Correspondence

MicroRNA in Acute Myeloid Leukemia

N Engl J Med 2008; 359:653-654August 7, 2008

Article

To the Editor:

Marcucci et al. (May 1 issue)1 highlight a possible association between the microRNA-181 family and innate-immunity genes. Of the 452 genes that correlated strongly with their microRNA summary value, 32 were predicted to be targets of the microRNA-181 family, according to the TargetScan database.2 Although some of these 32 genes indeed contribute to innate immunity, we write to underline the risk of using only one algorithm to predict microRNA targets and recommend the use of multiple algorithms to consolidate predictions. Thus, reanalysis of the published data with an alternative algorithm, miRanda,3 predicted 50 target genes for the microRNA-181 family, only 10 of which overlapped with the TargetScan predictions. Using DAVID,4 we found that these 50 genes were not statistically enriched for the gene-ontology category of “immunity” but rather for “protein binding” and “regulation of apoptosis.” Had the authors used miRanda instead of TargetScan, their conclusions regarding microRNA-181 target genes would have been different. In highlighting this discrepancy between algorithms, we wish to emphasize the need for greater circumspection in their application.

William J. Ritchie, Ph.D.
Stephane Flamant, Ph.D.
Centenary Institute, Sydney, NSW 2050, Australia

John E.J. Rasko, M.B., B.S., Ph.D.
Royal Prince Alfred Hospital, Camperdown, VIC 2050, Australia
j.rasko@centenary.usyd.edu.au

4 References

1. 1

   Marcucci G, Radmacher MD, Maharry K, et al. MicroRNA expression in cytogenetically normal acute myeloid leukemia. N Engl J Med 2008;358:1919-1928
   Full Text | Web of Science | Medline

2. 2

   Grimson A, Farh KK, Johnston WK, Garrett-Engele P, Lim LP, Bartel DP. MicroRNA targeting specificity in mammals: determinants beyond seed pairing. Mol Cell 2007;27:91-105
   CrossRef | Web of Science | Medline

3. 3

   John B, Enright AJ, Aravin A, Tuschl T, Sander C, Marks DS. Human MicroRNA targets. PLoS Biol 2004;2:e363-e363[Erratum, PloS Biol 2005;3(7):e264.]
   CrossRef | Web of Science | Medline

4. 4

   Dennis G Jr, Sherman BT, Hosack DA, et al. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol 2003;4:P3-P3
   CrossRef | Medline

Author/Editor Response

In our article, we showed that the summary value of the microRNA expression signature correlated with expression of 452 genes. Sixteen gene-ontology terms for biologic processes had at least half their members included among these 452 genes. Of the 16 terms, 15 included members participating in innate immunity. This finding did not depend on the prediction of microRNA targets.

We used TargetScan to predict which of the 452 genes were putative targets of microRNA-181, the most represented microRNA family in the signature. We noted that seven target genes (all negatively correlated with microRNA-181 expression) were related to innate immunity. The miRanda algorithm also predicts that four of these genes will be microRNA-181 targets, suggesting some similarity in the results obtained from TargetScan and miRanda. Although we agree that algorithms for predicting microRNA targets should be applied carefully, we believe that the optimal approach for definitively linking microRNAs to the regulation of specific genes is experimental validation at the bench.

Guido Marcucci, M.D.
Michael D. Radmacher, Ph.D.
Clara D. Bloomfield, M.D.
Ohio State University, Columbus, OH 43210
guido.marcucci@osumc.edu