**2. What Is HRF?**

Cytokine-like factors able to activate basophils in body fluids of allergic patients have been studied for many years [4]. Several chemokines were shown to induce histamine release from human basophils in an IgE-independent manner [5–7]. On the other hand, an IgE-dependent factor with histamine-releasing activity (HRF) was molecularly cloned by Susan MacDonald's group in 1995 [8]. Coincidentally, HRF happened to be identical to the protein termed translationally-controlled tumor protein (TCTP), fortilin, p21, and p23. It is often referred to as TCTP intracellularly and is required for cell cycle progression, proliferation, survival, and malignant transformation in a variety of cell types [9–14]. Extracellularly referred to as HRF (we follow this convention in this manuscript), it is an evolutionally conserved protein (96% identical between human and mouse proteins) composed of 172 amino acids with no known related proteins. Human HRF/TCTP is encoded by the *TPT1* gene on chromosome 13. Although numerous single nucleotide polymorphisms (SNPs) are associated with allergic diseases, no genetic associations with gene expression (eQTLs) are found in the *TPT1*

locus (http://dicew-database.org). Similar to antigen/IgE-mediated activation, HRF induces not only histamine release, but also IL-4 and IL-13 secretion from human basophils and IL-13 and TNF secretion from murine mast cells [15,16]. Despite the lack of a signal sequence, it is secreted as a cargo of extracellular vesicles (EVs), particularly in exosomes [17–20]. Intriguingly, the responsiveness of basophils to HRF depends on a particular type of IgE; IgE derived from certain atopic patients, termed IgE<sup>+</sup>, can prime basophils in response to HRF, but other IgE molecules, termed IgE<sup>−</sup>, are unable to do so [21]. The dichotomy of IgE+ vs. IgE− was discovered long before the molecular cloning of HRF, and several possibilities exist to explain the heterogeneity of IgE molecules: 1) structural di fferences in the constant regions of IgE, for example, by di fferences in glycosylation or alternative mRNA splicing at the ε chain 3- terminal region [22]; 2) IgE+ being an HRF-specific IgE antibody, that is, HRF acting as an IgE autoantigen; 3) IgE+ reactivity due to the presence of anti-IgE antibodies in the serum.

In contrast to an earlier report suggesting that HRF does not bind to IgE [23], Kashiwakura et al. showed that a subset of IgE and IgG molecules are able to directly bind to HRF via two Ig Fab-interacting sites: the N-terminal 19 residue stretch (N19) and the H3 helix [24]. These observations are in line with an earlier speculation that the dichotomy of IgE+ vs. IgE− may be caused by di fferences in IgE variable region subgroups [25]. However, another speculation that IgE+ reactivity is related to glycosylation of IgE [21] was not supported by the observation that mannose-specific lectins could not distinguish between basophils sensitized with IgE+ or with IgE− [26]. Despite these studies, it still remains possible that glycosylation at V H and VL regions might contribute to the IgE+ reactivity. In light of recent revelations regarding IgE glycosylation [27], the potential role of glycosylation may be worth revisiting.
