New Modular Method for Generating Multispecific Antibodies Will Require A Superior Design Suite

Designed to Cure

Protein structure of a human antibody of the IgG2 isoform generated in-vivo via VDJ recombination. Source: Wikipedia Commons

As biotechs everywhere flock to to the biologic gold rush, antibody design is at the tip of everyone’s tongues. Developing effective and safe biologics is a difficult process, requiring an abundance of laboratory infrastructure, binding studies, cultures and screening. Traditional biologics that have seen clinical success, such as rituximab, are being called revolutionary,1 but carry many side effects with them.

Researchers have long sought to reduce these side effects by  increasing the number of specificities of the antibodies used in therapy.2 Increasing the number of potential specificities allows for a therapeutic to have a smaller target population of cells. Multispecific biologics which have efficacy against multiple targets while retaining few side effects are now possible thanks to a new method for creating modular biologics using a stable scaffolding.3

Using this modular method to generate modular binding domains promises to revolutionize the biologic development process and enable the rapid prototyping and subsequent efficacy screening of multispecific biologics with relative ease, provided that there is sufficiently advanced software during the antibody design phase of development. In order to make the most use of the modular method, researchers must pair it with a software suite which can simulate binding, biochemistry, track the data generated during screening and analyze downstream biological efficacy.

How to Make a MATCH

The modular method for generating antibodies is known as “multispecific antibody-based therapeutics by cognate heterodimerization”, or MATCH for short. In a nutshell, MATCH uses a stable scaffolding of a generalized fragment variable region to accommodate two split variable binding domains. The split variable binding domains can be chemically changed depending on the requirements of the biologic’s intended application, and the resulting biologic can have up to four specificities.

The workflow of biologic creation using MATCH will look a bit different from traditional antibody generation, especially a few major points:

  • Biologics generated via MATCH will need to be simulated beforehand via software in order to pre-screen out variants which may have contradictory biochemistry that might prevent binding
  • MATCH-generated biologics will require a multiplicatively greater number of variants to be screened before suitable candidates are found, as multiple specificity binding domains must be tested simultaneously to verify efficacy
  • MATCH-generated biologic candidates which pass screening will also be carried forward into further research in multiplicatively greater in numbers as a result of differing ratios of binding affinities having different downstream biological effects
  • By virtue of having a gradient of differing antigen binding affinities among each individual biologic variant’s domains, downstream therapeutic effects and side effect profiles can be tuned

More MATCHes, More Simulations, More Data

Given that the biochemistry of antibody binding is subject to a plurality of different properties including electrostatic interactions between domains, biologic design has long demanded a robust software package that is capable of assisting researchers with the production of effective variants.4 As B-cell immunologists are aware, binding simulation, data management and tracking of screening candidates are essential when generating biologics.

The design of effective biologic variants for subsequent screening is of particular concern while using MATCH. The MATCH scaffolding system guarantees the ability for the biologic to accommodate four binding points, but does not guarantee that the biochemical interactions between the biologic’s binding points confer high affinity for their target antigens. For that, the split-variable binding domains must be designed separately before unification via the modular system. Following this step, further simulations of electrostatic interference and protein folding are required in order to determine if the hypothetical biologic is actually chemically possible and will have binding activity.

Informatically, using MATCH requires an extensive amount of simulation and predictive analysis:

  • Biochemically designing four separate split variable binding domains
  • Simulating the target antigens of each of the binding domains
  • Simulating each of the binding domains’ affinity for their target antigens and their binding
  • Combining the four split variable binding domains that have high affinity for their target antigen into the MATCH scaffolding fragment variable domain
  • Simulating the entire MATCH-generated biologic’s biochemical interactions with itself alone
  • Simulating the entire MATCH-generated biologic’s chemical interactions with the antigens targeted by each of its binding domains separately and then additively
  • Simulating the downstream effects of the MATCH-generated biologic’s multiple antigen binding
  • Simulating the PK of the biologic itself
  • Predicting the breakdown of the biologic
  • Simulating the PK of the biologic’s metabolites

Currently, many biochemical simulation and antibody design suites are separate from primary LIMS, leading to frustrating cross-referencing between software packages and dropped data.

An informatics system that is ready to be embedded with the MATCH process will have the following traits:

  • Integration with flexible and fully featured biochemical design and simulation suites to perform in silico analysis of binding efficacy before variant generation
  • Downstream biological effect prediction
  • Integration with traditional LIMS in order to plan experiments and track variant generation procedures
  • Multiparametric tracking of biologic candidates
  • Collaborative data distribution and analysis to facilitate parallel research branches

Traditional LIMS are not up to the task, and frequently falter at seamless collaboration during experimental planning due to clunky integration with simulation suites.While bridging multiple informatics systems is still possible within the scope of traditional biologic generation which seeks a maximally binding antibody for a single antigen, using MATCH will require an order of magnitude more integration with informatics and design systems due to the plurality of variables which must be tested against each other during screening and downstream efficacy testing.

Using BIOVIA’s Designed to Cure for Biologics, data about biologics generated using MATCH will be consumable as a library of multispecific antibody candidates to send forward to the clinic based off of tailored criterion decided before experimentation. Combining Designed to Cure for Biologics and MATCH will streamline the biologic generation pipeline and allow for patient-friendly biologics which confer fewer side effects. Contact us today to find out how BIOVIA can help you get the most out of the latest technology and help the patients of the future’s biologic therapies.

  1.  “Rituximab in ANCA-associated vasculitis: a revolution?“ July 27, 2011,;26/10/3077
  2. “The safety and side effects of monoclonal antibodies.“ April 2010,
  3. “Novel multispecific heterodimeric antibody format allowing modular assembly of variable domain fragments.“ October 27th, 2016,
  4. “Simulation of electrostatic effects in Fab–antigen complex formation.“ August, 2000,