The repeated genes are part of a duplicated segment of about 200 kB with greater than 99% identity, separated by a large gap. There are only two BAC end pairs spanning the gap. In contrast all other mammals that we examined, including other primates, have only one copy of the TMEM236 and MRC1 genes between the same flanking marker orthologues, without the gap. While a very recent duplication in humans cannot be ruled out, it seems much more likely that this is a mis-assembled region in the human genome and thus that all mammals carry only a single MRC1 gene. In species from other classes of terrestrial vertebrate, examination of the region of the genomes between the most highly similar homologues of the flanking markers revealed that some of these contained multiple, tandemly arranged diverged paralogues of MRC1. Xenopus tropicalis genomes contained only a single gene, the lizard Anolis carolinensis had three, while three birds and the painted turtle had five. This indicated duplication of the ancestral MRC1 gene in the avian lineage and its precursors. The most likely sequence of events would have been an initial duplication producing the ancestors of chicken MRC1L-A and MRC1L-B genes, followed by a much more recent duplication of the latter in the lizard, and by further early duplications in the common ancestor of birds and turtles. In this context, it is of note that the phylogenetic position of the turtle has been the subject of much debate over a number of decades. Whilst a recent report based on an analysis of microRNAs suggested that turtles form a clade with lizards, subsequent reports place them in the archosaur lineage with birds and the crocodylia. The more recent proposal is compatible with the simplest possible history of the MRC1 genes described in the present report. Chicken orthologues of the adjacent DEC205 and PLA2R genes, and of the MRC2 gene, are found elsewhere in the genome. The additional genes in the MRC1 locus are therefore not relocated orthologues of these genes. All the identified genes in all the species examined had intact reading GDC-0879 frames coding for proteins with the CTLD structure normally found in members of the mannose receptor family. All were found as spliced mRNAs in the chicken. Thus it is unlikely that any of the duplicated genes is a pseudogene, although differently spliced variants of the genes D and E transcripts were found in HD11 cDNA that had interrupted reading frames. The physical distances between the genes C, D and E were small, and the pattern of variation of their transcript levels in tissues was very similar. It may be that the transcription of these three genes is coordinately regulated by a shared set of upstream cis-acting elements. Indeed, the PCR amplifications used to confirm splice junctions would not have detected splicing between exons in different genes, so that the existence of splice variants that combine segments of the three genes, in a manner similar to the TWEPRIL transcripts from the TWEAK-APRIL genes in mouse, is not excluded. The HD11 cell line contained mRNA for all five MRC1L genes, but peptides from protein immunoadsorbed by KUL01 included only those from MRC1L-B. This would be consistent with the KUL01 epitope being exclusive to MRC1L-B. However, the similarities between the MRC1L paralogues, while low, are sufficient that we could not exclude the possibility of recognition of the product of one or more of the other genes in the context where KUL01 is applied as a macrophage marker. To test this possibility we conducted two further experiments. As shown in figure S6, treatment of HD11 cells with transfection reagents including a small interfering RNA with 25/25 nucleotide.