Abstract 2448

Poster Board II-425

Donor lymphocyte graft engineering to avoid graft-versus-host (GVH) reactivity while improving graft-versus-leukemia (GVL) immunity remains of central interest in allogeneic hematopoietic stem cell transplantation (HSCT). However, appropriate models to evaluate experimental concepts of donor lymphocyte allograft engineering in vivo are missing. We, therefore, established a human-murine chimeric transplantation model using immunodeficient NOD/SCID/IL2Rγcnull (NSG) mice to evaluate GVH reactivity of human T cell grafts in vivo. Moreover, since mechanisms of immune functions resembling human GVH immunity have not yet been addressed in detail in these humanized mice we started to analyse T cell trafficking and homing to lymphoid tissues across species barrier to examine the clinical relevance of T cell activity observed in this model.

To this end, skin substitutes composed of human primary allogeneic fibroblasts embedded in a collagen-based matrix were subcutaneously implanted into NSG mice to detect alloreactive specificities within the implant post adoptive transfer of MHC-mismatched or haploidentical donor T lymphocytes. The skin substitutes revealed murine vascularisation two weeks after implantation as demonstrated by immunohistological studies. Following transfer of human HLA-mismatched or haploidentical T lymphocytes, up to 23% of the T cells migrated into skin substitutes explanted 21 days post injection. As this T cell migration and homing involves both murine and human adhesion molecules we further analyzed the specific adhesion mechanisms underlying the egress of human T cells from the murine bloodstream.

Using laminar flow chamber experiments and real time video recordings we could first demonstrate that human anti-CD3/anti-CD28 preactivated T lymphocytes but not naive T cells bound and firmly adhered to the murine endothelial cell line bEND.3 (BEND3.EC) at shear stresses of up to 3.5 dynes/cm2. As controls, human umbilical vein endothelial cells (HUVEC) were used. Adhesion and transmigration was significally enhanced when both human and murine endothelial cells (ECs) were prestimulated with low doses of TNF-α (5-20ng/ml) to resemble an activated phenotype. Firm adhesion of activated T lymphocytes was suppressed following pretreatment with function-blocking anti-integrin-alpha 4 (CD49d, subunit of VLA-4) or anti-integrin-alpha L (CD11a, subunit of LFA-1) antibody (Ab) or when ECs were preincubated with anti-VCAM-1 (CD106). No inhibitory effects were observed when anti-Endoglin (CD105) Ab was included as specificity control suggesting that the integrin dimer VLA-4 and its counter-receptor VCAM-1 as well as the integrin dimer LFA-1 are required for the transmigration of human T cells across murine ECs. Primary ECs derived from murine aorta are currently used to confirm our results obtained with bEND.3 cells. As the same ligand-receptor pairs are described for human T-EC interaction these findings indicate a closely related mechanism of T cell extravasation in human and murine endothelium at least in our transplantation model.

In conclusion these results suggest that intravenously transferred activated human T cells migrate into allogeneic skin substitutes involving VCAM-1 and integrin-alpha-4 for firm adhesion followed by transmigration. In vivo studies investigating the effects of the function-blocking antibodies against VCAM-1 and integrin-alpha-4 in our model to confirm the in vitro results are in progress and will be reported. In addition, our human-murine chimeric NSG transplantation model may represent a promising tool to study human GVH biology and to evaluate T cell graft engineering in allogeneic HSCT.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.