Vaccine-induced, but not natural immunity, against the Streptococcal inhibitor of complement protects against invasive disease

Highly pathogenic emm1 Streptococcus pyogenes strains secrete the multidomain Streptococcal inhibitor of complement (SIC) that binds and inactivates components of the innate immune response. We aimed to determine if naturally occurring or vaccine-induced antibodies to SIC are protective against invasive S. pyogenes infection. Immunisation with full-length SIC protected mice against systemic bacterial dissemination following intranasal or intramuscular infection with emm1 S. pyogenes. Vaccine-induced rabbit anti-SIC antibodies, but not naturally occurring human anti-SIC antibodies, enhanced bacterial clearance in an ex vivo whole-blood assay. SIC vaccination of both mice and rabbits resulted in antibody recognition of all domains of SIC, whereas naturally occurring human anti-SIC antibodies recognised the proline-rich region of SIC only. We, therefore, propose a model whereby natural infection with S. pyogenes generates non-protective antibodies against the proline-rich region of SIC, while vaccination with full-length SIC permits the development of protective antibodies against all SIC domains.


Introduction 43
Invasive disease caused by the human specific pathogen, Streptococcus pyogenes, 44 also known as group A Streptococcus (GAS), has been increasing since the 1980s 45 and is associated with mortality of approximately 20% 1,2 . Strains expressing the 46 M1 protein, encoded by emm1, are overrepresented amongst invasive isolates, 47 and account for over 30% of cases of necrotising fasciitis and streptococcal toxic 48 shock syndrome 3 . The Streptococcal Inhibitor of Complement (SIC) is an 49 extracellular protein, almost uniquely expressed by emm1 S. pyogenes, and is one 50 of several virulence factors implicated in the propensity for emm1 isolates to 51 cause severe infection 4 . 52 53 Three distinct regions of SIC have been described; an N-proximal short repeat 54 region, a central long repeat region, and a C-proximal proline-rich region 4,5 . SIC 55 binds to the C5b67 complex of complement to inhibit the formation of the 56 membrane attack complex 4,6 . SIC also inhibits the function of host antimicrobial 57 factors including lysozyme, alpha and beta defensins, secretory leucocyte 58 protease inhibitor, LL-37 and histones, and additionally has a role in bacterial 59 adherence to epithelial cells 5,7-10 . While transcriptomic and mutagenesis studies 60 have suggested a role for SIC in invasive disease in vivo 11,12 , expression of SIC has 61 not been directly linked to invasiveness of S. pyogenes in the clinical setting. 62 7 of infection and in distant tissues were quantified 48 hours after infection 130 ( Figure 3). Bacterial counts recovered from the nose were similar between both 131 sets of animals, indicating that there was no differences in dose or local bacterial 132 replication between the two groups ( Figure 3A). Compared to mice that received 133 sham vaccination, mice immunised with SIC1.300 had significantly reduced 134 bacterial counts in the spleen ( Figure 3B) and liver ( Figure 3C). Although 3/10 135 SIC-immunised mice were bacteraemic, compared to 7/10 sham immunised 136 mice, there was no significant difference in bacterial counts in the bloodstream 137 ( Figure 3D). To assess the protective role of vaccine-induced immunity against SIC ex vivo, 149 polyclonal rabbit anti-SIC serum was generated using recombinant SIC 1:300. 150 Cross reactivity with other SIC variants and native SIC from emm1 GAS culture 151 supernatant was confirmed by ELISA and western blotting (Supplementary 152 Figure S1). To determine if vaccine induced rabbit antibodies and/or naturally 153 occurring human antibodies to SIC could enhance clearance of emm1 S. pyogenes 154 8 in the ex vivo whole blood assay, rabbit and human anti-SIC antibodies were 155 affinity purified from rabbit polyclonal serum and commercially-available pooled 156 human immunoglobulin (ivIg) respectively. ELISA and western blot analysis 157 confirmed that anti-SIC antibody from both humans and rabbits, that was affinity 158 purified using SIC1.300, was able to detect full length recombinant (r) SIC of 159 three different variants and also native SIC from emm1 S. pyogenes culture 160 supernatant, suggesting that immunogenicity was not restricted to a single SIC 161 variant (Supplementary Figure S2). 162

163
Purified rabbit and human anti-SIC antibodies were added to human whole 164 blood from healthy individuals and growth of emm1 S. pyogenes was assessed 165 over 3h using a modified Lancefield assay. Rabbit anti-SIC antibody reduced 166 growth of the emm1 isolates H584 and AP1 compared to rabbit IgG isotype 167 control antibody ( Figure 5A). In contrast, human anti-SIC IgG did not inhibit 168 growth of the emm1 isolate H584 in whole human blood, compared to a control 169 antibody ( Figure 5B). We further evaluated naturally occurring SIC antibodies 170 using a panel of human sera previously determined to have high anti-SIC titres 16 . 171 There was no correlation between anti-SIC titre and bacterial growth inhibition 172 when heat-inactivated human serum from antenatal donors 16 was co-incubated 173 with emm1 S. pyogenes growing in human whole blood (Supplementary Figure  174 S3), further indicating that natural anti-SIC antibodies in human serum do not 175 promote opsonophagocytic killing of S. pyogenes. Together with the in vivo data 176 from mice, the findings indicated that immunisation of mice or rabbits with SIC 177 results in antibodies that have the potential to protect against emm1 S. pyogenes 9 infection. In contrast, naturally occurring human anti-SIC antibodies lacked the 179 protective activity that was observed for vaccine-induced antibody. 180 181 Human, rabbit and mouse anti-SIC antibodies detect different fragments of 182

SIC 183
To determine the basis for the apparent difference between natural human anti-184 SIC antibodies, and vaccine-induced rabbit or mouse anti-SIC antibodies, the 185 regions of SIC recognised by each of the anti-SIC antibodies were studied using 186 polypeptide fragments of SIC (fragments 1, 2 and 3), which are based on SIC from 187 the emm1 S. pyogenes strain AP1 (Supplementary Figure S4). Whilst purified 188 rabbit anti-SIC antibodies were able to recognise all three SIC fragments by an 189 ELISA-based assay, purified natural human anti-SIC was able to detect only 190 fragment 3, with limited detection of fragment 1 or fragment 2 ( Figure 6A). 191 Serum from mice that had been immunised with full length rSIC1.300 for the 192 earlier infection challenge experiments, also detected all three SIC fragments 193 similar to findings in the rSIC1.300-immunised rabbit serum ( Figure 6B). These 194 findings were confirmed by Western blot; rabbit and mouse anti SIC detected all 195 three fragments while human anti-SIC detected only fragment 3 ( Figure 6C). 196 Furthermore, when the anti-SIC response was quantified from individual donor 197 human antenatal sera, the predominant response was against SIC fragment 3 198 ( Figure 6D). 199 200 Immunisation with SIC fragments does not protect against invasive disease 201 To ascertain whether immunity to any single SIC fragment would be sufficient to 202 induce protective immunity, mice were immunised with recombinant SIC 203 fragment 1, fragment 2, fragment 3 or a sham vaccine containing PBS and 204 adjuvant. Fragment 3 immunization was complicated by an unexpected 205 hypersensitivity reaction in 5/8 mice. Following immunisation, strong reactivity 206 was obtained against SIC fragment 1 and fragment 3 respectively with serum 207 from mice immunised with the homologous SIC fragment when analysed by 208 ELISA ( Figure 7A In contrast to vaccine-induced immunity, natural human immunity to SIC is 227 directed against only one domain of SIC, and this is insufficient to confer 228 immunity. 229 230 Multiple functions have been attributed to SIC which all appear to aid bacterial 231 evasion of host innate immunity 6,8,9 . The upregulation of sic by invasive emm1 232 isolates that had undergone a mutation in the bacterial two-component regulator 233 covRS 11 suggested that SIC may contribute to S. pyogenes invasiveness as part of 234 the covRS regulon. Whilst previous studies have examined the role of the 235 variation in sequence and size of SIC 17-21 , we instead hypothesised that variation 236 in SIC expression levels would reflect the invasive phenotype of a strain. SIC 237 expression by single bacterial isolates has been quantified in two separate 238 reports 8,9 , however, to our knowledge this is the first report to examine SIC 239 expression in a wider collection of invasive and non-invasive emm1 GAS clinical 240 isolates. Although SIC expression levels varied widely, levels were not overall 241 significantly higher amongst invasive isolates. Of note, invasive isolates 242 represented almost two thirds of the strains investigated and it is likely that only 243 some have mutations in covRS. Among a subset of isolates for which covRS 244 sequencing was undertaken we did however observe significantly heightened 245 expression of SIC. We have, for the first time, also demonstrated that SIC is 246 detectable in infected murine thigh tissue, which was the site of infection, in 247 infected animals. We attempted to detect SIC in muscle from a patient with Whilst human anti-SIC antibodies are abundant in populations, our data suggests 280 that these antibodies are not protective. One possible explanation for this is that 281 following immunisation with full length SIC, protective antibodies are raised to 282 epitopes throughout the mature SIC protein, however, following natural infection, 283 antibodies are only generated against epitopes within fragment 3 of SIC. A 284 previous study using sera from 29 individuals, identified ten linear epitopes in 285 SIC1.01 that were identified by ≥ 50% of the human anti-SIC sera. These 286 epitopes were at sites in SIC in which polymorphisms commonly occur, and were 287 evenly distributed within the equivalent of SIC fragments 1, 2 and 3 14 . When 288 this was further analysed by phage display using two sera to detect natively 289 folded SIC peptides, 7/8 peptides recognised by one serum and 11/12 peptides 290 recognised by the other serum spanned regions present in the equivalent of SIC 291 fragment 3 (from residue 168 onwards). Our data using ELISA and western blot 292 confirm that epitopes within fragment 3 are the most readily recognised by 293 human anti-SIC that had been purified from ivIg pooled from over 1,000 donors 294 23 295 296 The reasons that human anti-SIC antibody responses are directed to a single 297 domain only are unclear. One possibility is that anti-SIC responses represent 298 cross-reactive responses to another, unrelated antigen, that has structural 299 similarity to fragment 3. Notably, SIC readily undergoes proteolysis by human 300 proteases such as human neutrophil elastase and bacterial proteases such as 301 14 SpeB 14,24 . The digestion of SIC by unknown bacterial or host factors within 302 human saliva 25 provides an alternative mechanism by which SIC fragments, 303 rather than full length SIC, might be presented to the immune system. 304

305
The lack of protection following immunisation with separate SIC fragments was 306 surprising, especially considering that several immune inhibitory functions of 307 SIC have been localised to the SRR and LRR contained within SIC fragments 1 and 308 2 respectively 5 . Indeed, immunisation with SIC fragment 2 elicited minimal 309 antibody response, despite epitopes in this fragment being detected with 310 antibodies raised against full length SIC, and also human anti-SIC serum in a 311 previous study 14 , suggesting that this fragment may form part of a discontinuous 312 epitope. Data generated using various biophysical methods indicate that SIC 313 contains low levels of regular secondary and tertiary structures (unpublished 314 data); this could mean that discontinous epitopes form long range contacts 315 resulting in either stable or dynamic tertiary structure assemblies. Resolution of 316 the complete folded structure of SIC would provide a better understanding of the 317 nature of these epitopes. Interestingly, few immune inhibitory functions of SIC 318 have been localised to the PRR of SIC contained within fragment 3. Thus, an 319 alternative explanation to the varying immunogenicity of SIC domains is that, in 320 humans, SIC fragments 1 and 2 bind to host ligands concealing these regions 321 from the host, whilst epitopes in SIC fragment 3 are abundantly available. Whilst 322 SIC binds to both human LL-37 and mouse cathelicidin 12 , previous studies 323 assessing other ligands of SIC have used only human proteins 5,8,9,22,24 . It 324 remains unclear whether SIC binds to other mouse proteins and hence the lack of 325 15 immunogenicity of fragment 2 compared to fragment 1 may be due to 326 differential binding to mouse proteins. To assess the antibody response of immunised mice, 96-well polystyrene plates 442 (Nunc, ThermoScientific, MA, USA) were coated with 100 ng of full length 443 SIC1.300 or SIC fragments 1, 2 or 3 overnight at 4°C, washed, blocked for one 444 hour with 3% normal goat serum (Sigma-Aldrich) diluted in PBS-0.1% Tween20 445 (PBST). Test sera were then added at a range of dilutions. Binding was detected 446 using HRP-conjugated goat anti-mouse IgG (Abcam) and incubated for one hour 447 at room temperature. The substrate (ONPG, Sigma-Aldrich) was added to wells, 448 the reaction was stopped with 3N HCl, and the OD A492 read with a μQuant 449 spectrophotometer (Biotek, VT, USA).
To compare cross detection of 450 recombinant SIC variants or SIC fragments by rabbit polyclonal anti-SIC1.300 451 serum or purified SIC antibody, plates were coated with SIC1.02, SIC1.301 and 452 SIC1.300 or SIC fragments 1,2 and 3 and binding was detected using 1 in 25,000 453 dilution of HRP-conjugated goat anti-rabbit IgG (Life Technologies, Paisley, UK). 454 To determine detection of SIC fragments by human purified anti-SIC antibody or 455 antenatal serum, plates were coated with SIC1.300 or SIC fragments 1, 2 and 3, and 456 binding detected using 1 in 25,000 dilution of HRP-conjugated goat anti-human 457 IgG (Sigma-Aldrich). 458 following blocking in PBST with 5% skimmed milk powder, SIC was detected by 468 probing with 1 in 10,000 dilution of rabbit polyclonal anti-SIC1.300 for two hours at 469 room temperature. Proteins were visualized by incubation in 1 in 50,000 dilution of 470 HRP-conjugated goat anti-rabbit IgG (Life Technologies) for one hour followed by 471