Abstract
Acquired thrombotic thrombocytopenic purpura (TTP) is an autoimmune disease due to autoantibodies (Auto-Ab) against ADAMTS13, resulting in a severe deficiency in enzyme activity (<10%). Little is known about the pathogenicity of anti-ADAMTS13 IgG directed against different domains of the enzyme, or how domain specificity alters between presentation, remission or TTP relapse. We developed novel assays to explore domain specificity and inhibitory potential of anti-ADAMTS13 IgG using samples from the UK TTP Registry.
Full-length ADAMTS13 and ADAMTS13 fragments (MD, MDTC and MDTCS - N-terminal domains; TSR2-8 and CUB1/2 - C-terminal domains) were expressed in HEK293T cells and purified. Microtitre plates were coated with ADAMTS13 or ADAMTS13 fragments, incubated with TTP plasmas and bound IgG detected. Domain specificity was further analysed by preincubation of patient samples with ADAMTS13 fragments, assessing the ability of fragments to compete for full-length ADAMTS13 binding on the plate. From a cohort of 92 acquired TTP patients at presentation, 100% (92/92) contained auto-Ab that recognised ADAMTS13 N-terminal domains (MDTCS). Of these, 39% (36/92) had auto-Ab recognising the N-terminal domains alone, 22% (20/92) also had anti-TSR2-8, and 39% (36/92) had anti-CUB1/2 auto-Abs. There was no correlation between auto-Ab domain specificity and mortality/disease severity (anti-ADAMTS13 IgG titre, plasma exchanges to remission, platelet count and cardiac involvement).
There was no evidence for auto-Ab recognising MD or MDTC domains in 25 patients tested, suggesting immunoreactivity for MDTCS was primarily mediated through the spacer domain. Consistent with this, TTP IgG inhibited proteolysis of VWF115 (a recombinant VWF A2 domain fragment), but failed to inhibit VWF106 proteolysis (VWF A2 domain fragment lacking the spacer domain binding exosite).
To examine the inhibitory potency of anti-ADAMTS13 auto-Ab, IgG isolated from TTP patient samples were titrated in reactions measuring MDTCS-mediated proteolysis of FRETS-VWF73. Twenty eight out of 43 presentation samples tested had inhibitory IgG (IC50 ranging from <0.2-5μM total IgG). Addition of excess of MDTC to 19/19 of these samples did not alter the IC50, confirming inhibition in these samples is mediated solely by anti-spacer antibodies. Intriguingly, 15/43 patients had no evidence of inhibitory auto-Ab, suggesting the pathogenic mechanism in these cases may be enhanced clearance of ADAMTS13. We therefore determined ADAMTS13 antigen concentrations by ELISA. All patients at presentation exhibited markedly reduced ADAMTS13 antigen levels (n=91, median 58ng/ml (range 0-450) vs. normal n=67, median 951ng/ml (515-1829); p<0.0001). Interestingly, patients with predominantly C terminal auto-Ab and no evidence of inhibitory auto-Ab had significantly lower ADAMTS13 antigen levels at first presentation than the rest of the cohort (median 2ng/ml (0-141) vs. 57ng/ml (0-356); p=0.006).
Cases where rituximab failed to normalise ADAMTS13 activity despite clinical remission were associated with the persistence of inhibitory anti-N-terminal auto-Ab. Two patients with persistently high anti-ADAMTS13 IgG but normal ADAMTS13 activity and antigen levels during remission had non-pathogenic anti-TSR2-8 antibodies that seemingly had no effect upon ADAMTS13 function or clearance. Nine out of 16 patients exhibited altered auto-Ab domain specificity upon relapse, with either loss or development of anti-C-terminal auto-Ab.
In conclusion, from our functional studies ADAMTS13 inhibition appears to be mediated exclusively by anti-spacer antibodies, with no evidence for the presence of auto-Ab that recognise/inhibit other N terminal domains. Importantly, we demonstrate ADAMTS13 antigen depletion is a crucial pathogenic mechanism in acquired TTP, accounting for a large proportion of the reduced ADAMTS13 activity detected in patient samples. Our results suggest that auto-Ab directed against the C-terminal domains of ADAMTS13 may be particularly important in mediating this clearance. There is continual development of the autoimmune response during treatment of TTP patients that can result in altered domain specificity at relapse.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.