@phdthesis{Uri2019,
  author    = {Uri, Anna},
  title     = {Differential requirement for CD28 co-stimulation on donor T cell subsets in mouse models of acute graft versus host disease and graft versus tumour effect},
  doi       = {10.25972/OPUS-16586},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-165863},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2019},
  abstract  = {Hematopoietic stem cell transplantation is a curative therapy for malignant diseases of the haematopoietic system. The patients first undergo chemotherapy or irradiation therapy which depletes the majority of tumour cells before they receive the transplant, consisting of haematopoietic stem cells and mature T cells from a healthy donor. The donor T cells kill malignant cells that have not been eliminated by the conditioning therapy (graft versus leukaemia effect, GvL), and, therefore, are crucially required to prevent relapse of the tumour. However, the donor T cells may also severely damage the patient's organs causing acute graft versus host disease (aGvHD). In mice, aGvHD can be prevented by interfering with the co-stimulatory CD28 signal on donor T cells. However, experimental models using conventional CD28 knockout mice as T cell donors or αCD28 antibodies have some disadvantages, i.e. impaired T cell development in the thymus of CD28 knockout mice and systemic CD28 blockade with αCD28 antibodies. Thus, it remains unclear how CD28 co-stimulation on different donor T cell subsets contributes to the GvL effect and aGvHD, respectively. We developed mouse models of aGvHD and the GvL effect that allowed to selectively delete CD28 on certain donor T cell populations or on all donor T cells. CD4+ conventional T cells (Tconv cells), regulatory T cells (Treg cells) or CD8+ T cells were isolated from either Tamoxifen-inducible CD28 knockout (iCD28KO) mice or their wild type (wt) littermates. Allogeneic recipient mice were then transplanted with T cell depleted bone marrow cells and different combinations of iCD28KO and wt T cell subsets. Tamoxifen treatment of the recipients caused irreversible CD28 deletion on the iCD28KO donor T cell population. In order to study the GvL response, BCL-1 tumour cells were injected into the mice shortly before transfer of the T cells. CD4+ Tconv mediated aGvHD was efficiently inhibited when wt Treg cells were co-transplanted. In contrast, after selective CD28 deletion on donor Treg cells, the mice developed a late and lethal flare of aGvHD, i.e. late-onset aGvHD. This was associated with a decline in iCD28KO Treg cell numbers around day 20 after transplantation. CD28 ablation on either donor CD4+ Tconv cells or CD8+ T cells reduced but did not abrogate aGvHD. Moreover, iCD28KO and wt CD8+ T cells were equally capable of killing allogeneic target cells in vivo and in vitro. Due to this sufficient anti-tumour activity of iCD28KO CD8+ T cells, they had a therapeutic effect in our GvL model and 25\% of the mice survived until the end of the experiment (day 120) without any sign of the malignant disease. Similarly, CD28 deletion on all donor T cells induced long-term survival. This was not the case when all donor T cells were isolated from wt donor mice. In contrast to the beneficial outcome after CD28 deletion on all donor T cells or only CD8+ T cells, selective CD28 deletion on donor CD4+ Tconv cells completely abrogated the GvL effect due to insufficient CD4+ T cell help from iCD28KO CD4+ Tconv cells. This study demonstrates that therapeutic inhibition of the co-stimulatory CD28 signal in either all donor T cells or only in CD8+ T cells might protect patients from aGvHD without increasing the risk of relapse of the underlying disease. Moreover, deletion of CD28 on donor Treg cells constitutes a mouse model of late-onset aGvHD which can be a useful tool in aGvHD research.},
  subject      = {Antigen CD28},
  language  = {en}
}
@phdthesis{RomerRoche2012,
  author    = {Romer Roche, Paula Sofia},
  title     = {Separation from self explains failure of circulating T-cells to respond to the CD28 superagonist TGN1412},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-74933},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2012},
  abstract  = {Stimulatory or superagonistic (SA) CD28-specific monoclonal antibodies (mAbs) are potent polyclonal activators of regulatory T cells and have proven highly effective as treatment in a wide range of rodent models for autoimmune and inflammatory diseases. In these models, a preferential activation of regulatory T cells was observed by in vivo administration of CD28SA. In stark contrast, human volunteers receiving TGN1412, a humanized CD28-specific mAb, experienced a life-threatening cytokine release syndrome during the first-in-man trial. Preclinical tests employing human peripheral blood mononuclear cells (PBMC) failed to announce the rapid cytokine release measured in the human volunteers in response to TGN1412. The aim of this thesis project was to find an explanation of why standard PBMC assays failed to predict the unexpected TGN1412-induced "cytokine storm" observed in human volunteers. CD28 superagonists can activate T cells without T cell receptor (TCR) ligation. They do depend, however, on "tonic" TCR signals received by MHC scanning, signals that they amplify. PBMC do not receive these signals in the circulation. Short-term in vitro preculture of human PBMC at a high cell density (HDC) resulted in massive cytokine release during subsequent TGN1412 stimulation. Restoration of reactivity was cell-contact dependent, associated with TCR polarization and tyrosine-phosphorylation, and blocked by HLA-specific mAb. In HDC, both CD4 T cells and monocytes functionally mature in a mutually dependent fashion. However, only CD4 memory T-cells proliferate upon TGN1412 stimulation, and were identified as the main source of pro-inflammatory cytokines. Importantly, responses to other T-cell activating agents were also enhanced if PBMC were first allowed to interact under tissue-like conditions. A new in vitro protocol is provided that returns circulating T-cells to a tissue-like status where they respond to TGN1412 stimulation, and it might represent a more reliable preclinical in vitro test for both activating and inhibitory immunomodulatory drugs. Finally, the surprising observation was made that the IgG1 "sibling" of TGN1412, which is of the poorly Fc receptor-binding IgG4 isotype, has a much lower stimulatory activity. We could exclude steric hindrance as an explanation and provide evidence for removal of TGN1112 from the T-cell surface by trans-endocytosis.},
  subject      = {T-Lymphozyten-Rezeptor},
  language  = {en}
}