Active and machine learning-based approaches to rapidly enhance microbial chemical production

In order to make renewable fuels and chemicals from microbes, new methods are required to engineer microbes more intelligently. Computational approaches, to engineer strains for enhanced chemical production typically rely on detailed mechanistic models (e.g., kinetic/stoichiometric models of metabolism)—requiring many experimental datasets for their parameterization—while experimental methods may require screening large mutant libraries to explore the design space for the few mutants with desired behaviors. To address these limitations, we developed an active and machine learning approach (ActiveOpt) to intelligently guide experiments to arrive at an optimal phenotype with minimal measured datasets. ActiveOpt was applied to two separate case studies to evaluate its potential to increase valine yields and neurosporene productivity in Escherichia coli. In both the cases, ActiveOpt identified the best performing strain in fewer experiments than the case studies used. This work demonstrates that machine and active learning approaches have the potential to greatly facilitate metabolic engineering efforts to rapidly achieve its objectives.     

Lung tissue engineering and preservation of alveolar microstructure using a novel casting method

Abstract

We used a rat model to decellularize and seed alveolar cells on a three-dimensional lung scaffold to preserve alveolar microarchitecture. We verified the preservation of terminal respiratory structure by casting and by scanning electron microscopy (SEM) of the casts after decellularization. Whole lungs were obtained from 12 healthy Sprague-Dawley rats, cannulated through the trachea under sterile conditions, and decellularized using a detergent-based method. Casting of both natural and decellularized lungs was performed to verify preservation of the inner microstructure of scaffolds for further cell seeding. Alveolar cell seeding was performed using green fluorescent protein (GFP) lung cells and non-GFP lung cells, and a peristaltic pump. We assessed cell seeding using histological and immunohistochemical staining, and enzymatic evaluation. All cellular components were removed completely from the scaffolds, and histological staining and SEM of casts were used to verify the preservation of tissue structure. Tensile tests verified conservation of biomechanical properties. The hydroxyproline content of decellularized lungs was similar to native lung. Histological and immunohistochemical evaluations showed effective cell seeding on decellularized matrices. Enzymatic measurement of trypsin and alpha 1 antitrypsin suggested the potential functional properties of the regenerated lungs. Casts produced by our method have satisfactory geometrical properties for further cell seeding of lung scaffolds. Preservation of micro-architecture and terminal alveoli that was confirmed by SEM of lung casts increases the probability of an effective cell seeding process.     

Predicting the structural integrity of bone defects repaired using bone graft materials

Bone defects create stress concentrations which can cause fracture under impact or cyclic loading. Defects are often repaired by filling them with a bone graft material; this will reduce the stress concentration, but not completely, because these materials have lower stiffness than bone. The fracture risk decreases over time as the graft material is replaced by living bone. Many new bone graft materials are being developed, using tissue engineering and other techniques, but currently there is no rational way to compare these materials and predict their effectiveness in repairing a given defect. This paper describes, for the first time, a theoretical model which can be used to predict failure by brittle fracture or fatigue, initiating at the defect. Preliminary results are presented, concentrating on the prediction of stress fracture during the crucial post-operative period. It is shown that the likelihood of fracture is strongly influenced by the shape of the defect as well as its size, and also by the level of post-operative exercise. The most important finding is that bone graft materials can be successful in preventing fracture even when their mechanical properties are greatly inferior to those of bone. Future uses of this technique include pre-clinical assessment of bone replacement materials and pre-operative planning in orthopaedic surgery.     

A general requirement for FcγRIIB co-engagement of agonistic anti-TNFR antibodies

Comment on: Li F, et al. Proc Natl Acad Sci USA 2012; 109:10966-71.     

Development of tibulizumab,a tetravalent bispecific antibody targeting BAFF and IL-17A for the treatment of autoimmune disease

ABSTRACT

We describe a bispecific dual-antagonist antibody against human B cell activating factor (BAFF) and interleukin 17A (IL-17). An anti-IL-17 single-chain variable fragment (scFv) derived from ixekizumab (Taltz®) was fused via a glycine-rich linker to anti-BAFF tabalumab. The IgG-scFv bound both BAFF and IL-17 simultaneously with identical stoichiometry as the parental mAbs. Stability studies of the initial IgG-scFv revealed chemical degradation and aggregation not observed in either parental antibody. The anti-IL-17 scFv showed a high melting temperature (Tm) by differential scanning calorimetry (73.1°C), but also concentration-dependent, initially reversible, protein self-association. To engineer scFv stability, three parallel approaches were taken: labile complementary-determining region (CDR) residues were replaced by stable, affinity-neutral amino acids, CDR charge distribution was balanced, and a H44-L100 interface disulfide bond was introduced. The Tm of the disulfide-stabilized scFv was largely unperturbed, yet it remained monodispersed at high protein concentration. Fluorescent dye binding titrations indicated reduced solvent exposure of hydrophobic residues and decreased proteolytic susceptibility was observed, both indicative of enhanced conformational stability. Superimposition of the H44-L100 scFv (PDB id: 6NOU) and ixekizumab antigen-binding fragment (PDB id: 6NOV) crystal structures revealed nearly identical orientation of the frameworks and CDR loops. The stabilized bispecific molecule LY3090106 (tibulizumab) potently antagonized both BAFF and IL-17 in cell-based and in vivo mouse models. In cynomolgus monkey, it suppressed B cell development and survival and remained functionally intact in circulation, with a prolonged half-life. In summary, we engineered a potent bispecific antibody targeting two key cytokines involved in human autoimmunity amenable to clinical development.     

Monoclonal antibodies in neuro-oncology

Monoclonal antibodies (mAbs) are used with increasing success against many tumors but, for brain tumors, the blood-brain barrier (BBB) is a special concern. The BBB prevents antibody entry to the normal brain; however, its role in brain tumor therapy is more complex. The BBB is closest to normal at micro-tumor sites; its properties and importance change as the tumor grows. In this review, evolving insight into the role of the BBB is balanced against other factors that affect efficacy or interpretation when mAbs are used against brain tumor targets. As specific examples, glioblastoma multiforme (GBM), primary central nervous system lymphoma (PCNSL) and blood-borne metastases from breast cancer are discussed in the context of treatment, respectively, with the mAbs bevacizumab, rituximab, and trastuzumab, each of which is already widely used against tumor outside the brain. It is suggested that success against brain tumors will require getting past the BBB in two senses: physically, to better attack brain tumor targets and conceptually, to give equal attention to problems that are shared with other tumor sites.     

De novo discovery of antibody drugs – great promise demands scrutiny

ABSTRACT

We live in an era of rapidly advancing computing capacity and algorithmic sophistication. “Big data” and “artificial intelligence”find progressively wider use in all spheres of human activity, including healthcare. A diverse array of computational technologies is being applied with increasing frequency to antibody drug research and development (R&D). Their successful applications are met with great interest due to the potential for accelerating and streamlining the antibody R&D process. While this excitement is very likely justified in the long term, it is less likely that the transition from the first use to routine practice will escape challenges that other new technologies had experienced before they began to blossom. This transition typically requires many cycles of iterative learning that rely on the deconstruction of the technology to understand its pitfalls and define vectors for optimization. The study by Vasquez et al. identifies a key obstacle to such learning: the lack of transparency regarding methodology in computational antibody design reports, which has the potential to mislead the community efforts