Turksen, K. and Troy, T.-C. (2004). Barriers built on claudins. J. Cell Sci.117, 2435-2447.
We wish to correct some errors that we inadvertently introduced in revising and editing our recent Commentary. In the text, we incorrectly cited (Wolburg et al., 2003) in reference to the downregulation of claudin-23 in gastric cancer; the reference should be (Katoh and Katoh, 2003) as correctly cited in Table 3. (Peacock et al., 1997) was mistakenly inserted in relation to the discussion of the number of claudin genes. We also wrote the number of human claudin genes as 24, the number currently ascribed to the mouse claudin genes (GenBank). The correct number ascribed to human claudin genes by (Katoh and Katoh, 2003) is 23, up from the originally predicted 20 (Venter et al., 2001). In this regard, it is worth noting that, since our Commentary was published, (Loh et al., 2004) have annotated the claudins in the teleost Fugu rubripes genome and reported 56 claudin genes, of which only 35 can be assigned orthology to 17 mammalian claudin genes, with the remaining 21 being specific to the fish lineage and most of the 56 expressed in a more or less tissue-specific fashion, or at particular developmental stages. This, along with other issues we raised, suggests that additional annotation of multiple other genomes and other functional genomics approaches will be useful to advance our understanding of claudin biology and physiology. Finally, based on a Clustal analysis of full-length claudins, we reported that there is a highly conserved WWCC motif of unknown function within the first loop of the claudins analysed; a motif also reported in alignments of claudins carried out by (Katoh and Katoh, 2003).
