None of the dyes mentioned are RNA-specific; as far as I know, there aren't
any dyes which are RNA-specific in the sense in which 7-AAD, the chromomycin
dyes (chromomycin A3, mithramycin, olivomycin) DAPI, and the Hoechst dyes
are DNA-specific.
Tricyclic dyes such as acridine orange and pyronin Y do not show enhanced
fluorescence on binding to RNA. Acridine orange forms a luminescent complex
with single-stranded RNA; pyronin Y does not, and is fluorescent when bound
to double stranded DNA or RNA. RNA-"specific" staining with pyronin Y is
achieved by blocking DNA binding with a Hoechst dye, 7-AAD, or methyl green;
in such cases, RNAse treatment abolishes most of the pyronin fluorescence.
Pyronin Y can also be used for staining reticulocytes, which, like other red
cells, shouldn't contain DNA, and micronucleated erythrocytes or cells with
Howell-Jolly bodies are considerably more fluorescent than retics and can
thus be discriminated.
Cyanine dyes such as thiazole orange and the TO-PRO and TOTO series show
some spectral shift and fluorescence enhancements by factors of hundreds to
thousands when bound to DNA or RNA; these dyes apparently do show some
fluorescence when bound to single-stranded RNA. Their utility as
reticulocyte stains depends on the absence of DNA from retics; see above.
Auramine also fluorescence enhances when bound to DNA and RNA; its
spectrum is also shifted markedly. I always wondered how TOA managed to use
auramine O in a retic counter with 488 nm excitation when the absorption
maximum is at about 435 nm, until I ran fluorescence spectra. The auramine
O nucleic acid complex absorbs somewhere around 470 nm, and 488 nm
excitation works. I also noticed that auramine O is about 15 times as
fluorescent when bound to polydGdC than when bound to polydAdT, but have not
had much luck using this base preference for chromosome staining.
Among red-excited dyes, I have used dicyanine A for retic staining; it
also fluoresces when bound to DNA or to RNA, and emits around 670 when
excited at 633-650 nm. B-D's thiazole blue, and NKK's NK321, are also
usable in this region, but the B-D dye is not commercially available.
TO-PRO-3 and TOTO-3 have the same spectra as thiazole blue and NK321, but
require permeabilization for reliable staining (although entry into intact
cells has recently been reported). However, thiazole blue apparently takes
hours to equilibrate with RNA in retics, while thiazole orange takes only
1/2 hour or so, making the former dye of questionable utility. The retic
staining method I used with dicyanine A was a nonequilibrium method which
factored in the staining kinetics.
-Howard