Generally, all RNAs are sensitive to exonucleases (with individual requirements and specificities) acting on the 5' and 3'-end. However, specific RNA modifications, secondary structures and bound proteins impede them mostly.
In eukaryotic cells, mRNA are protected by their poly-A tails because it is covered by poly-A binding proteins (cytoplasmic PABPC1-4 and nuclear PABPN in humans) and assumes an inaccessible conformation. There are however many specialized deadenylases (two known families of nucleases: DEDD, e.g. PAN2, CAF1, and EEP, e.g. CCR4) that trim and cut off the poly-A tails of mRNA and thus start degradation of the mRNA by 3'-endonucleases (e.g. the 3'-exosome in humans).
Please note that in prokaryotes, a short poly-A tail is destabilizing RNA and is added by PAPI during RNA degradation. This process has many similarities to degradation pathways in eukaryotes that involve addition of a short U-tail by TUTases.
Therefore, there is no universal protection of RNA by poly-A tails and to answer your second question: it depends on your experimental system whether a poly-A tail would be protective on single stranded RNA. Without more information, this question cannot be answered in a helpful manner, I am afraid. Consider whether you are going to use a cellular system in eukaryotes or prokaryotes or a cell-free in-vitro system.
For most assays, nucleotide modifications in synthetic RNAs are easier to control and thus could more reliably stabilize your RNA. Please consider stabilizing both 5'- and 3'-ends. Common nucleotide modifications to achieve this are available from many suppliers, with few examples here:
Correction: "thus start degradation of the mRNA by 3'-endonucleases (e.g. the 3'-exosome in humans)" should be "thus start degradation of the mRNA by 3'-exonucleases (e.g. the 3'-exosome in humans)" of course. =)