Acute Graft versus host diseas (aGVHD) is a major complication of stem cell transplantation. The disease is mediated by T cells and a higher incidence/severity would be expected when higher numbers of T cells are inoculated. However, the incidence of aGVHD in PBST, which carries about 10 times more T cells then BMT, is not higher than the one found in later. This finding indicates a modulatory role for G-CSF over T cell activity. We had previously shown that T cells from G-CSF treated PBSC donors do not produce g-IFN nor IL-4 and that this inhibition was mediated by low density, G-CSF activated, granulocytes. In order to test if in fact G-CSF activated granulocytes could inhibit disease, we first checked if G-CSF could generate low density granulocytes, in vivo and in vitro. Indeed, either in vivo(21mg /day - 5 days) or in vitro (150 ng -12hs) with G-CSF generates low density granulocytes which co-purify with the mononuclear cells in the ficollĀ® gradient. Moreover, as we had shown in humans, these low density cells, inhibit the production of g-IFN by anti-CD3 activated T cells on flow cytometry studies (17%-T cells alone versus 3% T cells with granulocytes 1:1).

Radiation quimaeras were set with (B6 X BALB/c)F1 as hosts reconstituted with T cell depleted C57Bl6 bone marrow, in the presence or absence of nylon wool selected spleen cells (NWSC), as T cell source, from normal or G-CSF treated mice. As previously shown by others, NWSC from G-CSF treated mice diminishes the incidence of acute disease on day 20 post-transplant, from 75 to 25%. In order to investigate if this inhibition was dependent on the activated granulocytes present in the NWSC from G-CSF treated mice, granulocytes were depleted with anti-GR1 and complement. In this case, the incidence of disease is the same or even higher (75% experiment#1 and 100% in experiment #2) than the one observed on the control group (NWSC from control mice). These results strongly suggest that activated granulocytes could indeed inhibit aGVHD.

We then generated activated granulocytes in vitro, by treating spleen derived high density granulocytes with 150ng of G-CSF for 12 hs. After the incubation period, a new ficollĀ® gradient was performed and the low density cells were obtained. T cell contamination on the second gradient was eliminated by anti-CD4 and CD8 complement lysis. These activated granulocytes were inoculated together with NWSC from control mice in the radiation quimaeras at a 1:1 ratio. In this case 100% disease inhibition was observed when compared to the positive control group, where 75% of the animals got sick. Our data indicate that activated granulocytes are the major mediators of the G-CSF immunossupressive effects and that these cells can be used as a novel immune modulator in clinical transplantation to prevent acute GVHD.

Author notes

Corresponding author

Sign in via your Institution