To the editor:

In the August 2003 issue of Nature Immunology1, as well as in later studies2, Voβhenrich et al. presented experimental data on the cytokine requirements of developing B cells, which led them to conclude that “TSLP [thymic stromal lymphopoietin] is the factor responsible for most of the fetal and perinatal B cell production that takes place when the IL-7–γc [interleukin 7–common γ-chain] signaling pathway is disrupted.”1 Although the data reported were technically sound and compatible with such a conclusion, the authors did not provide direct evidence to support (or exclude) the idea of a critical function for TSLP in IL-7-independent fetal B lymphopoiesis. The conclusions of Voβhenrich et al. were based on the demonstration that B lymphopoiesis was much more affected (tenfold more) in mice deficient in IL-7 receptor α-chain, essential for IL-7 as well as TSLP signaling, than in mice deficient in the common γ-chain (γc), required for IL-7 but not TSLP-mediated signaling3,4. However, these data could at best be considered strong indirect support for the idea of TSLP as the main cytokine driving IL-7-independent fetal B lymphopoiesis, as there could be other reasons for a difference in the phenotypes of γc-deficient mice and those deficient in the IL-7 receptor α-chain. Furthermore, Voβhenrich et al. used bone marrow of mice 4–12 weeks of age, not fetal liver, for their comparative in vivo analysis of B lymphopoiesis in these mice1,2. Instead, the extrapolation to the idea that TSLP is key to the fetal stages of B lymphopoiesis was based on the finding that fetal but not adult pro–B cells were responsive to TSLP in vitro1,2. In contrast, a lack of an important function for TSLP in adult B lymphopoiesis has been indicated by studies of TSLP receptor–deficient (Tpte2−/−) mice5,6.

As fetal lymphopoiesis had not been examined in singly deficient Tslp−/− or Tpte2−/− mice, we investigated B lymphopoiesis in the livers of Tpte2−/− mice at embryonic day 17.5 but found no deficiency in Tpte2−/− fetuses at any stage of B cell development (Fig. 1a and Supplementary Fig. 1 online). Furthermore, when comparing B lymphopoiesis in the fetal livers of Il7−/− and Il7−/−Tpte2−/− mice, we obtained no evidence for substantial involvement of TSLP in IL-7-independent regulation of fetal pro–B cells or pre–B cells, whereas we noted a slight additional reduction in the number of immature B cells in Il7−/−Tpte2−/− fetuses relative to that in Il7−/− fetuses (Fig. 1b and Supplementary Fig. 1). Thus, although Voβhenrich et al. provided compelling evidence that fetal pro–B cells are highly responsive to TSLP1, our studies of Tpte2−/− and Il7−/−Tpte2−/− fetuses fail to support their claim that TSLP is the most important cytokine promoting IL-7-independent fetal B lymphopoiesis. Instead, although Voβhenrich et al. also concluded that “Flk-2 is involved, but TSLP is the main factor driving IL-7-independent fetal and perinatal lymphopoiesis,”1 we have done additional studies of mice deficient in the cytokine Flt3L (also called Flk-2 ligand) and IL-7 (Flt3l−/−Il7−/− mice) and of Flt3l−/−Tpte2−/− mice and have found that the reported complete loss of B-1 as well as B-2 B lymphopoiesis in Flt3l−/−Il7r−/− mice7 and Flk2−/−Il7r−/− mice1 is entirely due to the simultaneous loss of function of IL-7 and Flt3L (C.T.J. and S.E.W.J., unpublished observations). Collectively, our findings suggest that Flt3L rather than TSLP is the key regulator of IL-7-independent B lymphopoiesis and that intact TSLP function is insufficient to restore any detectable B lymphopoiesis in the absence of these two critical regulators of B cell progenitors.

Figure 1: Critical function for IL-7 but not TSLP in the regulation of fetal B cell progenitors.
figure 1

Mean numbers (+ s.d.) of pro–B cells (Pro-B; B220+CD43+AA4.1+CD19+IgM), pre–B cells (Pre-B; B220+CD43AA4.1+CD19+IgM) and immature B cells (Imm-B; B220+CD43AA4.1+CD19+IgM+) in fetal livers at embryonic day 17.5 for littermate wild-type mice (WT; n = 8) and Tpte2−/− mice (n = 4; each from three litters; a) and for littermate Il7−/− mice (n = 10) and Il7−/−Tpte2−/− mice (n = 10; each from five litters; b). *, P = 0.04. Corresponding flow cytometry plots are in Supplementary Figure 1.

Note: Supplementary information is available on the Nature Immunology website.