All animal studies conducted comply with federal and institutional (the Committee on the Ethics of Animal Experiments of the Nagoya City University) guidelines. The protocol was approved by the Committee on the Ethics of Animal Experiments of the Nagoya City University (Permit Number: H23M-07). All efforts were made to minimize suffering.

Streptococcus pyogenes strains 1529 and GT01 isolated as causative organisms from invasive diseases patients in Japan possess NADase active SPN [18,20]. S. pyogenes (GAS) strain SF370, which is prevalent as the database reference isolate (accession NC_002737), was provided by the courtesy of J. J. Ferretti [21,22]. Streptococcal strains were cultured in brain heart infusion (E-MC62, EIKEN Chemical Co., Tokyo, Japan) supplemented with 0.3% yeast extract (BD, Sparks, MD, USA) (BHI-Y) broth unless otherwise described.

The spn_SF370 of S. pyogenes strain SF370 and the other control DNAs were amplified by PCR with Extaq DNA polymerase (Takara Bio, Ohtsu, Japan) using the corresponding primers listed in Table 1. pGEM^®-T Easy vector system (Promega, Madison, WI, USA) was used for cloning of the PCR products purified by Gel extraction kit (Qiagen, Hilden, Germany). The high copy-number pGEM^®-T Easy vector contains T7 and SP6 RNA poly-

Table 1. Sequences of primers used in this study.

Continued

^aName of insert DNA fragment amplified by PCR (the expected size), used primer name and the sequence were described. The two PCR products of recA_SF370 and spy1193₁₅₂₉ contain a part of the gene, respectively, whereas the others have the hole genes indicated. spn_SF370 and recA_SF370 are PCR products from strain SF370. spn-ifs_GT01, and covRS_GT01 are PCR products derived from strain GT01. spy1193₁₅₂₉, covRS₁₅₂₉, vicRK₁₅₂₉, and vicK₁₅₂₉ are PCR products derived from strain 1529. S. pyogenes strain SF370 encode NADase-inactive SPN, whereas strains 1529 and GT01 encode NA-Daseactive SPN.

merase promoters flanking a multiple cloning region within the α-peptide coding region of the enzyme β-galactosidase http://www.promega.com/~/media/files/resources/protocols/technical%20manuals/0/pgem-t%20and%20pgem-t%20easy%20vector%20systems%20protocol.pdf?la = en. In the pGEM^®-T Easy vector system, recombinant clones are allowed to be directly identified by blue/white color screening on indicator plates.

For spn_SF370-cloning, we obtained three recombinants that have the correct size insert in the corresponding plasmids (named pGEM-spn_SF37026, pGEM-spn_SF37032, and pGEM-spn_SF37013; see Result section for additional detail on the creation of these plasmids).

Primestar Taq DNA polymerase (Takara) was used for the inverse PCR described previously [18]. Primers used are listed in Table 1. PCR product was self-ligated and used to transform E. coli strain DH5α.

The inserts of pGEM-spn_SF37026, pGEM-spn_SF37032, and pGEM-spn_SF37013 were digested with EcoRI, and subcloned into pLZ12-Km2 to yield pLZ-spn_SF37026, pLZspn_SF37032, and pLZ-spn_SF37013, respectively.

E. coli JM109 was used to propagate plasmid constructions. Non-polar inactivated mutant of spn was constructed via double-crossover allelic replacement in the chromosome of S. pyogenes SF370. To construct the plasmid for the spn knockout mutant, the 5’ end of spn (fragment 1) was amplified with oligonucleotide primers ngaGT-n1 (5’-GGCTAGCGAACAGATGTGAAGGTTCTG-3’) with an NheI restriction site and ngaGT-c1 (5’-TCCCCCGGGTTTCTCATGTAAACCACCT-3’) with an SmaI restriction site, and the 3’ end of spn (fragment 2) was amplified with ngaGT-n2 (5’-TCCCCCGGGATAGGAAGTAACAATATGT-3’) with an SmaI restriction site and ngaGT-c2 (5’-GGACTAGTATGTTAGCTTTCAATTGGGT-3’) with an SpeI restriction site. Oligonucleotides ngaGT-n1, ngaGT-c1, ngaGT-n2 and ngaGT-c2 contained a restriction site for NheI, SmaI, SmaI and SpeI, respectively, (shown in underline in the primer sequence). Fragment 2 was digested with SmaI and SpeI for insertion into multi-cloning site 2 of the pFW12 plasmid [23]. The resulting plasmid was then digested with NheI and SmaI, and both the spc2 DNA fragment containing aad9 (promoter less spectinomycin resistant gene), which was obtained from a SmaI digested fragment of pSL60-2 [24], and the NheI-SmaI-digested fragment 1 were inserted. This plasmid, pFW12::(spn::aad9), was a suicide vector for S. pyogenes. For the preparation of competent cells, strain SF370 was harvested at earlyto mid-log phase (OD₆₆₀ = 0.4 to 0.5) and washed twice with 0.5 M sucrose buffer. The constructed suicide vector pFW12::(spn::aad9) was used to transform strain SF370 by electroporation. The conditions of electroporation were 1.25 kV/mm, 25 μF capacitance and 200 Ω resistance, using Gene Pulser II (Bio-Rad, Hercules, CA, USA). After incubation at 37˚C for 3 h, competent cells were spread onto BHI agar plates containing 0.3% yeast extract and spectinomycin (final concentration 100 μg/ml). Selected colonies on the plates were cultured. Cultured bacteria were washed once with saline, resuspended in 10 mM Tris, 1 mM EDTA and boiled for 10 min. Genomic DNA was obtained from the supernatant of boiled bacteria. The double-crossover replacement was analyzed using genomic DNA by PCR and successful double-crossover replacement was further confirmed by DNA sequencing.

The ability of S. pyogenes to cause local skin lesions and necrosis in mice after skin inoculation was assessed using a procedure similar to that described elsewhere [25]. In brief, 3-week-old female ICR mice (10 - 12 g) were anesthetized with sevoflurane, and the skin of the left flank was bared by separating hair with alcohol swab, unless otherwise indicated. Bacteria (0.2 ml; 2 × 10⁷ cfu per mouse) grown in BHI-Y were injected with a 27- gauge needle just under the surface of the skin so that a superficial bleb was raised immediately below the skin surface. The number of colony-forming units injected was verified for each experiment by plating bacteria on BHI-Y or sheep blood agar plates (with or without kanamycin) and counting colony-forming units. Lesion sizes (length × width) were measured, with the length determined as the longest dimension of the lesion at day 3 or at the death time point.

Bacteria were recovered from the mice which survived until day 8. For mouse blood samples: 100 µl of blood from the heart was spread on a sheep blood agar plate. For the spleen samples: Spleen was homogenized and suspended with 100 µl of PBS and spread on a sheep blood agar plate. β-hemolysis positive colonies were calculated. Two colonies per plate were randomly selected and inoculated into BHI-Y broth supplemented with or without 50 µg/ml spectinomycin for microscopic analysis to confirm coccus morphology and chain arrangements characteristic of Streptoccoal species.

All animal procedures were approved by the Institutional Animal Care and Use Committee at Nagoya City University.

Data collected for virulence to mouse (survival days) were assessed using a log-rank comparison described previously [20]. R software was used for statistical analysis http://bioinf.wehi.edu.au/software/russell/logrank/. P-value ≤ 0.05 was considered significant.

S. pyogenes strain SF370 is a representative among NADase negative strains. In order to evaluate the role of NADase-inactive SPN, we firstly attempted to clone the spn_SF370 gene of the strain SF370 into a pGEM^®-T Easy vector in E. coli strain DH5α without support of the ifs_SF370 gene. The vector and the strain DH5α are compatible with the blue/white color screening for recombinants. Typically, we see about half of the transformants showing white color under our experimental condition when a DNA insert without toxicity to E. coli is used. We show data from two representative experiments in Table 2 (40.9% and 44.4% white colonies for recA_SF370 and spy1193₁₅₂₉, respectively; these genes have been cloned for our other studies around the same time as this study). In contrast, the four cloning experiments using the insert encoding spn_SF370 showed only 8.7%, 27.6%, 31.8% and 0.8% white colonies (Table 2). Eighty-five of the white colonies derived from the spn_SF370-insert were randomly selected for further plasmid analysis (named as pGEM-spn_SF3701 to 85), and only three colonies (4%) possessed the correct size insert (1.5 kbp) in the corresponding plasmids (pGEM-spn_SF37026, pGEM-spn_SF37032, and pGEM-spn_SF37013; see Table 3 and Figure S1). In contrast, in the experiments using control inserts, more than 50% of white colonies had the correct size inserts: 94%, 67%, 100%, 100%, 86%, 57% and 88% for spn-ifs_GT01, recA_SF370, spy1193₁₅₂₉, covRS_GT01, covRS₁₅₂₉, vicRK₁₅₂₉ and vicK₁₅₂₉, respectively (Table 3).

For the spn_SF370, the insert of the three plasmids (pGEM-spn_SF37026, pGEM-spn_SF37032, and pGEMspn_SF37013) were sequenced, and the following mutations were found: substitution of the start codon ATG to ACG in pGEM-spn_SF37026, the second codon AGA (Arg) to TGA (stop) in pGEM-spn_SF37032, and an adenine nucleotide was substituted to a guanine (G) at nucleotide 34 upstream from the adenine (A) of the start codon in pGEM-spn_SF37013, respectively (see most right column in Table 3). We propose these mutations may have been introduced for the following reasons: 1) A spontaneous mutation is often inserted in DNA fragment amplified by PCR or 2) spn_SF370 gene is toxic to E. coli cells, so mutations to make the gene inactive were given for a natural survival advantage. In order to examine those possibilitiesTable 2. The numbers of blue/white colonies.

^aPCR products of recA_SF370 and spy1193₁₅₂₉ contain a part of the gene, respectively, whereas the others have the hole genes indicated. ^bThe numbers of white (W), blue (B), and the total (W + B) colonies. ^c% white colonies. ^dsum of the colony numbers from 4 experiments.

Table 3. Cloning of the insert containing spn_SF370 gene into pGEM-T easy vector.

^aAs a control for spn_SF370, seven examples spn-ifs_GT01, recA_SF370, spy1193₁₅₂₉, covRS_GT01, covRS₁₅₂₉, vicRK₁₅₂₉, and vicK₁₅₂₉ were shown. See Table 1 about information for the insert DNAs. ^bZ-test was used in order to compare with spn_SF370. ^cNumber of insert without any mutation/number of sequenced insert DNA.

we attempted to change the mutated nucleotides back to the original nucleotides using the inverse PCR method described previously [18]. We performed inverse PCR with primers nga(SF370)-F and nga(SF370)-R (Tables 1 and 4) constructed to substitute the mutated second codon of the spn_SF370 on pGEM-spn_SF37032 to the original

Table 4. Physical maps of primers used for inverse PCR were shown.

The mutated nucleotides were attempted to change back to the original nucleotides. Primers used for the inverse PCR were shown by arrows. The primer’ nucleotide sequences were also shown in Table 1. Unsuccessful substitutions were shown in italic type. ^aOriginal nucleotide sequence of the junction site. The adenine and thymine nucleotides which were substituted in No. 32 and No. 26, respectively, were shown as bold “A” and “T”. The start (ATG) and second (AGA) codons were underlined. ^bThe mutated thymine nucleotide of the spn_SF370 on pGEM-spn_SF37032 was shown as bold “T”. ^cThe mutated cytosine nucleotide of the spn_SF370 on pGEM-spn_SF37026 was shown as bold “C”. ^dOriginal nucleotide sequence of the junction site. The adenine (A) nucleotide, which was substituted in No. 13, was shown in bold type. ^eThe mutated guanine nucleotide on the pGEM-spn_SF37013 was shown as bold “G”. ^fOriginal sequence of the junction site. The adenine nucleotide, which was substituted in No. 11 and No. 28, was shown as bold “A”. The 267^th codon (ACC) was underlined. ^gThe mutated guanine nucleotide of the vicK₁₅₂₉ on pGEM-vicRK₁₅₂₉11 was shown as bold “G”. ^hThe mutated guanine nucleotide of the vicK₁₅₂₉ on pGEM-vicK₁₅₂₉28 was shown as bold “G”.

codon AGA (R). The amplification product was selfligated and used to transform E. coli strain DH5α. Plasmids were prepared from randomly selected 14 transformants (named as 32 - 1 to 32 - 14). Five plasmids (32 - 2, 32 - 4, 32 - 8, 32 - 9 and 32 - 12) appeared to possess the correct size insert, whereas the other nine (32 - 1, 32 - 3, 32 - 5, 32 - 6, 32 - 7, 32 - 10, 32 - 11, 32 - 13, and 32 - 14) had the smaller size inserts based on the result seen during agarose gel electrophoresis (data not shown). In addition to the inserts of the five passed plasmids (32 - 2, 32 - 4, 32 - 8, 32 - 9 and 32 - 12), one of the dropped-out plasmids (32 - 1) which was added as a representative (internal) negative-control were sequenced. As shown in Table 4, the 32 - 1 had a large deletion and the other five contained a nucleotide mutation or deletion at the junction site. Additionally for pGEM-spn_SF37026, and pGEMspn_SF37013, we attempted to change the mutated nucleotides back to the original nucleotides by using same method with primers nga (SF370)-F and nga (SF370)-R, or primers nga(SF370)-F2 and nga(SF370)-R2 (Table 4). Two of the seven pGEM-spnSF₃₇₀26 derivative plasmids (26 - 1and 26 - 2) prepared from randomly selected transformants (named as 26 - 1 to 26 - 7) appeared to possess the correct size insert. However, both plasmids had a nucleotide deletion at the junction site (Table 4). For pGEM-spn_SF37013, four (13 - 1, 13 - 2, 13 - 4, and 13 - 5) of seven plasmids prepared from randomly selected transformants appeared to possess the correct size insert. However, one plasmid (13 - 4) was same as the original pGEM-spn_SF37013 and the other three possessed a nucleotides deletion or a nucleotide mutation at the junction site (Table 4). We have observed successful substitution of a corresponding nucleotide for more than 50% of the transformants checked in other recent experiments. Two representative examples are shown below. As described above (see Table 3), the four and the seven types of pGEM-T easy derivatives having vicRK₁₅₂₉ and vicK₁₅₂₉ have been previously constructed, respectively (named as pGEM-vicRK₁₅₂₉11, pGEM-vicRK₁₅₂₉12pGEM-vicRK₁₅₂₉13, pGEM-vicRK₁₅₂₉15 and pGEMvicK₁₅₂₉22, pGEM-vicK₁₅₂₉23, pGEM-vicK₁₅₂₉24, pGEMvicK₁₅₂₉25, pGEM-vicK₁₅₂₉26, pGEM-vicK₁₅₂₉27, pGEMvicK₁₅₂₉28: each plasmid has a mutation(s) in somewhere of vicRK₁₅₂₉ or vicK₁₅₂₉). Among the plasmids, pGEMvicRK₁₅₂₉11 and pGEM-vicK₁₅₂₉28 which have both a mutation changing the codon 276 of vicK₁₅₂₉ gene from a ACC (encoding “T”) to a GCC (encoding “A”) were used for the control experiments performed with primers vicK-F and vicK-R (Table 4). We observed a successful substitution of the corresponding nucleotide in one of two transformants (11 - 1 and 11 - 2) analyzed for pGEM-vicRK₁₅₂₉11, two of two transformants (28 - 1 and 28 - 2) in pGEM-vicK₁₅₂₉28 (Tables 4 and 5). In the case of spn_SF370 gene_, we never observed the expected nucleotide change when using any of three plasmids as template described above (Tables 4 and 5). These results suggest that spn_SF370 gene is toxic to E. coli cells.

In strain SF370, the ifs_SF370 allele has a nonsense mutation in the codon for leucine 24 to produce a truncated open reading frame [11]. In order to determine whether the truncated ifs_SF370 open reading frame can successfully inhibit the toxicity of spn_SF370 gene in E. coli cells, we attempted to clone a spn-ifs_SF370 operon into the pGEM^®- T Easy vector. The original forward primer ngaGTn1Nhe previously used for cloning of spn_SF370 gene and three altered reverse primers (nga-c8xho, IFS-R(EcoRI), and slo-c2) to include ifs were tested (Table 1). However, we did not obtain any recombinants having the expected insert in the size (data not shown).

We attempted to clone spn_SF370 by using plasmid pLZ12- Km2 instead of pGEM^®-T Easy. pLZ12-Km2 which has a rolling circle type of replication can be successfully maintained in both E. coli and S. pyogenes [26]. Firstly, each insert DNA of pGEM-spn_SF37026, pGEM-spn_SF37032, and pGEM-spn_SF37013 was subcloned into pLZ12-Km2 (named as pLZ-spn_SF37026, pLZ-spn_SF37032, and pLZspn_SF37013 respectively). By using these plasmids as template for inverse PCR, we attempted to change the mutated nucleotides back to original nucleotides. Ligated DNA was introduced into E. coli strain DH5α and S. pyogenes strain 1529. We obtained a handful of and many transformants in E. coli and S. pyogenes, respecttively. Therefore, all E. coli transformants obtained were investigated, while only a subset of transformants in S. pyogenes were further investigated. We did not observe successful substitution in any of the E. coli transformants

Table 5. Substitution of the mutated nucleotides back to the original one.

^aPlasmids used for inverse PCR. ^bNumber of plasmid having the successful substitution/number of the analyzed plasmid (%). N/A: not applicable.

regardless of the plasmid templates used (Table 5). We did observe successful substitution in the S. pyognenes transformants for all three of the plasmid templates used (Table 5). Finally, the spn_SF370 gene was successfully cloned only when S. pyogenes strain 1529 was used as host. These results suggest that the insert DNA encoding spn_SF370 has some toxic effect to E. coli, but not S. pyogenes.

The toxicity of spn_SF370 could be related with Riddle et al.’s claim that SPN has a NADase-independent function. In order to further explore this hypothesis, spn_SF370 gene was replaced with an antibiotics marker and the resulting strain SF370∆spn was used to infect mice.

When the SF370∆spn was inoculated in mouse skin, the mortality (36%) of the infected mice was not significantly different from the infection with the parental strain SF370 (14%) (p = 0.214, see Table 6). In addition, there were not significant differences in the lesion sizes of mouse skin infected with either SF370∆spn or strain SF370 (data not shown). Furthermore, bacteria were recovered from blood and spleen of the surviving mice on day 8 (Table 7). The bacterial number recovered from the mice infected with SF370∆spn was not reduced compared with the case infected with the parental strain SF370.