I know a lot of people will have seen this from other sources, but the people at Am Institut für Hygiene at Münster University have some very useful strain typing data on HUSEC publicly available here. You can see that there is an entry of HUSEC of serotype O104:H4, ST678 (same as the current outbreak), strain 01-09591 from 2001, carrying stx2 & terE but not saa; this matches the current HUSEC strains. On June 6, Münster university announced they had sequenced this older strain, however no data or analysis has been released yet.
Münster have also released a multiplex PCR for typing the outbreak strain, which incorporates primers for stx2, terD, rfbO (for O104) and flicH4 (for H4). So this will capture the presence of Shiga toxin, tellurite resistance, and O104:H4 serotype. Seems pretty comprehensive, although I’m not sure how useful the detection of tellurite resistance is since it seems to be quite widely distributed in enterohemorrhagic E. coli (see my tree of ter operon here).
Yesterday, BGI released their own typing scheme, targeting stx2 and what they call AFF_I-2 (primers F-CAGTCAGGAAAGTGGGAAAGC, R-AGGTCTGTATAGTGCCGTCTG). I’m pretty sure this is a typo and they mean AAF/I, i.e. aggregative adhesive fimbriae type I. Their primers target the aggC gene of the AAF/I gene cluster, see EMBL file showing off agg operon and primer binding sites here.
This is exactly what I suggested a few days ago based on the detection of stx2 and AAF/I fimbrial cluster (agg operon) in the outbreak strain (as opposed to the AAF/II operon found in Ec55989). The rationale for me was that this combination tells you that you have an enteroaggregative E. coli (and a specific type of aggregative fimbriae) in combination with Shiga toxin, which tells you that it is not a typical EHEC, and is more informative than any old EAEC + shiga toxin. I don’t have access to databases which would show how rare this combination is, but given that the AAF/I cluster is not present in any GenBank genomes, I would hazard a guess it is at least somewhat rare.
So it seems to me the best option would be a combination of the two schemes – take the Münster scheme and replace terD primers with aggC primers, to end up with stx2, aggC, rfbO and flicH4. This will give a definitive diagnosis of serotype O104:H4 enteroaggregative E. coli (AAF/I type) producing Shiga-toxin.