Complement Biosensors Can be Used to Identify Classical Pathway and Alternative Pathway Dysregulation in Complement-Mediated Thrombotic Microangiopathy

Complement-mediated thrombotic microangiopathy or hemolytic uremic syndrome (CM-TMA/CM-HUS) is a thrombotic microangiopathy commonly characterized by germline variants or acquired antibodies to complement proteins and regulators. However, these variants are neither necessary nor sufficient, and many patients are found to have variants of undetermined significance complicating definitive diagnosis. New tools are needed to provide a global assessment of complement dysregulation and better differentiate TMA into categories that might predict response to complement inhibitor therapies.

We developed an array of cell-based complement “biosensors'' by selective removal of complement regulatory proteins in an autonomously bioluminescent HEK293 cell line: a PIGA knockout (PIGAKO, deficient in CD55 and CD59), a CD46 knockout (CD46KO, deficient in CD46), or a double knockout (DKO, knockout of PIGA and CD46 resulting in deficiency of CD46, CD55, and CD59).

First, we demonstrate the ability of these biosensors to detect complement-mediated cytolytic effects in healthy control (HC) sera comparing the wild type (n=19), CD46KO (n=60), PIGAKO (n=61), and the DKO (n=29) knockout cell lines. We also demonstrate this cytolytic activity is blocked by heat inactivation, isolated alternative pathway (AP) buffer (GVB-MgEGTA), sutimlimab (C1s inhibitor) and eculizumab (C5 inhibitor), suggesting a classical pathway (CP) stimulus. For CM-TMA samples (acute n=6, remission n=17) we demonstrate significantly increased CP stimulus compared to HC, acute thrombotic thrombocytopenic purpura (n=6), or other hospitalized controls (n=10). This stimulus is blocked by sutimlimab and eculizumab, but not by alternative pathway inhibitors including, AMY-101 (C3 inhibitor), iptacopan (factor B inhibitor, FBi), danicopan (factor D inhibitor, FDi), or even a combination of all three. We confirm these findings by flow cytometry on wild type HEK293 and TF1 cells, showing increased C4d deposition that is blocked by sutimlimab. Further mechanistic studies utilizing DTT, IdeS, protein G columns, and commercial preparations of immunoglobulin fractions identify the culprit immunoglobulin species as IgM and flow cytometry studies show a trend towards increased IgM binding in CM-TMA compared to HC sera.

We next show the utility of performing the assay in isolated AP buffer (GVB-MgEGTA). HC sera or CM-TMA sera in AP buffer on the CD46KO or PIGAKO cell lines demonstrates complete protection from cytolytic activity by residual surface complement regulators. However, on the DKO cell line in AP buffer, HC sera (n=22) or CM-TMA sera (n=18) demonstrates AP cytolytic activity that is increased in the case of FHAA or germline variants and blocked by AP inhibitors (C3, FDi, or FBi). Of 7 acute CM-TMA samples studied in both all pathway and AP buffer, 1 shows isolated CP dysregulation, 1 shows isolated AP dysregulation, and 5 of 7 show both CP and AP dysregulation. Of 9 remission CM-TMA samples in which genetic testing was performed, 5 have identified variants. Of those with variants, 2 show both AP and CP dysregulation and 2 show isolated AP dysregulation. Whereas in those without an identified variant (n=4), none show ongoing AP dysregulation and 2 show ongoing CP dysregulation.

Finally, given the lack of an easily accessible method to monitor complement inhibitors for therapeutic efficacy, we show that utilization of the DKO cell line in either all pathway buffer or isolated AP buffer can demonstrate protection of membrane surfaces from either CP or AP activity. Such an approach may provide an individualized assay to guide dosing strategy in the case of nonresponse to therapy.

In summary, these biosensors can be used as a sensitive method for investigating individual mechanisms of complement pathway activation in CM-TMA and help monitor therapeutic complement blockade. Identification of a CP stimulus in CM-TMA provides a potential explanation for ~50% of CM-TMA patients who lack an AP “driving” variant and suggests at least a subset of CM-TMA is characterized by a breakdown of IgM immunologic tolerance.