Recombination Creates Human Beta2c Coronavirus Sub-clades

Recombinomics Commentary 19:30
March 29, 2013

NAMRU-3 has released a full sequence (Jordan-N3) for nCoV from the first known fatality, a nurse (45F) in the Jordan SARS-like cluster linked to an ICU.  The sequence was most closely related to the sequence (EMC/2012) from the first confirmed case (60F), who was from Bisha, Saudi Arabia (KSA) and died in Jeddah.  In addition to these two full sequences, the Health Protection Agency (HPA) in the UK had sequenced two additional cases, a Qatari (49M) who developed symptoms while performing Umrah in August 2012, and a UK resident who also developed symptoms while performing Umrah in January, 2013 (England1 and England2, respectively)..

The first three sequences defined a consensus sequence which had an identity of more than 99.9% with the two England sequences (21 differences + 6 base deletion and 26 differences, respectively).  In contrast, the EMC/2012 sequence had an identity of less than 99.8% due to 70 polymorphisms which were concentrated in the orf 1ab gene.  The Jordan-N3 sequence matched most of the orf 1ab polymorphisms, but had many additional unique polymorphisms.  The 79 polymorphism in Jordan-N3 included 39 that matched EMC/12, two that matched England1, and 38 that were unique.

The distribution of the 70 Jordan-N3 polymorphisms was far from random.  Between positions 3134 and 10982 (all of which are in orf 1ab) 24/26 Jordan-N3 polymorphisms match EMC/2012.  In contrast, none of the 17 polymorphisms between positions 20208 and 26457 match EMC/2012 (one matches England1).  This clustering signals evolution via homologous recombination.  This same clustering was seen for EMC/12, where 24/27 of the polymorphisms between 3134 and 10982 match Jordan-N3 while 18/20 polymorphisms between 13678 and 266223 are unique.

The matches in orf 1ab are markedly different that the polymorphisms in the S gene.  EMC/12 has only two polymorphisms, both of which areunique and synonymous.  Jordan-N3 has 7 polymorphisms in the S gene and all are unique (and 4 are non-synonymous).

The matches between  EMC/2012 and Jordan-N3 define a separate sub-clade that may be circulating in western KSA, and reports of mild nCoV in Jeddah raise concerns that this sub-clade may be linked to a large number of nCoV infections.  In contrast, England1 and England2 may signal a separate clade circulating in eastern KSA.  Partial sequences from the first case (45M) in Riyadh are identical to the consensus sequence and therefore not informative.  This is also true for one of the partial sequences from the Essen case (the second Qatari patient, 45M, who was treated in Essen, Germany).  The second partial sequence includes a polymorphism that matches the consensus for a position (29714) that has a shared change in EMC/2012 and Jordan-N3 (T29714A).

Thus, recombination has generated two distinct nCoV sub-clades, which may be circulating in geographically distinct regions in KSA. Release of full sequences from the Riyadh and Essen cases would help define these sub-clades, as would sequences from the mild cases in Jeddah.