Reply to Hart et al.: MINCR and MYC: More than expression correlation

Based on novel correlation analyses in RNA-seq data from normal tissues and from Burkitt lymphoma (BL) cell lines, Hart et al. (1) suggest that MINCR is not a MYC-induced long noncoding (lnc) RNA. We thank the authors for providing additional analyses, but we do not think these data prove that MINCR is not a MYC-induced lncRNA.

Expression correlation alone is not a proof of coregulation. To claim that MINCR is a MYC-induced lncRNA, we based our argument on three observations: (i) MINCR is up-regulated by MYC in cell lines carrying MYC-inducible constructs (figure 2B in ref. 2); (ii) MINCR promoter is bound by MYC in published MYC ChIP-seq data (figure 1C in ref. 2); (iii) MINCR and MYC expression correlate in RNA-seq data from a cohort of MYC-positive B-cell lymphomas (figure 3C in ref. 2). We also show that MYC knock-down in two BL cell lines reduces MINCR expression (figure S1 in ref. 2).

If regulation by MYC of the identified lncRNA would have been restricted to B-cell lymphoma, we would have been more cautious in choosing the name MINCR (MYC-induced long noncoding RNA). But we do observe MINCR regulation by MYC also in cells of very different origin in respect to B-cells, like retinal pigment epithelial cells, and the analysis of ENCODE ChIP-seq data show clearly that MYC binds to MINCR promoter in all cell lines analyzed. Furthermore, we observe a correlation between MYC and MINCR expression in pancreatic cancer as well. At the same time, this correlation is not observed in ovarian carcinomas, chronic lymphocytic leukemia, and pancreatic endocrine neoplasm (figure 3C in ref. 2). Therefore, we are not surprised to see an absence of correlation in normal tissues (1). The lack of correlation in BL cell lines may be because of the broad heterogeneity that usually characterizes cancer cell lines, in terms of patient age, ethnicity, and chemo-treatment. In contrast, our BL cohort is very homogeneous in terms of patient age, gene-expression profiling, and diagnosis (table S1 in ref. 2).

Concerning the second criticism, we did not intend to offend any researcher working in the same field and we deeply apologize if this was the case. In our publication (2) we only wanted to stress the novelty of the combined analysis of MYC-inducible cell lines and RNA-sequencing data coming from lymphomas. And we did indeed cite the other studies (including that from the Vogt group) (3–6) that analyzed MYC-regulated lncRNA expression before. We also apologize if someone felt we were trying making readers think the P493-6 RNA-seq data were generated by ourselves. For all analyses the information on data source is written in the Materials and Methods of ref. 2, which unfortunately had to be shortened and moved to the Supporting Information for lack of space. The use of the Gene Expression Omnibus identifiers was simply thought to be a more direct way to address the readers to the data source.