I (Kevin Cook) had questions for David Van Vactor concerning UAS-microRNA “sponge” constructs described in FBrf0228710 and several other constructs that were sent to Bloomington, but didn't make it into the paper. These constructs express an RNA transcript consisting of multiple catenated copies of a sequence complementary to a sequence in the stem region of the stem-loop structure of a microRNA. Expression of the sponge construct leads to elimination of microRNA transcripts.
In the following descriptions, “5P” refers to the stem sequence closest to the 5’ end of the microRNA and “3P” refers to the complementary stem region closest to the 3’ end. Upon microRNA processing, the “dominant” sequence is the stem sequence that is biologically active in gene regulation. The “less dominant” or “*” sequence is the stem sequence complementary to the dominant sequence that is not active or less active in gene regulation.
Me: In the set of stocks you sent, there are several constructs that appear to correspond to multiple genes. For example, you have a P{UAS-mCherry.mir-2a.sponge.V2} construct when there are two "mir-2a" genes: mir-2a-1 and mir-2a-2. Is the repeated sequence in the construct present in both mir-2a-1 and mir-2a-2 so that both genes are knocked down?
Van Vactor: miR-2a-1 and miR-2a-2 are defined in miRBase (www.mirbase.org/) by different 5P strands, which in this case is the less dominant (*) strand. The dominant strand (miR-2a3P) is identical for both family members. The sponge construct was designed to target the dominant mature miRNA strand. See the sequence alignments in the associated file. 
Me: The other cases I found are:
P{UAS-mCherry.mir-2b.sponge.V2} for the mir-2b-1 & mir-2b-2 genes
P{UAS-mCherry.mir-6.sponge.V2} for the mir-6-1, mir-6-2 & mir-6-3 genes
P{UAS-mCherry.mir-13b.sponge.V2} for the mir-13b-1 & mir-13b-2 genes
P{UAS-mCherry.mir-281.sponge.V2} for the mir-281-1 & mir-281-2 genes
P{UAS-mCherry.mir-983.sponge.V2} for the mir-983-1 & mir-983-2 genes
Van Vactor: These transgenes work the same way as the P{UAS-mCherry.mir-2a.sponge.V2} construct works for the mir-2a-1 and mir-2a-2 genes.
The P{UAS-mCherry.mir-276a.sponge.V2} and P{UAS-mCherry.mir-276b.sponge.V2} constructs are special cases. They target miR-276a-3P and miR-276b-3P respectively (see sequence alignments in associated file). However, in this case the miR-276b sponge sequence is targeting the less dominant (*) strand – apparently, for miR-276b the dominant strand is 5P. In addition, because the only difference between miR-276a-3P and miR-276b-3P is at position 10, both sponges target both miRNAs since position 10 is in the sponge bulge (not complementary to the miRNA).  See the sequence alignment in the associated file.
Me: Is the UAS-scramble.sponge construct in your manuscript the same as P{UAS-eGFP.Scramble-SP} (http://flybase.org/reports/FBtp0070473.html) described in Loya et al. (http://flybase.org/reports/FBrf0209364.html), or did you build a new mCherry version to match the set you sent?
Van Vactor: This is a new mCherry-marked, second generation scrambled sponge construct, which matches the UAS-mCherry.microRNA.sponge collection. It consists of 20 catenated copies of a scrambled microRNA sequence.