A Novel Process for the Expansion of Functional Human ILC2 Cells for GCHD Prevention

Recently, we have shown that type 2 mouse innate lymphoid cells (ILC2) ameliorate acute GVHD in a mouse model (Bruce et al, 2017). Moreover, it has been shown by us and others that not only is the recovery of ILC2 cells slow in both mouse and humans allogeneic recipients but also the presence of active human ILC2 cells in correlates with lower incidence of acute GVHD. Thus, we sought out to determine if human ILC2 cells could be enriched and expanded in vitro for potential use as a therapy to prevent or treat acute GVHD in humans. Toward this goal, we utilized human peripheral blood cells from apheresis units. We compared three methods to enrich ILC2 cells: RosetteSep kit, EasySep kit and EasySep NK56 kit. In the apheresis unit, ILC2 cells ranged from 0.023-0.062% of the total cells. In comparing the three enrichment methods, the percentage of enriched ILC2 cells varied from 1.1-6.9% with the EasySep ILC2 kit giving the best yields (6.9% and 4.7e4 total ILC2 cells). Interestingly, although the RosetteSep had the lowest percentage enriched, it had a higher yield than the EasySep NK Kit due to the higher yield of total cells (3.4 e6 compared to 4.3 e5 with 3.7 e4 and 2.2e4 ILC2 cells respectively). We next investigated optimal culture conditions. Our initial culture conditions were with α-MEM media with a cytokine cocktail of IL-2, IL-33 and IL-25. Using these conditions, the range of expansion was 500-1500 fold from the initial number of ILC2 at the end of the culture on Day 10. We tested a number of different interventions for optimization: addition of OP9-DL1 as a stromal cell component, addition of IL-7, and addition of TSLP. These results showed that OP9DL1, IL-7 or TSLP did not improve expansion or change the phenotype of the ILC2 cells at the end of the 10 day culture. We also tested different medias (α-MEM and IMDM) with different concentrations of the cytokines (based on the results by Camelo et al, 2017). We found that cells that were enriched using the RosetteSep and then cultured in α-MEM media with 50 ng/ml of IL-33, IL-25 and 20 IU/Ml IL-2 resulted in the best ILC2 expansion (>1000-fold) within 10 days. The final product was >90% pure ILC2 cells (lineage negative, CD161 and Gata-3 expression). These cells produced IL-13 and IL-5 in culture upon stimulation with PMA/ionomycin but did not produce IFN-gamma. Notably, between 50-80% of the ILC2 cells lost expression of CRTH2 (CD294) while in culture, similar to results found by Camelo et al. (2017). A representative experiment is shown in Figure 1.

We tested whether injection of human ILC2 cells ameliorated acute GVHD in a humanized mouse model. In this model, PBMCs are used as a positive control for GVHD. Two different approaches were taken. In the first approach we used third party hILC2 cells. On Day 0, irradiated female NSG mice received 15e6 PBMCs +/- 15e6 ILC2 cells. Mice were monitored for GVHD scores and survival. Recipients that received human ILC2 cells had significantly higher percentage mean initial body weight ratios at each time point beginning at Day 8 (P<0.05). At one month, 32% of controls (n=19) had succumbed to GVHD versus 12.5% (n=8) who had received third party human ILC2 cells. In the second approach we used donor-specific hILC2 cells, Irradiated (Day -1) male NSG mice were given 8e6 PBMC with or without donor matched ILC2 cells in 1:1 ratio on day 0. We found a single injection of matched ILC2 cells increased median survival by 10 days increasing survival from 19 days to 29 days for ILC2 treated animals (p<0.01). Taken together, we have discovered a novel method to rapidly expand human ILC2 cells in vitro. We have found that a single dose of hILC2 cells is active to decrease the severity of xenogeneic acute GVHD. These findings provide a rationale basis for the development of hILC2 cells to treat/prevent GI tract GvHD.

Figure 1. Apheresis cells were enriched using the ILC2 RosetteSep kit and were cultured in α-MEM media with 50 ng/ml of IL-33, IL-25 and 20 IU/Ml IL-2 for 10 days. The predominant population was ILC2 cells (A) that expressed GATA3 (B; ILC2 shown in red, negative control in blue) and upon stimulation with PMA and ionomycin expressed IL-13 but not IFN-gamma (C). The cells expanded >1000 fold in 10 days (D).

Figure 2. Survival of NSG mice following xeno-transplantation with PBMCs and donor matched expanded ILC2. Kaplan-Meier plot of survival following allo-SCT (n=4 per group) Log-rank (Mantel-cox) test used.

View largeDownload slide

View largeDownload slide

Close modal