Most children with B-cell precursor ALL (BCP-ALL) have an excellent prognosis but some patients are susceptible to CNS relapse, the mechanisms of which are largely unknown. Zap-70 is a tyrosine kinase mainly expressed in normal T- and NK-cells, but it is also expressed at low levels in normal B-cells. Zap-70 is however overexpressed in B-cell chronic lymphocytic leukemia (B-CLL) and is associated with an aggressive course of the disease. In B-CLL, Zap-70 was shown to up-regulate the expression of CCR7 (Calpe et al., 2011), which in turn allows T-lymphoblasts to enter the CNS (Buonamici et al., 2009). We hypothesized that Zap-70 mediated upregulation of CCR7 may enhance the homing and the survival of CNS-prone BCP-ALL cells in the CNS niche.
We first transduced the 697 BCP-ALL cell line with a lentiviral vector carrying a non-targeting GFP (shGFP) or a Zap-70 (shZap-70) specific shRNA. Down-regulation of Zap-70 resulted in a significantly reduced CCR7 mRNA and surface protein expression. 697shGFP and 697shZap-70 cells were then injected intravenously into NSG mice, which were sacrificed when leukemic symptoms developed. There were no statistically significant differences in leukemic bone marrow infiltration and survival between the groups. However, immunohistochemical scoring of CNS infiltration (Krause et al., 2015) revealed that 9/10 animals (90%) in the 697shGFP group were CNS+ and only 2/7 animals (29%) in the 697shZap-70 group showed a CNS+ status (p = 0.0345). This suggests that Zap-70 mediated reduction in CCR7 impairs the CNS-positive phenotype of 697 cells in vivo.
We next investigated if the Zap-70/CCR7 axis influences the ability of primary samples from patients to infiltrate the CNS of xenografts. Zap-70 mRNA was measured in pediatric BCP-ALL patients and patients with Zap-70 expression levels higher than the median were considered Zap-70high, and the remaining patients Zap-70low. 1 x 106 cells of 5 Zap-70high and 5 Zap-70low samples were injected intrafemorally into duplicate NSG mice. Interestingly, 7/10 (70%) of the mice injected with Zap-70high cells showed a CNS+ phenotype, whereas only 1/10 (10%) of the mice bearing Zap-70low samples were CNS+ (p = 0.0198). Importantly, leukemic cells of one Zap-70high patient purified from the xenograft CNS showed a marked upregulation of CCR7 expression as compared to cells isolated from bone marrow or spleen. These data suggest that the Zap-70/CCR7 axis is beneficial for patient BCP-ALL cells in the CNS of NSG mice.
To test whether Zap-70/CCR7 is associated with CNS involvement in patients, we analyzed Zap-70 and CCR7 mRNA expression in diagnostic primary material of 76 BCP-ALL patients. The cohort contained 21 CNS-positive and 55 CNS-negative patients and was previously published by Cario et al. (2007). Patients were grouped into either Zap-70high/Zap-70low or CCR7high/CCR7low groups according to expression levels higher or lower than the median of the respective gene. Zap-70 expression was highly correlated with CCR7 expression (p = 0.0003; Spearman r = 0.401). There were no statistically significant differences in outcome surrogates such as prednisone response or MRD and there was no correlation between Zap-70/CCR7 expression and outcome (relapse or death). However, a trend towards high CCR7 levels and the occurrence of death could be observed (p = 0.055). Interestingly, high expression of Zap-70 or CCR7 correlated with the presence of blasts in the cytospin of the initial cerebrospinal fluid sample (p = 0.014 and p = 0.045, respectively). Furthermore, there was a significant correlation between high Zap-70 and a higher CNS status and a trend between high CCR7 and a higher CNS status according to AIEOP-BFM 2009 criteria (p = 0.024 and p = 0.098, respectively). We finally performed comparative correlations in 48/76 patients (63.2%) that were Zap-70high/CCR7high (n=24) or Zap-70low/CCR7low (n=24). Most importantly, a significant correlation between Zap-70high/CCR7high expression and a higher CNS status could be observed (p = 0.009). These data suggest that a high expression of Zap-70 or CCR7 is associated with CNS infiltration in BCP-ALL patients and that measurements of Zap-70 and CCR7 expression should be combined.
Altogether, we show that Zap-70/CCR7 is important for BCP-ALL cells to enter the CNS and to survive in that niche. We propose Zap-70/CCR7 as a mechanism of CNS leukemia and a diagnostic marker with potential therapeutic implications.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.