation and represents data for one individual mouse, representative of the entire group. Mice were repopulated with huPBMC-DQ8, containing 40% CD3+ T cells, 9% CD19+ B cells, 5% CD56+ NK cells and 6% CD14+ monocytes/macrophages. One week after repopulation, no difference was detectable between NRG and NRG Aβ–/–DQ8tg recipient mice. In both strains, more murine CD45+ cells (muCD45 > 80%) than huCD45+ cells were present. As shown in Fig. 1, huCD45+ cells increased throughout the experiment, while muCD45+ cells decreased correspondingly (data not shown). Detailed analysis demonstrated that huCD45+ cells in NRG as well as NRG Aβ–/–DQ8tg mice consist mainly of CD3+ T cells (>98%). Other human immune cells such as NK cells (CD56+), monocytes (CD14+) or B cell types (CD5-CD19+, CD5+CD19+) could not be detected in either strain even at the earliest time-point (day 3) (data not shown), although these subtypes were present among the donor huPBMC-DQ8 cells. Thus, human T cells repopulate both strains selectively.

figure

Figure 2. Human peripheral blood mononuclear cells (PBMC) repopulation of recipient mice. Donor blood cells were analysed by flow cytometry before the isolation of mononuclear cells (top row) or following adoptive transfer as peripheral blood cells, present on day 5 after repopulation. Data from one individual animal, representative of the indicated groups, are shown.

Download figure to PowerPoint

Humanized NRG Aβ–/–DQ8tg mice show delayed onset of GVHD

Engraftment of huPBMC into NRG mice results in the development of GVHD soon after transplantation [12]. Hence, NRG and NRG Aβ–/–DQ8tg mice repopulated with haplotype-matched huPBMC-DQ8 were monitored over time for signs of disease by determining individual disease scores [32]. Disease symptoms scored were hunched posture, ruffled hair and reduced mobility, ranked according to severity. Figure 3a shows disease scores over time of individual mice following their repopulation. Seven days after repopulation, NRG mice showed the first signs of disease while NRG Aβ–/–DQ8tg mice demonstrate such only from day 9 onwards. Furthermore, NRG mice progress rapidly from initial symptoms to severe GVHD disease (score > 3) within 12–19 days after transfer, whereas NRG Aβ–/–DQ8tg mice never reached a clinical score of >3 before day 28 after transfer (except one animal that had already scored 3 at day 14; however, this mouse was considerably smaller than all other mice). The progress of disease also correlated with weight loss of the individual animals. Figure 3b presents a parameter for each mouse in the group that indicates the weight loss linked to the time in the experiment. Weight loss was significantly different among the strains (P = 0·0018), with NRG mice having lost more weight (mean parameter 4·8) compared to NRG Aβ–/–DQ8tg mice (mean parameter 3·0).

figure

Figure 3. Graft-versus-host disease (GVHD)in peripheral blood mononuclear cells (PBMC) humanized mice. (a) A clinical scoring system (as per [32], but excluding weight, see M + M) was used to follow the course of GVHD. Animals were graded at the time-points indicated. NOD-Rag1–/– γc–/– (NRG) Aβ–/–DQ8tg mice are shown in green; NRG mice in black. (b) For each individual animal, as represented by individual symbols, a ‘parameter weight’ was calculated where the difference of the initial weight and the weight at the last day of the experiment was divided by the time in the experiment (in days). The difference between the groups is significant (P = 0·0018). (c) Alanine transaminase (ALT) levels of each individual mouse in U/l at the end of the experiment. The differences were significant (P-value: 0·0150).

Download figure to PowerPoint

figure

Figure 4. Survival of the mice. NOD-Rag1–/– γc–/– (NRG) and NRG Aβ–/–DQ8tg mice were repopulated with 5 × 107 human peripheral blood mononuclear cells (PBMC)-DQ8. The difference between the groups was significant (P = 0·0012).

Download figure to PowerPoint

Apart from external signs of disease and weight loss, the pathology caused by GVHD usually becomes evident in organs such as liver, intestine, kidney and skin. A very convenient diagnostic parameter is the presence of the liver-specific enzyme alanine transferase (ALT) in the serum, occurring when there is liver damage. Accordingly, to monitor the progress and severity of GVHD development, we analysed the serum of all mice for ALT activity (Fig. 3c). Strikingly, there was only a mild increase of ALT (mean: 200 U/l) in NRG Aβ–/–DQ8tg recipients, while NRG recipients showed a much higher concentration of ALT (mean: 1300 U/l) compared to non-humanized mice (non-hu; mean: 120 U/l). This indicates a more advanced progress of GVHD in NRG mice compared to NRG Aβ–/–DQ8tg mice following their repopulation with DQ8-matched PBMCs. These data suggest a survival advantage of HLA class II-matched mice over those expressing xenogenic murine MHC class II.

Essentially, the disease score and weight loss are a reflection of the ongoing GVHD leading eventually to death. In this study, a weight loss of more than 20%, compared to the initial weight and independent of other symptoms, required us to euthanize the animals by statutory order and was taken as the end of survival. Indeed, NRG Aβ–/–DQ8tg mice survived significantly longer (mean survival 28·5 days) after huPBMC-DQ8 engraftment than do NRG mice (mean survival 17 days) (Fig. 4). Thus, although NRG Aβ–/–DQ8tg mice repopulated to a higher level, the onset of disease symptoms and development of fetal GVHD disease was delayed.

Increased human CD8 T cells in NRG mice when GVHD commenced

Both human CD4+ and CD8+ T cells have been shown to contribute to GVHD development in murine recipients [25]. Adoptive transfer of NRG Aβ–/–DQ8tg mice with DQ8-matched donor PBMCs represents, with respect to HLA-DQ8, an HLA-class II-matched transplantation which should alleviate CD4+ T cell-mediated GVHD. In contrast, donor CD8+ T cells still face xenogenic MHC class I in both recipient mouse strains. Thus, it was interesting to determine whether the GvHD, mounting more slowly in NRG Aβ–/–DQ8tg recipients, could be correlated with differences in donor T cell subsets repopulating the two strains. While exclusively human CD3+ T cells accumulated in both strains, there was no difference between strains with regard to human CD4+ or CD8+ T cells at an early time-point after repopulation (Fig. 5, day 5). However, from day 9 after repopulation onwards, the contribution of human CD8+ T cells among CD3+ cells increased specifically in NRG mice, such that by day 14 the CD8+ T cells increased twice as much compared to day 5 (60 versus 30%, respectively). Such a dramatic shift towards CD8+ T cells did not occur in NRG Aβ–/–DQ8tg mice receiving the same DQ8+ donor PBMCs. In essence, the ratio of human CD4+ and CD8+ T cells reversed within 14 days after repopulation of NRG mice, but remained relatively stable in NRG Aβ–/–DQ8tg recipients. It is concluded that the expansion of human CD8+ T cells is an early sign of xenogenic GVHD.

figure

Figure 5. Repopulation by CD4+ and CD8+ T cells at different time-points following adoptive human peripheral blood mononuclear cells (huPBMC)-DQ8 transfer. The engraftment by huPBMC-DQ8 was monitored with respect to human CD4+ and CD8+ T cells by flow cytometry on the time-points indicated following transfer. Each symbol represents an individual mouse. Statistically significant differences are noted in the figure.

Download figure to PowerPoint

Human CD8+ T cell infiltration into organs is increased in NRG recipients

As we found that human CD8+ T cells are a population expanding at an early time when GVHD develops in NRG mice, we asked whether these T cells are responsible for the liver damage, detected as an increased in serum ALT levels (see Fig. 3c). Therefore, we analysed liver sections by immunohistochemical staining (IHC) for human CD8 (Fig. 6a). A massive, high-grade infiltration by mononuclear cells, many being CD8+ and spreading into the peripheral liver parenchyma, is seen in NRG recipients (Fig. 6a, bottom panels). In some sections, single hepatocytes were found to be necrotic: a hallmark for ongoing liver injury. In contrast to the NRG mice, infiltrates were less pronounced in NRG Aβ–/–DQ8tg mice, also showing far fewer CD8+ T cells (Fig. 6a). Non-humanized mice (non-hu) showed no infiltrates (Fig. 6a, top panels). The skin is a further organ affected typically by GVHD. In both mouse strains we observed macroscopically alterations of skin texture such as hyperkeratosis, scleroderma and desquamation, as used for clinical score grading. As expected, histological examination confirmed these observations. The skin surface appeared undulated and signs of fibrosis, folliculitis and steatitis were evident within the hypodermis [see arrows in Fig. 6, haematoxylin and eosin (H&E) staining]. Notably, these observations tended to be more severe in NRG control mice compared to NRG Aβ–/–DQ8tg mice. As GVHD is a systemic disease, we consequently also detected huCD8 T cells in other organs, such as kidney and intestine. Again, infiltrates were less pronounced in NRG Aβ–/–DQ8tg mice compared to NRG mice (Fig. 6a).

figure

Figure 6. Human CD8+ T cells infiltrating organs. (a) Sections from liver, kidney, intestine and skin of human peripheral blood mononuclear cell (huPBMC)-DQ8 transplanted mice, taken at the end of the experiment, were examined by haematoxylin and eosin staining (H&E, left panels) as well as immunohistochemistry (IHC) for human CD8 cells (in blue, right panels). Cell infiltrates are indicated by an arrow. Genotypes of the recipient mice are indicated. As reference, one non-humanized mouse is included (non-Hu). (b) Cellular infiltrates in the liver, kidney, intestine and skin were scored with respect to CD8+ cells, as identified by IHC as in (a) (scale range of 0–3, see [33]) and data are summarized graphically. Bars representing the mean level of infiltration per organ and recipient mouse group and the corresponding standard deviation are shown. Differences were significant for liver (P = 0·0099), intestine (P = 0·0112) and kidney (P = 0·0467), but not for skin (P = 0·7431).

Download figure to PowerPoint

To quantify the huCD8+ cell infiltrates we used a published score [33]. Livers of NRG mice exhibited a significantly higher infiltration by human CD8+ T cells (mean score: 2·15) compared to those of NRG Aβ–/–DQ8tg mice (mean score: 1·36). In addition, kidneys and intestines of NRG mice were also infiltrated more severely by huCD8+ cells (mean score: 1·05 and 1·00, respectively) compared to NRG Aβ–/–DQ8tg mice (mean score: 0·58 and 0·42, respectively). This tendency of a more pronounced infiltration in NRG mice was also seen for the skin, although the difference was not statistically significant (mean score: 1·45 versus 1·33 in NRG versus NRG Aβ–/–DQ8tg mice, respectively). Taken together, the delayed onset and mild progression of GVHD in NRG Aβ–/–DQ8tg mice could be due to a delay in the activation and expansion of xenoreactive CD8+ cells.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

In this study, we examined the effect of replacing murine MHC class II by HLA class II (DQ8) on the development of GVHD upon adoptive transfer of DQ8-positive human PBMCs into immunodeficient recipient mice (NRG Aβ–/–DQ8tg versus conventional NRG mice). The presence of HLA-DQ8 in NRG Aβ–/–DQ8tg recipient mice augmented significantly the overall repopulation rate by human PMBCs compared to conventional NRG mice. The cellular subset capable of engraftment was skewed exclusively towards CD3+ T cells in both mouse strains. Despite this, the striking difference between the two strains was the time-frame until GVHD became fatal. While repopulation of NRG mice with xenogeneic human DQ8-PBMCs resulted in rapid induction of GVHD and poor survival of the animals, HLA-class II-matched transfer of PBMCs into NRG Aβ–/–DQ8tg recipients resulted in a milder form of GVHD, such that the mice survived significantly longer. Indeed, liver destruction, as measured by serum ALT level, was less pronounced in NRG Aβ–/–DQ8tg recipients compared to that seen in NRG mice. This observed liver destruction correlated with huCD8+ T cell infiltration into the liver. Similarly, as expected for a systemic disease, huCD8+ T cells were also prominent in other organs such as kidney, intestine and skin. The delayed onset and mild progression of GVHD in the haplotype-matched recipients corresponded to the delay in the expansion of human CD8+ cells, most probably reacting towards the xenogeneic murine MHC class I.

Mechanistically, two scenarios can be envisioned for the reason that NRG Aβ–/–DQ8tg mice develop an attenuated form of GVHD only. Clearly, these scenarios must account for the fact that xenoreactive CD8+ T cells are apparently activated less efficiently in the DQ8 mice, despite having changed the xenoreactive recognition for class II MHC only, while xenogenic class I is still present. One explanation could be that the introduction of DQ8 and removal of murine class II reduced the frequency and thus the helper-activity of xenoreactive CD4+ T cells. This would be expected, as upon HLA class II being matched, the frequency of CD4+ T cells being activated would be much smaller than when confronted by xenogenic murine class II. In the NRG Aβ–/–DQ8tg recipients the CD4+ T cells would thus recognize murine peptides presented by DQ8, and this situation would mimic a class II-matched scenario where CD4+ T cells would react solely towards murine ‘minor histocompatibility antigens’. The lower frequency of activated CD4+ T cells may then not suffice to allow for an efficient mounting of the xenoreactive response of CD8+ T cells. Alternatively, upon the presence of DQ8, regulatory CD4+ T cells present in the donor inoculum may be induced due to their ability to interact with their restricting HLA class II, DQ8. In this way they could, initially, keep the GVHD-mediating T cells under control. However, it is unclear whether reactivity towards xenogenic class II versus matched class II, but presenting a multitude of foreign murine peptides as disparate minor histocompatibility antigens would favour preferentially either conventional CD4+ T helper or regulatory T cells in the transfer setting probed in this study.

Human interferon gamma (IFN-γ) levels in the serum of recipient mice were elevated shortly after the transfer of DQ8-PBMCs. This was equally true for both NRG and NRG Aβ–/–DQ8tg strains, and IFN-γ levels remained unaltered throughout the experiment (data not shown). These data favour a scenario in which the xenoreactive CD8+ T cell activation is responsible for the fatal GVHD induction in both strains, but due to class II haplotype matching changing the quality or quantity of the CD4+ T cell response, the xenoreactive CD8+ T cells take longer to mount their response in the DQ8-matched recipients. A clear mechanistic solution will require further analyses with respect to CD4+ T cell subsets, their cytokine profile and analyses with respect to the regulatory capability of transferred CD4+ regulatory T cells (Treg).

Unfortunately, HLA class II expression did not enable the PBMC-DQ8 transplanted DQ8 mice to develop a humoral immune response, which requires the collaboration of T with B cells (data not shown). Because B cells did not survive in this adoptive PBMC model, this is expected. Also, it limits the usefulness of this model for testing purposes, such as testing vaccines for which a humoral immune response has been shown to be essential to mediate protection. If a regimen can be established allowing for preservation of the B cell subset, it will be an interesting retest for this immune function. Currently, however, even though tests relying upon humoral responses are not possible, this does not mean that CD4+ T cell responses are not occurring. Thus, direct assays for CD4+ T cell function, such as lymphokine production in response to test antigens, could well be possible. It could also allow testing of whether a donor was primed to the given antigen, and thus became immune, during the testing of new therapeutic vaccines relying upon a cellular immune reaction. This mouse model could provide a personalized animal model to test vaccine efficacies in vivo. Potentially, the transfer of PBMCs of vaccinated people followed by a challenge infection in the mouse could provide indications of the effectiveness of cell-mediated vaccines. In this respect, the mouse model described in this study could be of considerable value for human immunodeficiency virus (HIV) vaccine testing, as HIV has a very limited host tropism and replicates almost exclusively in human CD4+ T cells.

Finally, NRG Aβ–/–DQ8tg mice are a useful model to test experimentally for modalities reducing GVHD in partially allogeneic or minor histocompatibility disparate settings. Similar to recently published data, the engraftment could be limited to CD4+ cells to focus upon the contribution to GVHD by these cells [33]. A further refinement would be to cross NRG Aβ–/– DQ8tg with MHC class I knock-out mice. In these, the CD8+-mediated component of GVHD would be eliminated, and this could make the mice suitable even for long-term studies.

Overall, this newly generated mouse strain shows prolonged survival and delayed onset of GVHD after transplantation with haplotype-matched human PBMCs. Thus, it is a superior model with which to study GVHD, and it could be valuable to investigate CD4+ T cell responses for certain human vaccines and pathogens.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References

We thank Heike Baumann, Christine von Rhein and Sophie Wald for excellent technical assistance and Kay-Martin Hanschmann for help with statistical analysis. This work was funded in part by the German Federal Ministry of Health.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure
  9. References
  • 1
    Bosma GC, Custer RP, Bosma MJ. A severe combined immunodeficiency mutation in the mouse. Nature 1983; 301:527530.
  • 2
    Lapidot T, Pflumio F, Doedens M, Murdoch B, Williams DE, Dick JE. Cytokine stimulation of multilineage hematopoiesis from immature human cells engrafted in SCID mice. Science 1992; 255:11371141.
  • 3
    Mosier DE, Gulizia RJ, Baird SM, Wilson DB. Transfer of a functional human immune system to mice with severe combined immunodeficiency. Nature 1988; 335:256259.
  • 4
    Lieber MR, Hesse JE, Lewis S et al. The defect in murine severe combined immune deficiency: joining of signal sequences but not coding segments in V(D)J recombination. Cell 1988; 55:716.
  • 5
    Blunt T, Finnie NJ, Taccioli GE et al. Defective DNA-dependent protein kinase activity is linked to V(D)J recombination and DNA repair defects associated with the murine scid mutation. Cell 1995; 80:813823.
  • 6
    Blunt T, Gell D, Fox M et al. Identification of a nonsense mutation in the carboxyl-terminal region of DNA-dependent protein kinase catalytic subunit in the scid mouse. Proc Natl Acad Sci USA 1996; 93:1028510290.
  • 7
    Mombaerts P, Iacomini J, Johnson RS, Herrup K, Tonegawa S, Papaioannou VE. RAG-1-deficient mice have no mature B and T lymphocytes. Cell 1992; 68:869877.
  • 8
    Shinkai Y, Koyasu S, Nakayama K et al. Restoration of T cell development in RAG-2-deficient mice by functional TCR transgenes. Science 1993; 259:822825.
  • 9
    Dorshkind K, Pollack SB, Bosma MJ, Phillips RA. Natural killer (NK) cells are present in mice with severe combined immunodeficiency (SCID). J Immunol 1985; 134:37983801.
  • 10
    Ljunggren HG, Karre K. In search of the ‘missing self’: MHC molecules and NK cell recognition. Immunol Today 1990; 11:237244.
  • 11
    Kirberg J, Berns A, von Boehmer H. Peripheral T cell survival requires continual ligation of the T cell receptor to major histocompatibility complex-encoded molecules. J Exp Med 1997; 186:12691275.
  • 12
    van Rijn RS, Simonetti ER, Hagenbeek A et al. A new xenograft model for graft-versus-host disease by intravenous transfer of human peripheral blood mononuclear cells in RAG2–/– gammac–/– double-mutant mice. Blood 2003; 102:25222531.
  • 13
    Gimeno R, Weijer K, Voordouw A et al. Monitoring the effect of gene silencing by RNA interference in human CD34+ cells injected into newborn RAG2–/– gammac–/– mice: functional inactivation of p53 in developing T cells. Blood 2004; 104:38863893.
  • 14
    Legrand N, Cupedo T, van Lent AU et al. Transient accumulation of human mature thymocytes and regulatory T cells with CD28 superagonist in ‘human immune system’ Rag2(–/–)gammac(–/–) mice. Blood 2006; 108:238245.
  • 15
    Legrand N, Weijer K, Spits H. Experimental models to study development and function of the human immune system in vivo. J Immunol 2006; 176:20532058.
  • 16
    Makino S, Kunimoto K, Muraoka Y, Mizushima Y, Katagiri K, Tochino Y. Breeding of a non-obese, diabetic strain of mice. Jikken Dobutsu 1980; 29:113.
  • 17
    Hudson WA, Li Q, Le C, Kersey JH. Xenotransplantation of human lymphoid malignancies is optimized in mice with multiple immunologic defects. Leukemia 1998; 12:20292033.
  • 18
    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; 21:843858.
  • 19
    Ogasawara K, Hamerman JA, Hsin H et al. Impairment of NK cell function by NKG2D modulation in NOD mice. Immunity 2003; 18:4151.
  • 20
    Matsumura T, Kametani Y, Ando K et al. Functional CD5+ B cells develop predominantly in the spleen of NOD/SCID/gammac(null) (NOG) mice transplanted either with human umbilical cord blood, bone marrow, or mobilized peripheral blood CD34+ cells. Exp Hematol 2003; 31:789797.
  • 21
    Ito M, Kobayashi K, Nakahata T. NOD/Shi-scid IL2rgamma(null) (NOG) mice more appropriate for humanized mouse models. Curr Top Microbiol Immunol 2008; 324:5376.
  • 22
    Wege AK, Melkus MW, Denton PW, Estes JD, Garcia JV. Functional and phenotypic characterization of the humanized BLT mouse model. Curr Top Microbiol Immunol 2008; 324:149165.
  • 23
    Ito R, Katano I, Kawai K et al. Highly sensitive model for xenogenic GVHD using severe immunodeficient NOG mice. Transplantation 2009; 87:16541658.
  • 24
    Imadome K, Yajima M, Arai A et al. Novel mouse xenograft models reveal a critical role of CD4+ T cells in the proliferation of EBV-infected T and NK cells. PLoS Pathog 2011; 7:e1002326.
  • 25
    King MA, Covassin L, Brehm MA et al. Human peripheral blood leucocyte non-obese diabetic-severe combined immunodeficiency interleukin-2 receptor gamma chain gene mouse model of xenogeneic graft-versus-host-like disease and the role of host major histocompatibility complex. Clin Exp Immunol 2009; 157:104118.
  • 26
    Hayward SL, Bautista-Lopez N, Suzuki K, Atrazhev A, Dickie P, Elliott JF. CD4 T cells play major effector role and CD8 T cells initiating role in spontaneous autoimmune myocarditis of HLA-DQ8 transgenic IAb knockout nonobese diabetic mice. J Immunol 2006; 176:77157725.
  • 27
    Cosgrove D, Gray D, Dierich A et al. Mice lacking MHC class II molecules. Cell 1991; 66:10511066.
  • 28
    Nabozny GH, Baisch JM, Cheng S et al. HLA-DQ8 transgenic mice are highly susceptible to collagen-induced arthritis: a novel model for human polyarthritis. J Exp Med 1996; 183:2737.
  • 29
    Shultz LD, Saito Y, Najima Y et al. Generation of functional human T-cell subsets with HLA-restricted immune responses in HLA class I expressing NOD/SCID/IL2r gamma(null) humanized mice. Proc Natl Acad Sci USA 2010; 107:1302213027.
  • 30
    Jaiswal S, Pearson T, Friberg H et al. Dengue virus infection and virus-specific HLA-A2 restricted immune responses in humanized NOD-SCID IL2rgammanull mice. PLoS ONE 2009; 4:e7251.
  • 31
    Danner R, Chaudhari SN, Rosenberger J et al. Expression of HLA class II molecules in humanized NOD.Rag1KO.IL2RgcKO mice is critical for development and function of human T and B cells. PLoS ONE 2011; 6:e19826.
  • 32
    Cooke KR, Hill GR, Crawford JM et al. Tumor necrosis factor-alpha production to lipopolysaccharide stimulation by donor cells predicts the severity of experimental acute graft-versus-host disease. J Clin Invest 1998; 102:18821891.
  • 33
    Covassin L, Laning J, Abdi R et al. Human peripheral blood CD4 T cell-engrafted non-obese diabetic-scid IL2rgamma(null) H2-Ab1 (tm1Gru) Tg (human leucocyte antigen d-related 4) mice: a mouse model of human allogeneic graft-versus-host disease. Clin Exp Immunol 2011; 166:269280.