Ss. Potter et Ww. Branford, EVOLUTIONARY CONSERVATION AND TISSUE-SPECIFIC PROCESSING OF HOXA-11 ANTISENSE TRANSCRIPTS, Mammalian genome, 9(10), 1998, pp. 799-806
We previously described the existence of abundant, processed, polyaden
ylated murine Hoxa 11 antisense transcripts. Of particular interest, i
n the developing limbs the antisense transcripts were observed to be p
resent in a pattern complementary to that of the sense transcripts, su
ggesting a possible regulatory function (Hsieh-Li et al. 1995). We hav
e analyzed the human HOXA 11 genomic locus, showing strong evolutionar
y conservation of regions potentially encoding antisense transcripts.
Human HOXA 11 fetal kidney antisense cDNAs were identified and sequenc
ed, demonstrating the evolutionary conservation of Hoxa 11 antisense t
ranscription. As for the mouse, the human antisense RNAs were polyaden
ylated and showed several alternative processing patterns, but shared
the sequences of a common 3' exon. The evolutionary conservation of th
e opposite strand transcripts strongly suggests function. A significan
tly long open reading frame was observed, but mouse-human comparisons
argued against true coding function. Murine kidney Hoxa 11 antisense t
ranscription and processing was also examined, revealing tissue-specif
ic differences between limb and kidney. A novel procedure, designated
Race in Circles, was devised and used to define mouse limb antisense t
ranscription start sites. Furthermore, comparisons of human, mouse, an
d chicken sense transcript Hoxa 11 homeobox nucleotide sequences and t
heir respective encoded homeodomains indicate a very strong selective
pressure in vertebrates against mutations that result in coding change
s. Given the significant differences in amino acid sequences of the ho
meodomains of different Hox genes, this observation argues for individ
ual homeodomain functional specificity.