The CD33xCD123xCD70 Multispecific CD3-Engaging DARPin MP0533 Induces Selective T Cell-Mediated Killing of AML Leukemic Stem Cells.

The prognosis of patients with acute myeloid leukemia (AML) is limited, especially for elderly or unfit patients not eligible for hematopoietic stem cell (HSC) transplantation. The disease is driven by leukemic stem cells (LSCs), which are characterized by clonal heterogeneity and resistance to conventional therapy. These cells are therefore believed to be a major cause of progression and relapse. We designed MP0533, a multispecific CD3-engaging DARPin (designed ankyrin repeat protein) that can simultaneously bind to three antigens on AML cells (CD33, CD123, and CD70), aiming to enable avidity-driven T cell-mediated killing of AML cells co-expressing at least two of the antigens. In vitro, MP0533 induced selective T cell-mediated killing of AML cell lines, as well as patient-derived AML blasts and LSCs, expressing two or more target antigens, while sparing healthy HSCs, blood, and endothelial cells. The higher selectivity also resulted in markedly lower levels of cytokine release in normal human blood compared to single antigen-targeting T-cell engagers. In xenograft AML mouse models, MP0533 induced tumor-localized T-cell activation and cytokine release, leading to complete eradication of the tumors while having no systemic adverse effects. These studies show that the multispecific-targeting strategy used with MP0533 holds promise for improved selectivity towards LSCs and efficacy against clonal heterogeneity, potentially bringing a new therapeutic option to this group of patients with high unmet need. MP0533 is currently being evaluated in a dose-escalation phase 1 study in patients with relapsed or refractory AML (NCT05673057).

The trispecific DARPin ensovibep inhibits diverse SARS-CoV-2 variants.

The emergence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants with potential resistance to existing drugs emphasizes the need for new therapeutic modalities with broad variant activity. Here we show that ensovibep, a trispecific DARPin (designed ankyrin repeat protein) clinical candidate, can engage the three units of the spike protein trimer of SARS-CoV-2 and inhibit ACE2 binding with high potency, as revealed by cryo-electron microscopy analysis. The cooperative binding together with the complementarity of the three DARPin modules enable ensovibep to inhibit frequent SARS-CoV-2 variants, including Omicron sublineages BA.1 and BA.2. In Roborovski dwarf hamsters infected with SARS-CoV-2, ensovibep reduced fatality similarly to a standard-of-care monoclonal antibody (mAb) cocktail. When used as a single agent in viral passaging experiments in vitro, ensovibep reduced the emergence of escape mutations in a similar fashion to the same mAb cocktail. These results support further clinical evaluation of ensovibep as a broad variant alternative to existing targeted therapies for Coronavirus Disease 2019 (COVID-19).

Structure-Guided Design of a Peptide Lock for Modular Peptide Binders.

Peptides play an important role in intermolecular interactions and are frequent analytes in diagnostic assays, also as unstructured, linear epitopes in whole proteins. Yet, due to the many different sequence possibilities even for short peptides, classical selection of binding proteins from a library, one at a time, is not scalable to proteomes. However, moving away from selection to a rational assembly of preselected modules binding to predefined linear epitopes would split the problem into smaller parts. These modules could then be reassembled in any desired order to bind to, in principle, arbitrary sequences, thereby circumventing any new rounds of selection. Designed Armadillo repeat proteins (dArmRPs) are modular, and they do bind elongated peptides in a modular way. Their consensus sequence carries pockets that prefer arginine and lysine. In our quest to select pockets for all amino acid side chains, we had discovered that repetitive sequences can lead to register shifts and peptide flipping during selections from libraries, hindering the selection of new binding specificities. To solve this problem, we now created an orthogonal binding specificity by a combination of grafting from ß-catenin, computational design and mutual optimization of the pocket and the bound peptide. We have confirmed the design and the desired interactions by X-ray structure determination. Furthermore, we could confirm the absence of sliding in solution by a single-molecule Förster resonance energy transfer. The new pocket could be moved from the N-terminus of the protein to the middle, retaining its properties, further underlining the modularity of the system.

Curvature of designed armadillo repeat proteins allows modular peptide binding.

Designed armadillo repeat proteins (dArmRPs) were developed to create a modular peptide binding technology where each of the structural repeats binds two residues of the target peptide. An essential prerequisite for such a technology is a dArmRP geometry that matches the peptide bond length. To this end, we determined a large set (n=27) of dArmRP X-ray structures, of which 12 were previously unpublished, to calculate curvature parameters that define their geometry. Our analysis shows that consensus dArmRPs exhibit curvatures close to the optimal range for modular peptide recognition. Binding of peptide ligands can induce a curvature within the desired range, as confirmed by single-molecule FRET experiments in solution. On the other hand, computationally designed ArmRPs, where side chains have been chosen with the intention to optimally fit into a geometrically optimized backbone, turned out to be more divergent in reality, and thus not suitable for continuous peptide binding. Furthermore, we show that the formation of a crystal lattice can induce small but significant deviations from the curvature adopted in solution, which can interfere with the evaluation of repeat protein scaffolds when high accuracy is required. This study corroborates the suitability of consensus dArmRPs as a scaffold for the development of modular peptide binders.

Computationally Designed Armadillo Repeat Proteins for Modular Peptide Recognition.

Armadillo repeat proteins (ArmRPs) recognize their target peptide in extended conformation and bind, in a first approximation, two residues per repeat. Thus, they may form the basis for building a modular system, in which each repeat is complementary to a piece of the target peptide. Accordingly, preselected repeats could be assembled into specific binding proteins on demand and thereby avoid the traditional generation of every new binding molecule by an independent selection from a library. Stacked armadillo repeats, each consisting of 42 aa arranged in three α-helices, build an elongated superhelical structure. Here, we analyzed the curvature variations in natural ArmRPs and identified a repeat pair from yeast importin-α as having the optimal curvature geometry that is complementary to a peptide over its whole length. We employed a symmetric in silico design to obtain a uniform sequence for a stackable repeat while maintaining the desired curvature geometry. Computationally designed ArmRPs (dArmRPs) had to be stabilized by mutations to remove regions of higher flexibility, which were identified by molecular dynamics simulations in explicit solvent. Using an N-capping repeat from the consensus-design approach, two different crystal structures of dArmRP were determined. Although the experimental structures of dArmRP deviated from the designed curvature, the insertion of the most conserved binding pockets of natural ArmRPs onto the surface of dArmRPs resulted in binders against the expected peptide with low nanomolar affinities, similar to the binders from the consensus-design series.

Structure and Energetic Contributions of a Designed Modular Peptide-Binding Protein with Picomolar Affinity.

Natural armadillo repeat proteins (nArmRP) like importin-α or ß-catenin bind their target peptides such that each repeat interacts with a dipeptide unit within the stretched target peptide. However, this modularity is imperfect and also restricted to short peptide stretches of usually four to six consecutive amino acids. Here we report the development and characterization of a regularized and truly modular peptide-specific binding protein, based on designed armadillo repeat proteins (dArmRP), binding to peptides of alternating lysine and arginine residues (KR)n. dArmRP were obtained from nArmRP through cycles of extensive protein engineering, which rendered them more uniform. This regularity is reflected in the consistent binding of dArmRP to (KR)-peptides, where affinities depend on the lengths of target peptides and the number of internal repeats in a very systematic manner, thus confirming the modularity of the interaction. This exponential dependency between affinity and recognition length suggests that each module adds a constant increment of binding energy to sequence-specific recognition. This relationship was confirmed by comprehensive mutagenesis studies that also reveal the importance of individual peptide side chains. The 1.83 Å resolution crystal structure of a dArmRP with five identical internal repeats in complex with the cognate (KR)5 peptide proves a modular binding mode, where each dipeptide is recognized by one internal repeat. The confirmation of this true modularity over longer peptide stretches lays the ground for the design of binders with different specificities and tailored affinities by the assembly of dipeptide-specific modules based on armadillo repeats.

Structures of designed armadillo-repeat proteins show propagation of inter-repeat interface effects.

The armadillo repeat serves as a scaffold for the development of modular peptide-recognition modules. In order to develop such a system, three crystal structures of designed armadillo-repeat proteins with third-generation N-caps (YIII-type), four or five internal repeats (M-type) and second-generation C-caps (AII-type) were determined at 1.8 Å (His-YIIIM4AII), 2.0 Å (His-YIIIM5AII) and 1.95 Å (YIIIM5AII) resolution and compared with those of variants with third-generation C-caps. All constructs are full consensus designs in which the internal repeats have exactly the same sequence, and hence identical conformations of the internal repeats are expected. The N-cap and internal repeats M1 to M3 are indeed extremely similar, but the comparison reveals structural differences in internal repeats M4 and M5 and the C-cap. These differences are caused by long-range effects of the C-cap, contacting molecules in the crystal, and the intrinsic design of the repeat. Unfortunately, the rigid-body movement of the C-terminal part impairs the regular arrangement of internal repeats that forms the putative peptide-binding site. The second-generation C-cap improves the packing of buried residues and thereby the stability of the protein. These considerations are useful for future improvements of an armadillo-repeat-based peptide-recognition system.

Crystal structures of designed armadillo repeat proteins: implications of construct design and crystallization conditions on overall structure.

Designed armadillo repeat proteins (dArmRP) are promising modular proteins for the engineering of binding molecules that recognize extended polypeptide chains. We determined the structure of a dArmRP containing five internal repeats and 3rd generation capping repeats in three different states by X-ray crystallography: without N-terminal His6 -tag and in the presence of calcium (YM5 A/Ca(2+) ), without N-terminal His6 -tag and in the absence of calcium (YM5 A), and with N-terminal His6 -tag and in the presence of calcium (His-YM5 A/Ca(2+)). All structures show different quaternary structures and superhelical parameters. His-YM5 A/Ca(2+) forms a crystallographic dimer, which is bridged by the His6 -tag, YM5 A/Ca(2+) forms a domain-swapped tetramer, and only in the absence of calcium and the His6 -tag, YM5 A forms a monomer. The changes of superhelical parameters are a consequence of calcium binding, because calcium ions interact with negatively charged residues, which can also participate in the modulation of helix dipole moments between adjacent repeats. These observations are important for further optimizations of dArmRPs and provide a general illustration of how construct design and crystallization conditions can influence the exact structure of the investigated protein.

Spontaneous self-assembly of engineered armadillo repeat protein fragments into a folded structure.

Repeat proteins are built of modules, each of which constitutes a structural motif. We have investigated whether fragments of a designed consensus armadillo repeat protein (ArmRP) recognize each other. We examined a split ArmRP consisting of an N-capping repeat (denoted Y), three internal repeats (M), and a C-capping repeat (A). We demonstrate that the C-terminal MA fragment adopts a fold similar to the corresponding part of the entire protein. In contrast, the N-terminal YM2 fragment constitutes a molten globule. The two fragments form a 1:1 YM2:MA complex with a nanomolar dissociation constant essentially identical to the crystal structure of the continuous YM3A protein. Molecular dynamics simulations show that the complex is structurally stable over a 1 µs timescale and reveal the importance of hydrophobic contacts across the interface. We propose that the existence of a stable complex recapitulates possible intermediates in the early evolution of these repeat proteins.

Modular peptide binding: from a comparison of natural binders to designed armadillo repeat proteins.

Several binding scaffolds that are not based on immunoglobulins have been designed as alternatives to traditional monoclonal antibodies. Many of them have been developed to bind to folded proteins, yet cellular networks for signaling and protein trafficking often depend on binding to unfolded regions of proteins. This type of binding can thus be well described as a peptide-protein interaction. In this review, we compare different peptide-binding scaffolds, highlighting that armadillo repeat proteins (ArmRP) offer an attractive modular system, as they bind a stretch of extended peptide in a repeat-wise manner. Instead of generating each new binding molecule by an independent selection, preselected repeats - each complementary to a piece of the target peptide - could be designed and assembled on demand into a new protein, which then binds the prescribed complete peptide. Stacked armadillo repeats (ArmR), each typically consisting of 42 amino acids arranged in three α-helices, build an elongated superhelical structure which enables binding of peptides in extended conformation. A consensus-based design approach, complemented with molecular dynamics simulations and rational engineering, resulted in well-expressed monomeric proteins with high stability. Peptide binders were selected and several structures were determined, forming the basis for the future development of modular peptide-binding scaffolds.

Sigma factor mimicry involved in regulation of general stress response.

Bacteria have evolved regulatory traits to rapidly adapt to changing conditions. Two principal regulatory mechanisms to modulate gene expression consist of regulation via alternative sigma factors and phosphorylation-dependent response regulators. PhyR represents a recently discovered protein family combining parts of both systems: a sigma factor-like domain of the extracytoplasmic function (ECF) subfamily linked to a receiver domain of a response regulator. Here we investigated the mode of action of this key regulator of general stress response in Methylobacterium extorquens. Our results indicate that PhyR does not act as a genuine sigma factor but instead controls gene expression indirectly through protein-protein interactions. This is evident from the analysis of additional proteins involved in PhyR-dependent gene regulation. We demonstrated that the ECF sigma factor-like domain of PhyR interacts with a protein, designated NepR, upon phosphorylation of the PhyR receiver domain. Using transcriptome analysis and phenotypic assays, we showed that NepR is a negative regulator of PhyR response. Furthermore, we provide biochemical and genetic evidence that NepR exerts this inhibitory effect through sequestration of the ECF sigma factor sigma(EcfG1). Our data support an unprecedented model according to which PhyR acts as a mimicry protein triggering a partner-switching mechanism. Such a regulation of general stress response clearly differs from the two known models operating via sigma(S) and sigma(B). Given the absence of these master regulators and the concomitant conservation of PhyR in Alphaproteobacteria, the novel mechanism presented here is most likely central to the control of general stress response in this large subclass of Proteobacteria.

The oligomeric state of c rings from cyanobacterial F-ATP synthases varies from 13 to 15.

We isolated the c rings of F-ATP synthases from eight cyanobacterial strains belonging to four different taxonomic classes (Chroococcales, Nostocales, Oscillatoriales, and Gloeobacteria). These c rings showed different mobilities on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), probably reflecting their molecular masses. This supposition was validated with the previously characterized c(11), c(14), and c(15) rings, which migrated on SDS-PAGE in proportion to their molecular masses. Hence, the masses of the cyanobacterial c rings can conveniently be deduced from their electrophoretic mobilities and, together with the masses of the c monomers, allow the calculation of the c ring stoichiometries. The method is a simple and fast way to determine stoichiometries of SDS-stable c rings and hence a convenient means to unambiguously determine the ion-to-ATP ratio, a parameter reflecting the bioenergetic efficacy of F-ATP synthases. AFM imaging was used to prove the accuracy of the method and confirmed that the c ring of Synechococcus elongatus SAG 89.79 is a tridecameric oligomer. Despite the high conservation of the c-subunit sequences from cyanobacterial strains from various environmental groups, the stoichiometries of their c rings varied between c(13) and c(15). This systematic study of the c-ring stoichiometries suggests that variability of c-ring sizes might represent an adaptation of the individual cyanobacterial species to their particular environmental and physiological conditions. Furthermore, the two new examples of c(15) rings underline once more that an F(1)/F(o) symmetry mismatch is not an obligatory feature of all F-ATP synthases.