Science at Mabylon

The Mabylon vision

Genetics, genomics and epigenomics can explain only a fraction of the human disease space. Not only is susceptibility to infections crucially dependent on the immune repertoire present in any individual, but antibodies can be predictive of a plethora of diseases unrelated to infectious agents.

Conversely, autoantibodies against select endogenous targets can protect against specific diseases. Hence the “antibody repertome” is an important and understudied aspect of Personalized Health.

The superiority of antibodies derived from humans is emerging in a growing number of clinically relevant instances. For example, human antibodies against cytomegalovirus are highly neutralizing and protective - whereas their murine counterparts are not. Also, the human-derived antibody, Aducanumab, is currently considered to be one of the most promising drugs against Alzheimer's disease. There is ample reason to expect that the list of diseases optimally treated with human antibodies will continue to grow.

Comprehensive human antibody screens

Mabylon is screening large cohorts of individuals giving informed consent to the utilization of surplus blood samples (>100’000 individuals/year) for the presence of autoantibodies against a panel of endogenous antigens. In collaboration with UZH, we avail of a high-throughput pipeline which can screen up to 2’000 patients/day for autoantibodies. Each sample is tested at dilutions ranging across 5 orders of magnitude using a Labcyte Echo acoustic dispenser integrated in a fully automated, robotic workstation with plate washers, centrifuges and readers.

Analyses of binding profiles, affinity estimations and stringent quality controls are performed in real time with customized software developed at the University of Zurich in collaboration with the Swiss Institute of Bioinformatics. Positive hits are being correlated quantitatively to multidimensional matrices of encrypted clinical data including inter alia diagnoses, family history, and medication.

Antibodies to therapeutically relevant targets (e.g. proteins involved in neurodegeneration) are being cloned from memory B-cells of high-titer patients. Single-cell cloning allows for perfect matching between the respective heavy and light chains of each antibody. The availability of the genetic information allows for lab-scale production and for testing therapeutic efficacy in preclinical models of disease.

Drug development

Emerging lead antibodies, with the desired affinity and selectivity, are expected to be directly convertible into candidate drugs. Our approach is fundamentally novel and relies on the availability of large cohorts of human individuals. Rather than using molecular markers to stratify patients with a given diagnosis (as in conventional personalized medicine), we are interrogating depersonalized data from clinical information systems (such as the main database of USZ) for clinical hallmarks in patients exhibiting specific autoimmune profiles.

This unbiased, hypothesis-free approach has the potential to uncover entirely new pathogenetic principles which may not be inferred from prior knowledge. It will also yield novel biomarkers of disease that cannot be discovered by any other methodology.

Mabylon has established the capabilities necessary to execute the work program that we envisage, including successful IRB submission and approval by the local bioethics authorities (Kantonale Ethik-Kommission of the Canton of Zurich). The screens described above are performed daily on 5 days/week.