Effect of Antioxidant N-Acetylcysteine Treatment on the Phenotype of the TCL-1 Transgenic Mouse Model of B-CLL through Reactive Oxygen Species (ROS) Modulation

Abstract 3581

Thioredoxin-interacting protein (TXNIP) binds and inactivates thioredoxin (TRX) leading to increased reactive oxygen species (ROS) production. It has been previously shown that patients with CLL have high levels of ROS compared to normal individuals and that TRX has a protective function in CLL. In an initial pilot study of 17 patients (Koshy et al., AACR Annual Meeting, 2010) we found that TXNIP levels (avg. 0.82 ± 0.18, range 0.45–1.15) and ROS (avg. 7967 ± 2515, range 2340–12340) were positively correlated (ρ = 0.92). Although not statistically significant, TXNIP expression was higher in patients under observation than those receiving treatment (0.89 ± 0.18 vs. 0.76 ± 0.15).

For further insights into the role of TXNIP in CLL, we chose the TCL-1 transgenic mouse model of B-CLL. These animals develop a CD5⁺ B-cell leukemia and express therapeutic targets and display sensitivity to therapeutic agents relevant in the treatment of CLL. Mice develop a clonal leukemia characterized by splenomegaly and increased lymphocyte counts in peripheral blood at a median age of 10 months.

To first determine if TXNIP was altered in the progression of CLL in the TCL-1 mice, we isolated peripheral blood (PB) and bone marrow (BM) from mice of varying ages; 3 months (disease free), 7 months (pre-leukemic phase) and 11 months (end-stage disease). As shown previously, end-stage animals developed severe splenomegaly (1.32±0.82g vs. 0.16±0.08; p<0.05) compared to disease-free. These animals also demonstrated a 1.4 fold increase in TXNIP RNA in BM (0.62±0.08 vs. 0.87±0.12, p<0.05) and 1.5 fold in PB cells (0.81±0.06 vs. 1.18±0.20; p<0.05) associated with a 3.2 fold increase in ROS in BM (15212±1522 vs. 46812±11563, p=0.07)and 3.9 fold increase in PB (11150±560 vs. 39230±8671; p=0.7) compared to animals with no evidence of disease.

Although ROS levels are regulated by several cellular redox mechanisms, we have shown in a breast cancer model, that TXNIP upregulation play a major role in ROS control through its interaction with TRX. We chose to alter the ROS levels by antioxidant therapy on animals with end-stage disease. The reason for this choice was to assess a faster way to interfere with the final product of TXNIP upregulation, rather than the protein itself. We wanted to prove that by reducing ROS we would have changed the course of the disease. For this purpose, we administered the antioxidant, N-acetylcysteine (NAC), in the drinking water and measured the changes in both ROS and TXNIP. Mice of approximately 10 months of age were given fresh NAC water every other day. After a treatment period of 4 weeks animals were sacrificed. Peripheral blood and splenocytes were stained for ROS and B220 expression. Bone marrow cells were stained for ROS, B220 and bone marrow lineage markers and evaluated by flow cytometry.

Upon gross examination of animals, spleen weight was found to be reduced 2-fold in NAC treated animals compared to that of untreated (0.69±0.40 g vs. 1.3±0.77, p<0.05) although still higher than that expected of a healthy animal of the same age (approximately 0.25g). We found that ROS was reduced in the spleen 1.5 fold (59±29 vs. 81±41; p<0.05). B220 expression was increased 1.8 fold (1155±611 vs. 836±416; p<0.05) in spleen cells. In peripheral blood, we found that NAC reduced ROS levels 1.7 -fold (51±19 vs. 29±14; p<0.05) and B220 was decreased by 1.5-fold (857±213 vs. 575±105). Total bone marrow ROS levels were decreased 2- fold by NAC treatment (79±31 vs. 38±13; p<0.01). In the bone marrow, low levels of ROS expression have been associated with primitive hematopoietic stem cells. Indeed, in our studies we found the bone marrow could be separated into a ROS^hi and ROS ^lo population and that NAC treatment increased the percentage of ROS^lo cells. The ratio of ROS^hi/ROS^lo was decreased by 1.5 fold (1.7±0.4 vs. 1.1±0.2; p<0.05) This ROS^lo population encompassed the Lin⁻, c-kit⁺, Sca-1⁺ cells. However while NAC treatment induced changes in ROS levels, as expected, TXNIP RNA levels were not found to differ between NAC-treated and untreated animals in any of the cell types tested.

In conclusion, both TXNIP and ROS levels were increased with progression of the disease. Antioxidant treatment of animals with end-stage disease resulted in decreased ROS in all cell types tested, associated with phenotype changes of the disease. This finding sets the ground for further work in identifying therapeutic tools that interfere with TXNIP-ROS axis.