The two most common histological subtypes of malignant germ cell tumour are distinguished by global microRNA profiles, associated with differential transcription factor expression

Table 3 Transcription factors up-regulated in YSTs versus germinomas

Transcription Factor		Paediatric Dataset		Adult Dataset		Motif Scanning		Gene Function
Accession	Name	Rank (n = 594)	Log₂ fold change	Rank (n = 730)	Log₂ fold change	Maximum motif bit score	e -value
NM_005257	*GATA6*	20	+4.72	13	+4.85	-0.28	0.032	Marker of early endodermal differentiation; transcriptional regulator of differentiation and proliferation
NM_031283	*TCF7L1*	29	+2.52	36	+2.17	-	-	Wnt pathway signalling
NM_003069	*SMARCA1*	52	+2.49	52	+3.27	-	-	Regulates transcription by altering chromatin structure. Involved in development and differentiation
NM_001146283	*TCF7L2*	59	+2.77	64	+3.09	-0.58	0.184	Wnt pathway signalling
NM_001002295	*GATA3*	63	+3.63	146	+3.66	-1.73	0.173	Marker of early endodermal differentiation; transcriptional regulator of differentiation and proliferation
NM_003108	*SOX11*	229	+2.62	313	+1.61	-	-	SRY (sex determining region Y)-box 11; embryonic development and cell fate; involved in tumorigenesis
NM_000280	*PAX6*	316	+2.77	600	+1.13	-	-	Paired box protein 6; marker of neuro-ectodermal differentiation
NM_005524	*HES1*	355	+1.41	175	+2.01	-	-	Hairy and enhancer of split 1; regulates growth and proliferation
NM_000325	*PITX2*	356	+1.41	274	+1.69	-	-	Paired like homeodomain 2; regulates terminal differentiation
NM_001031804	*MAF*	365	+1.39	234	+1.83	-3.93	0.032	Musculo-aponeurotic fibrosarcoma oncogene; involved in development and terminal differentiation

The 10 transcription factors significantly up-regulated in both paediatric and adult YSTs versus germinomas, including their log₂ fold changes and ranks in the lists of differentially expressed mRNAs, are listed. The iMotifs values represent the maximum motif bit scores for each of the four individual transcription factors with predicted binding sites in the 10 kb miR-302 promoter region, together with derived empirical e-values. In iMotifs the optimal bit score is normalised to zero, with all other bit scores having negative values. Accordingly, the closer a score to zero, the greater is the significance of the predicted transcription factor binding site.

ISSN: 1476-4598