Anti-ADAMTS13 Antibodies and a Novel Heterozygous p.R1177Q Mutation in a Case of Pregnancy-Onset Immune-Mediated Thrombotic Thrombocytopenic Purpura

Abstract

In this study, we investigated a case of pregnancy-onset thrombotic thrombocytopenic purpura (TTP). The patient had severely decreased ADAMTS13 ( a d isintegrin a nd m etalloprotease with t hrombo s pondin type 1 motif, member 13) activity levels during acute phase and the presence of inhibitory anti-ADAMTS13 autoantibodies was demonstrated, which led to the diagnosis of immune-mediated TTP. However, ADAMTS13 activity was only mildly restored during remission, although inhibitory anti-ADAMTS13 antibodies were no longer detected. We hypothesized that genetic abnormalities could account for this discrepancy between ADAMTS13 activity and antigen. Genetic analysis revealed the presence of two heterozygous substitutions on the same allele: a single nucleotide polymorphism (SNP) c.2699C > T (p.A900V), located in the beginning of the T5 domain, and a mutation c.3530G > A (p.R1177Q) located in the third linker region of ADAMTS13. In vitro testing of those substitutions by expression of recombinant proteins revealed a normal secretion but a reduced ADAMTS13 activity by the novel p.R1177Q mutation, which could partially explain the subnormal activity levels found during remission. Although changes in the linker region might induce conformational changes in ADAMTS13, the p.R1177Q mutation in the third linker region of ADAMTS13 did not expose a cryptic epitope in the metalloprotease domain. In conclusion, we report on an immune-mediated pregnancy-onset TTP patient who had inhibitory anti-ADAMTS13 autoantibodies during acute phase, but not during remission. Genetic analysis confirmed the diagnosis of immune-mediated TTP and revealed the novel p.R1177Q mutation which mildly impaired ADAMTS13 activity.

Keywords: SNPs and mutations; anti-ADAMTS13 autoantibodies; pregnancy-onset TTP; thrombotic thrombocytopenic purpura.

Fig. 1

ADAMTS13 activity, antigen, anti-ADAMTS13 autoantibodies and gene analysis from the TTP patient. Patient plasma samples of acute (Ac) and remission phases (one plasma sample just after the acute episode [remission 1, R1] and one plasma sample almost 2 years after achieving complete remission [remission 2, R2]) were analyzed. ( A ) ADAMTS13 activity was measured using the FRETS-VWF73 assay. NHP was used as a reference and set as 100% ADAMTS13 activity. Below 10% (dotted line) indicates severely decreased activity ( n  = 3). ( B ) Plasma ADAMTS13 antigen levels were measured using ELISA. NHP was used as a reference and set as 1 µg/mL ( n  = 3). ( C ) Anti-ADAMTS13 autoantibodies were detected by adding plasma to wells coated with rhADAMTS13. The dotted line indicates the background ( n  = 1). ( D ) Western blot was also used to detect anti-ADAMTS13 autoantibodies. rhADAMTS13 was loaded onto an SDS-polyacrylamide gel. After transfer, the membrane containing rhADAMTS13 was incubated with plasma samples, or purified human anti-ADAMTS13 antibody II-1 as a positive control. ( E ) A mixing experiment was performed to detect inhibitory anti-ADAMTS13 autoantibodies in plasma from the patient. Heat-inactivated patient plasma was added to fresh NHP (50 V%) and residual ADAMTS13 activity was determined using the FRETS-VWF73 assay ( n  = 3). ( F ) Representation of the domain structure of ADAMTS13 with the metalloprotease domain (M), disintegrin-like domain (D), a first thrombospondin type 1 repeat (T1), cysteine-rich domain (C), spacer domain (S), seven additional T domains (T2-T8), and 2 CUB domains. The three flexible linker regions are indicated by L1, L2, and L3. The SNP c.2699C > T (p.A900V) is located in the beginning of the T5 domain, while the mutation c.3530G > A (p.R1177Q) is located in the third linker region between T8 and CUB1 domains. The sequence chromatograms are shown below the ADAMTS13 domain structure and show the DNA sequence of the SNP and the mutation and their heterozygous nature.

Fig. 2

No effect on ADAMTS13 secretion and conformation, but decreased ADAMS13 activity by p.R1177Q mutation. ( A ) CHO cells were transiently transfected with plasmids pcDNA6.1-ADAMTS13-WT, pcDNA6.1-ADAMTS13-p.A900V, pcDNA6.1-ADAMTS13-p.R1177Q, pcDNA6.1-ADAMTS13-p.A900V/p.R1177Q or pcDNA6.1-ADAMTS13-p.A900V + pcDNA6.1-ADAMTS13-p.R1177Q, and a pmaxGFP plasmid to determine transfection efficiency. ADAMTS13 antigen levels in the expression medium were determined through ELISA. NHP was used as a reference and set as 1 µg/mL. Kruskal–Wallis compared with WT ( n  = 3). ( B ) A FRETS-VWF73 assay was performed to determine activity of WT and p.A900V, p.R1177Q, and p.A900V/R1177Q ADAMTS13 variants. Kruskal–Wallis compared with WT, * p  = 0.0283, ** p  = 0.0005 ( n  = 9–14). ( C ) The cryptic epitope of antibody 6A6 in the metalloprotease domain of ADAMTS13 is accessible in the truncated ADAMTS13 variant MDTCS (ADAMTS13 without T2 till CUB domains; left), but not in full length ADAMTS13 (right). ( D ) Conformation of ADAMTS13 variants was studied by adding WT, p.A900V, p.R1177Q, and p.A900V/p.R1177Q ADAMTS13 (6.25nM) to coated 6A6. MDTCS was used as a positive control. Kruskal–Wallis compared with WT, p  < 0.05 ( n  = 3).