Analysis of multiple genomes of GAS shows that about 10% of the genome can be attributed to genetic material acquired horizontal gene transfer . Multiple mobile genetic elements as prophages, ICE elements and ancient pathogenicity islands are part of GAS metagenome [3, 24]. Lack of detected natural transformation of GAS, despite proposed mechanism mediated via quorum sensing mechanism,  stresses the importance of transduction and conjugation processes in HGT.
Since late 1970s multiple authors were studying plasmid conjugal transfer between various streptococcal species [26–28]. Later, based on sequence analyses and experimental rationale, horizontal transfer of genes/regions between GAS and GGS was implied [29–31]. Finally, recent publications report conjugative transfer of ICE elements in human and animal isolates of GAS, GBS, GGS, GCS and Streptococcus uberis [32, 33].
Our work demonstrates that genetic element RD2 from GAS strain MGAS6180 (serotype M28) can be horizontally transferred in the laboratory to other GAS strains by filter mating. The transfer frequency is comparable with inter-species transfer of ICESt3 . However, we cannot exclude that the transfer frequency was influenced by the inactivation of M28_Spy1325-1326 genes. The genes encode putative extracellular proteins and can act as aggregation factors, in particular, M28_Spy1325 has homology to enterococcal conjugative plasmid pAM373 aggregation factor . However, because we used filter mating technique that can at least partially circumvent the need of aggregation factor in the conjugation process, the lack of M28_Spy1325-1326 genes does not have to affect transfer frequency during filter mating.
Presented study provides experimental support for the idea that the presence of RD2 in GAS strains of very diverse phylogenetic backgrounds that have not shared a recent common ancestor have acquired this element by lateral gene flow occurring in nature. Our results also show that RD2-like regions are present in multiple Lancefield group C and group G strains, additional evidence for horizontal dissemination of RD2 in natural populations of streptococci. Of note, the detection of an RD2-like element in group B , C and G streptococci (this work) is consistent with early reports of the production of the R28 antigen in these organisms [5, 36].
We believe that RD2 has spread and been maintained in genetically diverse organisms in part because proteins encoded by this genetic element confer a survival advantage to the recipient organism. RD2 encodes at least seven proteins that are secreted into the extracellular environment, including several likely to participate in host-pathogen interactions such as cell adhesion. It is plausible that at least two of these proteins confer a survival premium. The best characterized is protein R28 encoded by M28_Spy1336. The RD2 protein has been shown to promote adhesion of GAS to human epithelial cells grown in vitro and confer protective immunity in a mouse model of invasive disease, together providing evidence that the R28 protein is a virulence factor [5, 6]. Another RD2 encoded gene involved in virulence is M28_Spy1325. The protein is a member of the antigen I/II family of adhesions made by oral streptococci. It is made in vivo during invasive GAS infection, and binds GP340, a heavily glycosylated protein present in human saliva . Similar to the R28 protein, immunization with recombinant purified M28_Spy1325 protect mice from experimental invasive infection, and the protein is made during human invasive infections [1, 8]. Although far less is known about the other secreted extracellular proteins made by RD2, serologic analysis indicates that M28_Spy1306, M28_Spy1326 and M28_Spy1332 also are made during human invasive infections .
Although our work did not define the exact molecular mechanism(s) mediating horizontal gene transfer of RD2, the structure of the element and its transfer by filter mating point toward conjugation as a key process. Parts of RD2 share substantial homology with ICESt1  and ICESt3  conjugative elements from S. thermophilus. ICESt1 and ICESt3 elements have homology in sequence and organization with conjugative transposon Tn916 from Enterococcus faecalis . Interestingly, a large intergenic region between M28_Spy1321 and M28_SpyM28_Spy1322 ORFs contains multiple palindromic sequences and might function as origin of transfer (oriT) as the equivalent region of Tn916 has been shown  or has been suggested to function as such .
We note that the copy number of RD2 increased rapidly and substantially following treatment of strain MGAS6180 with mitomycin C, a DNA damaging agent known to induce the SOS response and stimulate horizontal transfer of ICEs [23, 41–43].
Interestingly, Ubeda et al. have reported that other factors as antibiotic treatment can mediate SOS response in staphylococci and promote horizontal dissemination of pathogenicity island-encoded virulence factor genes . The postulated mechanism of SOS-induced induction and transfer of ICESt1/3 elements involves autoproteolysis of cI type repressor Arp1 [23, 45]. As the RD2 element encodes multiple cI type repressors  it is plausible that the mechanism of RD2 induction is mediated by SOS-induced proteolysis or autoproteolysis of one of the RD2 cI regulators. The induction of RD2 was not observed after treatment with hydrogen peroxide i.e. in the condition of oxidative stress that is known to induce phages [46–48]. That suggests rather LexA dependent mechanism induced by DNA damage.
In conclusion, RD2 is a medium host range mobile element that is shared between multiple unrelated serotypes of GAS and other pathogenic streptococcal species. As a consequence of several extracellular secreted proteins encoded by RD2, the element may confer a selective advantage on organisms that acquire this element by horizontal gene transfer.