IgG Alloantibodies Against RBC Induced By Pregnancy or Transfusion Have Unique Glycosylation Patterns Which Correlate with Clinical Outcome of Hemolytic Disease of the Fetus or Newborn

Hemolytic disease of the fetus or newborn (HDFN) is a potentially life-threatening disease in which red blood cells (RBC) of the child are destroyed by maternal IgG alloantibodies. HDFN caused by anti-D antibodies is well known, but may also occur due to other RBC antibody specificities, with anti-K, anti-c and anti-E being most frequently implicated (Koelewijn et al. Transfusion 2008;48(5):941-52). Besides alloantibody-mediated RBC destruction, anti-K also impairs erythropoiesis. The composition of the sugar moiety of the IgG-Fc tail influences the binding affinity to IgG Fc receptors (FcγR) on effector cells. Particularly, the absence of core fucose increases the binding of IgG to FcγRIIIa and FcγRIIIb. In previous studies we observed a correlation between fucosylation of anti-D and anti-platelet antibodies with severity of disease in patients with HDFN (n=8) or fetal or neonatal alloimmune thrombocytopenia (FNAIT) (n=48) (Kapur et al Br.J.Haematol. 2014 166(6): 936-945; Kapur et al Blood 2014;123(4):471-80).

The aim of this study was to investigate whether skewed antibody glycosylation also applies to other anti-RBC antibodies and if this is related to the immunization route and disease severity.

Clinical data were collected for 146 women with anti-K, c or E antibodies. Anti-K (n=30), anti-c (n=69) and anti-E (n=47) antibodies were purified from serum by acid elution after incubation of the serum with matching positive RBC using antigen-negative RBC as negative-control for specificity of the procedure. Anti-RBC specific IgG antibodies and total IgG antibodies were subjected to tryptic digestion and resulting IgG1 Fc-derived glycopeptides analyzed by MALDI-TOF-Mass Spectrometry.

Lowered fucosylation was observed for anti-K IgG (average 75.7% compared to 92.1% of total IgG, p<0.0001), but less than observed for anti-D IgG1 (52.29%, p=0.0008). Also fucosylation of anti-E IgG was lowered (91.1%, p=0.02), but significantly less pronounced then anti-K and anti-D. Surprisingly, anti-c fucosylation was not significantly skewed (89.9%, p=0.25), but with lowered fucose down to 60% in a few anti-c cases. Anti-E is in most cases naturally occurring, which was definitively proven in 17 cases without history of transfusion or E-positive pregnancy. Anti-K is in most cases induced by transfusion (Koelewijn et al BJOG 2009;116(10): 1307-1314). Analyzing only cases of mothers with blood transfusion history and antigen negative fathers (n=15) also showed lowered fucosylation (p=0.014), suggesting that signatures of low IgG-fucosylation are not related to pregnancy responses.

In 113 antigen-positive newborns at risk for HDFN (K: n=16, c: n=62, E: n=35), the glycosylation pattern of the specific IgG correlated with anaemia and severity of jaundice. For anti-K, galactosylation and sialylation correlated with hemoglobin (r=-0.78 ,p=0.0015, and r=-0.59, p=0.0303, resp.). Similar correlations were found with hematocrit levels. A low anti-c fucosylation correlated with increased bilirubin level (r=-0.38, p=0.0186) and severity of jaundice (r=-0.35, p=0.0103). Newborns exposed to anti-c that required a blood transfusion showed significantly lower galactosylation and higher bisection (p=0.03, p=0.01 resp.).

In conclusion, we found similar skewing of antigen-specific IgG1 glycosylation for anti-E and anti-K as observed for anti-D and anti-platelet antibodies, although less pronounced. This skewing was not related to the immune response being induced in pregnancy, as similar features were also observed after transfusion. Interestingly, natural anti-E antibodies, probably evoked against non-blood borne antigens and recognizing E by molecular mimicry are far less skewed, possibly explaining why the presence of this antibody rarely has clinical consequences. The less pronounced skewing of glycosylation of anti-c might also explain the lower risk of severe HDFN mediated compared to anti-D. Anti-K glycosylation influenced clinical outcome, but in contrast to anti-Rh antibodies, it was not fucosylation but galactosylation and sialylation that correlated with anaemia. This might be related to the differences in the mechanism by which these antibodies induce anemia. Importantly, the clinical outcome of HDFN is strongly affected by IgG1-glycosylation, which is proving to become an important biomarker in screening for severe cases of HDFN.