MERS-CoV potential animal reservoirs and related bat coronaviruses

Contents


What is the nearest non-human host relative?

A paper published on Aug 22nd reports a short fragment (182 nucleotides in length) of coronavirus sequence recovered from a sample from an individual Taphozous perforatus or Egyptian tomb bat that was collect a short distance from the home and work location of the first reported case of MERS-CoV infection (Bisha in Western Saudi Arabia). This sequence is reported to be identical across its 182 nucleotides with the same bit of the MERS-CoV genome sequenced from this patient (referred to as EMC-2012).

Memish ZA, Mishra N, Olival KJ, Fagbo SF, Kapoor V, Epstein JH, et al. (2013) Middle East respiratory syndrome coronavirus in bats, Saudi Arabia. Emerg Infect Dis. http://dx.doi.org/10.3201/eid1911.131172

Some analysis of this finding and what it might tell us about the direct source of MERS-CoV infections is here.

The previous closest phylogenetic match was discribed in this paper:

Ithete NL, Stoffberg S, Corman VM, Cottontail VM, Richards LR, Schoeman MC, et al. (2013) Close relative of human Middle East respiratory syndrome coronavirus in bat, South Africa. Emerg Infect Dis. Oct 2013.

The bat is of the species Neoromicia zuluensis, and was sampled in South Africa. This paper also presents some longer fragments of sequence from the group of European pipistrelle bat viruses that were the former closest non-human viruses to MERS-CoV. The new study presents 816 nucleotide partial RdRp (the virus's polymerase enzyme) with nucleotide positions 14544 - 15359 in the ORF1ab gene of MERS-CoV. This is a highly conserved region of the genome and is used to look at the deep relationships between coronaviruses subtypes and as a diagnostic test. The study also has a shorter fragment (249 nucleotides) that spans the end of the spike (S) gene and into the NS3a gene (positions 25099 - 25347 in the MERS-CoV genome). Although much shorter, this is a much more variable region of the genome and helps with the closer phylogenetic relationships.

The phylogeny constructed from the partial RdRp as shown in the Ithete et al. study:

13-0946-F1.jpg

The new virus is labelled 'BtCoV/PML/2011/Neo_zul/RSA/2012' (this is a slight typo in the paper as the year of collection was 2011). The new sequences have the Genbank accession number, KC869678, but at the time of writing, had not been released. I am grateful to Felix Drexler of The University of Bonn Medical Centre for sending me the sequences.

RdRp genetic differences

between: and: amino acid
differences
272 aa
nucleotide
differences
816 bp
nucleotide divergence
substitutions/site
BtCoV/PML/2011/Neo_zul/RSA/2011
Neoromicia zuluensis bat CoV
www.ncbi.nlm.nih.gov/nuccore/KC869678
KSA/Bisha/EMC_2012
Human, MERS-CoV
www.ncbi.nlm.nih.gov/nuccore/JX869059
1 55 0.067
KSA/Bisha/EMC_2012
Human, MERS-CoV
www.ncbi.nlm.nih.gov/nuccore/JX869059
Qatar/Doha/England1/2012
Human, MERS-CoV
www.ncbi.nlm.nih.gov/nuccore/KC667074
0 2 0.0025

This demonstrates the extreme selective constraints there are on the RdRp gene region with only one amino acid difference between the new bat virus and MERS-CoV but 55 nucleotide changes. Between the most divergent pair of MERS-CoV there are only 2 nucleotide differences and no amino acid differences.

Spike/NS3a genetic differences

between: and: amino acid
differences
83 aa
nucleotide
differences
249 bp
nucleotide divergence
substitutions/site
BtCoV/PML/2011/Neo_zul/RSA/2011
Neoromicia zuluensis bat CoV
www.ncbi.nlm.nih.gov/nuccore/KC869678
KSA/Bisha/EMC_2012
Human, MERS-CoV
www.ncbi.nlm.nih.gov/nuccore/JX869059
11 36 0.14
KSA/Bisha/EMC_2012
Human, MERS-CoV
www.ncbi.nlm.nih.gov/nuccore/JX869059
Qatar/Doha/England1/2012
Human, MERS-CoV
www.ncbi.nlm.nih.gov/nuccore/KC667074
1 3 0.012

For the spike/NS3a gene fragment we see a far less extreme ratio of non-synonymous (amino acid changing) to synonymous changes of about 1:3. The overall nucleotide rate is about 2 times higher in this fragment too once corrected for the sequence length.

What does this tell us about the origin of MERS-CoV?

To answer this it is useful to convert genetic divergence into a time scale using what we know about the rate of MERS-CoV evolution to estimate when the most recent common ancestor (MRCA) existed. The TMRCA (time of MRCA) is the time when the ancestors of the MERS-CoV and the Neoromicia zuluensis bat CoV were in the same host individual. If we divide the genetic divergence of a pair of virus sequences by the rate of evolution, we will estimate the total amount of evolutionary time between the two viruses. The TMRCA is then approximately half this.

On this page we estimate a rate of evolution for MERS-CoV at about 1.5x10-3 subst/site/year. This is based on the amount of genetic change between the few available MERS-CoV genomes. It is approximate and imprecise and is likely to change as more MERS-CoV viruses are sequenced. However, this is is likely to be at the upper range of plausible rates so for these purposes will be conservative (the fastest possible rate will give the most recent possible estimate of the TMRCA).

For the RdRp gene region, the genetic distance between BtCoV/PML/2011/Neo_zul/RSA/2011 and MERS-CoV/KSA/Bisha/EMC_2012 is 0.067 substitutions per site. Dividing this by the rate 0.0015 substitutions per site per year gives a divergence time of 44.7 years giving a TMRCA of about 22 years ago.

For the spike/NS3a gene the genetic distance between BtCoV/PML/2011/Neo_zul/RSA/2011 and MERS-CoV/KSA/Bisha/EMC_2012 is 0.14 substitutions per site. Dividing this by the rate 0.0015 substitutions per site per year gives a divergence time of 93.3 years giving a TMRCA of about 47 years ago. 

Why the descrepancy? The rate of evolution we are using is an average over the entire genome whereas the RdRp gene region is the most constrained and the slowest evolving. If we allow for a two-fold slower nucleotide rate for this gene region then the TMRCA estimate for RdRp becomes 44 years ago which is much more consistent with the spike/NS3a estimate.

These rough estimates suggest that the common ancestor for MERS-CoV and the South African Neoromicia zuluensis bat CoV existed at least 44 years ago, possibly longer. So, whilst an insteresting and useful finding, unfortunately it is limited in how much it can tell us about the nature of any animal reservoir for MERS-CoV in the Arabian Peninsula. It is clear that MERS-CoV ultimately emerged from bats (as all the other coronaviruses in this group are found in bats and now we have a reasonably close relative from a bat) but we can't yet determine the species or geographical range of the direct source of MERS.

Secondly, this is a single virus from a bat in South Africa so the phylogenetic tree can tell us nothing about the direction of movement – it is possible the common ancestor was in the Middle East or South Africa or somewhere else all together (the next most related bat coronaviruses are from Europe).

Caveats:

There are many assumptions being used in this simple analysis but most of them are likely to have the effect of making the TMRCA an under-estimate (i.e., too recent). For example, the rate of evolution assumed is at the upper bound of plausible rates and other estimates for other coronaviruses such as SARS have been lower. The genetic distance estimate is a simple measure of observed number of nucleotide differences divided by the length of sequence but this ignores the fact that multiple substitutions can occur at the same site masking the true number of differences (the so-called multiple-hit problem). This underestimate in the true genetic distance can be corrected for but at the divergence levels we are seeing here the effect will be small and would have the effect of pushing the TMRCA back a bit. We are not correcting for heterogeneity in rates amongst sites or differences between codon positions or nucleotide substitutions. Again, all of these will result in an underestimate of the genetic divergence.

Antibodies to MERS-CoV like coronavirus in camels from Oman

A paper by Reusken et al in Lancet Infectious Diseases adds another clue to the animal reservoir for MERS-CoV. In this study the authors find that 50 out of 50 serum samples from Omani dromedary camels showed the presence of antibodies that strongly cross-react to the spike protein of human-isolated MERS-CoV (and don't cross-react with either SARS-CoV or human coronavirus OC43). Perhaps more confusingly 15 of 105 samples from Spanish camels from the Canary Islands also tested positive. 

This finding does not necessarily mean that the virus that elicited the antibodies in the camels was the precisely MERS-CoV as these serology tests are not that specific (for example it may be that antibodies raised by the bat coronavirus above would also cross-react with MERS-CoV spike. However the presence of a related virus in camels from the same geographical region (and widely distributed over Oman) is very suggestive of a link with MERS. Precisely what the link is is not yet clear. The camels in Oman could have been infected from the same source as the humans. The camel virus could be an independent jump from bats by a related virus. To distinguish these possibilities, sequences from infected camels would be needed and given the likely short period of infection for any individual, many samples will need to be tested.

Ultimately, camels and other livestock make more sense than bats as the direct source of multiple zoonotic infections because of the much higher contact between humans and camels. Camels are moved around the world with Australia and Africa being major exporters of camels to the Middle East. However, if the Middle East is generally a net importer for camels then this may explain the geographically constrained nature of the MERS-CoV cases thus far. 

Middle East respiratory syndrome coronavirus neutralising serum antibodies in dromedary camels: a comparative serological study

Chantal BEM Reusken PhD,Bart L Haagmans PhD,Marcel A Müller PhD,Prof Carlos Gutierrez PhD,Gert-Jan Godeke BSc,Benjamin Meyer MSc,Doreen Muth PhD,V Stalin Raj PhD,Laura Smits-De Vries MSc,Victor M Corman MD,Jan-Felix Drexler MD,Saskia L Smits PhD,Yasmin E El Tahir PhD,Rita De Sousa PhD,Janko van Beek MSc,Prof Norbert Nowotny PhD,Kees van Maanen PhD,Ezequiel Hidalgo-Hermoso DVM,Berend-Jan Bosch PhD,Prof Peter Rottier PhD,Prof Albert Osterhaus PhD,Christian Gortázar-Schmidt PhD,Prof Christian Drosten MD,Prof Marion PG Koopmans PhD

The Lancet Infectious Diseases - 9 August 2013 

doi:10.1016/S1473-3099(13)70164-6

 

Share: