This happens many a times, mostly in case of highly divergent sequences, to which BLAST algorithm cannot find highly similar (default Megablast) sequences in the GenBank. To overcome this, you have to change the stringency of your search, that is you have to optimize the search for less stringent search such as "More dissimilar sequences (discontiguous megablast)" or "Somewhat similar sequences (blastn)", whcih you can select just above the BLAST tab, under the "OPTIMIZE FOR" section.
In addition you also search your sequences through the NCBI CD search tool, which will help you to identify any conserved protein coding domain in your NUCLEOTIDE sequences. You can find this tool at :