The Authors’ Reply

We thank Visentin and coauthors1 for pointing out that our findings2 are in line with their earlier observations. Because the current clinical objective is to detect the antibodies (Abs) binding to peptide-associated β2-microglobulin-associated HLA-I heavy chains (pepA-β2aHC), the native HLA-I conformation expressed on the surface of resting cells, the presence of other conformational variants on single antigen beads (SAB) prevents the distinction between clinically relevant and irrelevant HLA Abs, in pre- and posttransplant patients, when the mean fluorescence intensities is below a certain threshold depending on the specific bead. The indiscriminate usage of the term anti-denatured (dn) HLA Abs adds additional concern. El-Awar et al3 have referred peptide-free β2-microglobulin-free HLA-I heavy chains (pepF-β2fHC) as natural antibodies or “anti-dnHLA Abs.” To distinguish pepA-β2aHC from pepF-β2fHC, the regular beads are subjected to acid or alkali or heat treatments, which not only releases β2-microglobulin and a peptide from the heavy chain (HC) but also alters the physical configuration of the HC, depending on the treatments. Failure to recognize this alteration has created ambiguity and misperception, as in a recent publication.4 Opposite conclusions were inferred from the impact of anti-dnHLA Abs using different denaturation methods. Visentin et al5 have used “acid-treated beads” to conclude that anti-dnHLA Abs do not impair graft survival,1 whereas Cai et al6 used the “heat-treated (90°C) beads” to state that “complement-fixing antibodies against denatured HLA and MICA antigens are associated with antibody mediated rejection.” The β2fHC are also expressed naturally (“open conformers”7) on activated lymphocytes,8 and serve as ligands for leukocyte receptors.9 Therefore, the naturally occurring (or expressed) β2fHC cannot be referred to as dnHLA. Perhaps beads with only β2fHC may specifically identify clinically relevant β2fHC Abs.

It is true that monoclonal antibodies TFL-006 provided lower mean fluorescence intensities than monoclonal antibodies HC-10 on acid-treated beads (displaying only pepF-β2fHC). Obviously, the acid denaturation (acid-dn) affected the epitope recognition by TFL-006 more strongly than the one recognized by HC-10. The following reasons support the above contention:

• TFL-006, raised by immunizing the folded β2m-free HLA-E^R107 HC,10 recognizes a conformational epitope on β2fHC. In contrast, HC-10 was raised against denatured papain-fragments of HLA-B7 and HLA-B40 HC.11 Therefore, the TFL-006 epitope may be more sensitive to the acid-dn of SAB compared to the HC-10 epitope.

• Although the TFL-006 epitope (¹¹⁷AYDGKDY¹²³) is located in the α2-domain of HC,12 the antigenic determinant (R⁶²) of HC-10 is located in the α1-domain. Upon acid, alkali and heat denaturation of HC, the α1-domain is most susceptible to structural changes, because the α2- and α3-domains contain disulfide bonds. Such denaturation of the α1-domain does not abolish HC-10 reactivity, as Perosa et al13 documented that HC-10 recognizes a linear epitope, but the extreme denaturation of the α1-domain may hinder the accessibility of TFL-006 to its epitope on α2.

Moreover, the epitope of TFL-006 is shared by HLA-A/-HLA-B/HLA-Cw/HLA-E/HLA-F and HLA-G,12 and the location of the sequence (shown in figures14) is cryptic in β2aHC because β2-microglobulin masks the epitope; TFL-006 recognizes only the β2fHC HLA-I. On the other hand, the crypticity of the HC-10 epitope is only dependent on the presence of a peptide in the groove, and the epitope is not shared by all HLA-I antigens coated on SAB. HC-10 recognizes both peptide-free β2-microglobulin-associated HLA-I heavy chains (pepF-β2aHC) and β2fHC.

Visentin et al5 described patients’ sera which contain anti-dnHLA directed against all the beads which are not recognized by TFL-006; however, the HLA-I specificity is not explicit. Nonetheless, we inferred that “anti-dHLA” that can stain iBeads corresponded to antibodies directed against pepF-β2aHC because W6/32 (which confirms the presence of pepA-β2aHC and pepF-β2aHC) and HC-10 (which confirms the presence of pepF-β2aHC and β2fHC) bind but TFL-006 (confirming the absence of β2fHC) does not. By the process of elimination, only pepA-β2aHC and pepF-β2aHC are present on iBeads. In addition, such “anti-dHLA that can stain iBeads” but are not associated with positive FCXM could be similar to HC-10. Therefore, we concluded that “anti-dHLA that can stain iBeads” mimic HC-10 reactivity in SAB and FXCM. We are aware that some peculiar epitopes can be conserved on all HLA-I conformational variants, but it needs verification as to whether Abs against these epitopes are positive with FXCM.

Finally, we conclude that (i) the manufacturing process of SAB produces a highly heterogeneous product, with varying densities (or concentrations) of the different variants, which impedes recognition of the conformational specificity of anti-HLA Abs; (ii) there is a need for the resurgence of iBeads to detect HLA Abs specifically targeting pepA-β2aHC; (iii) a consensus is needed for the terminology used to describe anti-HLA Abs against different conformational variants and (iv) the term “anti-dnHLA Abs” should only be restricted to Abs tested on physicochemically treated SABs, and the name of the treatment should be specified (ie’, anti–acid-dnHLA Ab).