Group A streptococci have evolved multiple strategies to evade human antibodies, making it challenging to create effective vaccines or antibody treatments. Here, we have generated antibodies derived from the memory B cells of an individual who had successfully cleared a group A streptococcal infection. The antibodies bind with high affinity in the central region of the surface‐bound M protein. Such antibodies are typically non‐opsonic. However, one antibody could effectively promote vital immune functions, including phagocytosis and in vivo protection. Remarkably, this antibody primarily interacts through a bivalent dual‐Fab cis mode, where the Fabs bind to two distinct epitopes in the M protein. The dual‐Fab cis‐binding phenomenon is conserved across different groups of M types. In contrast, other antibodies binding with normal single‐Fab mode to the same region cannot bypass the M protein's virulent effects. A broadly binding, protective monoclonal antibody could be a candidate for anti‐streptococcal therapy. Our findings highlight the concept of dual‐Fab cis binding as a means to access conserved, and normally non‐opsonic regions, regions for protective antibody targeting.

The generation of antibodies targeting an external threat is a hallmark of immunity. Bacteria have evolved multiple strategies to evade the effect of antibodies, making it challenging to create effective vaccines or antibody treatments. Here, we have generated antibodies derived from the memory B cells of an individual who had successfully cleared a group A streptococcal infection. The antibodies bind with high affinity to the central region on the streptococcal surface-bound M protein. One antibody could effectively promote vital immune functions, including phagocytosis and in vivo protection. Remarkably, this antibody only interacts through dual-Fab cis mode, where both Fabs bind to two distinct epitopes in a single protein. In contrast, traditional antibody binding to a single epitope does not manage to bypass the M protein's virulent effects. Dual-Fab cis binding is an unexpected mechanism for protective antibody function with important implications for vaccine and therapeutic antibody development.

Generating and analyzing overlapping peptides through multienzymatic digestion is an efficient procedure for de novo protein using from bottom-up mass spectrometry (MS). Despite improved instrumentation and software, de novo MS data analysis remains challenging. In recent years, deep learning models have represented a performance breakthrough. Incorporating that technology into de novo protein sequencing workflows require machine-learning models capable of handling highly diverse MS data. In this study, we analyzed the requirements for assembling such generalizable deep learning models by systemcally varying the composition and size of the training set. We assessed the generated models’ performances using two test sets composed of peptides originating from the multienzyme digestion of samples from various species. The peptide recall values on the test sets showed that the deep learning models generated from a collection of highly N- and C-termini diverse peptides generalized 76% more over the termini-restricted ones. Moreover, expanding the training set’s size by adding peptides from the multienzymatic digestion with five proteases of several species samples led to a 2–3 fold generalizability gain. Furthermore, we tested the applicability of these multienzyme deep learning (MEM) models by fully de novo sequencing the heavy and light monomeric chains of five commercial antibodies (mAbs). MEMs extracted over 10000 matching and overlapped peptides across six different proteases mAb samples, achieving a 100% sequence coverage for 8 of the ten polypeptide chains. We foretell that the MEMs’ proven improvements to de novo analysis will positively impact several applications, such as analyzing samples of high complexity, unknown nature, or the peptidomics field.

No abstract

Generating and analyzing overlapping peptides through multienzymatic digestion is an efficient procedure for de novo protein using from bottom-up mass spectrometry (MS). Despite improved instrumentation and software, de novo MS data analysis remains challenging. In recent years, deep learning models have represented a performance breakthrough. Incorporating that technology into de novo protein sequencing workflows require machine-learning models capable of handling highly diverse MS data. In this study, we analyzed the requirements for assembling such generalizable deep learning models by systematically varying the composition and size of the training set. We assessed the generated models' performances using two test sets composed of peptides originating from the multienzyme digestion of samples from various species. The peptide recall values on the test sets showed that the deep learning models generated from a collection of highly N- and C-termini diverse peptides generalized 76% more over the termini-restricted ones. Moreover, expanding the training set's size by adding peptides from the multienzymatic digestion with five proteases of several species samples led to a 2-3 fold generalizability gain. Furthermore, we tested the applicability of these multienzyme deep learning (MEM) models by fully de novo sequencing the heavy and light monomeric chains of five commercial antibodies (mAbs). MEM models extracted over 10000 matching and overlapped peptides across six different proteases mAb samples, achieving a 100% sequence coverage for 8 of the ten polypeptide chains. We foretell that the MEM models' proven improvements to de novo analysis will positively impact several applications, such as analyzing samples of high complexity, unknown nature, or the peptidomics field.