Isolation of monoclonal antibodies from anti-synthetase syndrome patients and affinity maturation by recombination of independent somatic variants

Abstract

The autoimmune disease known as Jo-1 positive anti-synthetase syndrome (ASS) is characterized by circulating antibody titers to histidyl-tRNA synthetase (HARS), which may play a role in modulating the non-canonical functions of HARS. Monoclonal antibodies to HARS were isolated by single-cell screening and sequencing from three Jo-1 positive ASS patients and shown to be of high affinity, covering diverse epitope space. The immune response was further characterized by repertoire sequencing from the most productive of the donor samples. In line with previous studies of autoimmune repertoires, these antibodies tended to have long complementarity-determining region H3 sequences with more positive-charged residues than average. Clones of interest were clustered into groups with related sequences, allowing us to observe different somatic mutations in related clones. We postulated that these had found alternate structural solutions for high affinity binding, but that mutations might be transferable between clones to further enhance binding affinity. Transfer of somatic mutations between antibodies within the same clonal group was able to enhance binding affinity in a number of cases, including beneficial transfer of a mutation from a lower affinity clone into one of higher affinity. Affinity enhancement was seen with mutation transfer both between related single-cell clones, and directly from related repertoire sequences. To our knowledge, this is the first demonstration of somatic hypermutation transfer from repertoire sequences to further mature in vivo derived antibodies, and represents an additional tool to aid in affinity maturation for the development of antibodies.

Keywords: Aminoacyl tRNA synthetase; antibody; autoimmune disease; bioinformatics; cellular immune response; protein evolution; repertoire; single cell.

Figure 1.

Assay format used for single-cell antibody discovery. A two-step screening assay was performed. In step 1, antibody-secreting cells were co-incubated with protein A beads to capture secreted antibodies and were detected using fluorescently labeled WHEP domain-MBP and WHEP domain-deleted HARS (left). In step 2, fluorescently labeled full-length HARS was added to confirm binding to the full-length protein (right). The two phenotypes observed during screening are depicted in the schematic: specificity to the WHEP domain (top panel, Phenotype 1) and specificity outside the WHEP domain of HARS (bottom panel, Phenotype 2)

Figure 2.

Frequency and pairing of V and J genes within single cell (b) and repertoire (a) heavy chains. V genes are shown on the left and J genes are shown on the right of each diagram. V/J pairing is represented by a horizontal-colored line joining the respective V and J gene. Single-cell data was collapsed to unique chains. (A) V/J gene frequency and pairing in repertoire heavy chains from Donor 1, including both self-reactive, and non-self-reactive clones. (B) V/J gene frequency and pairing in heavy chains from recovered self-reactive single-cell clones from Donor 1

Figure 3.

Comparison of heavy chain features from single cells, collapsed to unique sequences, to 5887 unique sequences recovered in the repertoire. Significance between Donor 1 repertoire and Donor 1 single cells calculated by Kolmogorov Smirnov test. * = p < .05, ** = p < .005, ns = not significant. Comparison of: (a) nucleotide SHM, (b) heavy chain CDR3 length, (c) heavy chain positive residues per CDR3 (lysine and arginine), (d) heavy chain negative residues per CDR3 (glutamate and aspartate)

Figure 4.

(a) Alignment of Ab14 and Ab92 heavy chain V region to germline sequences IGHV3-11*06 and IGHJ4*02. Identities are represented as dots, SHM substitutions are shown. V(D)J junction represented as Xs. (b) Alignment of Ab14 and Ab92 light chain V region to germline IGKV1D-12*01 and IGKJ2*01

Figure 5.

Affinity of parent antibodies to recombinant dimeric HARS(aa1-506) at 150, 50, 16.67, 5.56, and 1.85 nM, fit to bivalent analyte model. Fit lines are displayed in red, experimental data in gray scale. Representative SPR trace of: (a) Ab92, (b) Ab14, (c) Ab92 with H97Y mutation sourced from Ab14, (d) Ab14 with T99V mutation sourced from Ab92. (e) Table of kinetic constants. (f) Summary table of sites of SHM in Ab14 and resulting affinities when transferred to Ab92. Mutations numbered according to Kabat numbering scheme

Figure 6.

Iso-affinity plot of SHM variants tested for four selected antibodies: Ab31, Ab54, Ab62 and Ab77. Indicated affinities to recombinant dimeric HARS(aa1-506) at 150, 50, 16.67, 5.56, and 1.85 nM, fit to bivalent analyte model. Parent antibodies are displayed as colored crosses, while individual mutants are shown as colored blue circles (Ab31 variants), red squares (Ab54 variants), up green arrows (Ab62 variants) and down purple arrows (Ab77 variants)