Lppnx lncRNA: The new kid on the block or an old friend in X-inactivation choice?

Dear Editor

I would like to raise a few points for reflection and discussion based on the recently published manuscript by Hierholzer et al. (1).

Is the Lppnx locus described in this study different from the Linx locus described previously? (2, 3) The Lppnx promoter deletion (0.6 kb) is contained within the deletion of the Linx promoter (2 kb), with very similar effects on the expression of its associated transcript and on Xist and choice of X to be inactivated (3). Similarly, characterization of the Lppnx transcript shows that it is all equivalent to the Linx transcript described previously (2): Expression occurs in both sexes and is associated with pluripotency (restricted to ICM in vivo and down regulated upon differentiation of ES cells ex vivo); transcripts are more abundant in the nucleus than cytosol and have no protein-coding potential.

Is the Lppnx/Linx RNA important for Xist regulation? The authors favor such hypothesis based on a promoter deletion; however, deleting the promoter of a lncRNA locus can also eliminate important genomic cis-regulatory elements (4). We have previously complemented experiments of Linx promoter deletion with a Linx promoter inversion, which showed that the absence of Linx transcription and transcript did not lead to skewed Xist expression ratios and choice patterns, contrary to the promoter deletion (3). Thus, before further investigations are pursued, it is preliminary (and maybe misleading) to state that the lncRNA underlies the effects reported.

Moreover, the mechanisms proposed by the authors are not incompatible with a genomic cis-regulatory element, namely the chromatin contacts with the Xist-intron1 region and the loading of pluripotency factors at this region and others. It remains unclear based on the data presented whether such chromatin contacts are significant (statistically and/or biologically), and it will be interesting to investigate whether the effects on Xist-intron1 are in cis as expected.

Importantly, the fact that deletion of the Xist-intron1 region in Lppnx-deficient ES cells rescues the expected Xist ratios does not indicate that Lppnx/Linx acts via Xist-intron1. Several elements are known to affect Xist ratios and can do so independently of each other (3, 5, 6); if a positive “skewer” is deleted on the same chromosome in which a negative skewer was previously deleted (or vice versa), their effects are expected to rescue each other’s, and this does not mean that they act via each other.

Finally, is the Lppnx/Linx locus the elusive Xce locus (7, 8)? The authors narrowed down the Xce locus to an 80-kb region (without reporting how) and, through genetic dissections, showed that only Lppnx/Linx and not the other loci present within the large 80-kb region (Cdx4 and Chic1) has an impact on Xist expression. Interestingly, when deleting the Lppnx/Linx promoter, the authors observed different effects on Xist expression depending on which Xce allele harbored the deletion. This could suggest that the Xce effects are determined by more than the Lppnx/Linx locus itself.