You need to use different internal control to choose right one  and I would suggest you to do normal PCR and run into agarose gel to confirm your result if you get same result maybe you have feedback loop with you gene and miRNA.

The miRNA may be affecting a transcription factor responsible for regulating all of your genes of interest.  You can try to address this by searching for transcription factor binding motifs in the regulatory regions of these genes.  There are a bunch of tools that let you do this (search Google/Pubmed for "transcription factor prediction").  

There are also a few cases where miRNAs have been showed to have directly upregulate mRNA expression, contrary to the canonical miRNA method of action.  Obviously, this is less likely but still possible.

http://www.ncbi.nlm.nih.gov/pubmed/22053081

http://www.ncbi.nlm.nih.gov/pubmed/20410487

Is it possible to check protein levels of the miRNA targets? It could be that the miRNA is blocking translation of the target mRNA, rather than degrading the mRNA.

Two important factors may affect your findings that were obvious in my studies as well.

First make sure that your target site i.e. miRNA binding site -usually ~7 bases in the target gene 3'-UTR or coding region (unusually) does exist or intact. You can do simple PCR and sequencing or functional validation for miR-target binding by luciferase reporter assay. It's important before you conclude any findings because different cells or tissue types do not express or have mutated miRNA target site or may have variable 3'UTR lengths or alternative polyadenylation mechanisms so that cells can avoid the regulation by specific miRNAs) in order to fine tune gene regulation. see Science 2008, 320(5883): 1643-7 AND RNA Biol. 2012 May;9(5):563-76.

Another important factor is the use of correct endogenous control for normalization of miRNA expression level by qPCR. The endogenous control RNAs, including small nucleolar and small nuclear (snoRNAs/snRNAs) that are highly conserved and widely expressed in different tissues, are commonly used in the normalization of miRNA expression. Recent evidence has shown that these small control RNAs can be differentially expressed and frequently play roles in cancer cells or in pathogenensis (Br J Cancer 2011, 104(7): 1168-77). Therefore using a single snoRNA or snRNA as a reference gene may introduce variability in results and analysis. To overcome this problem, a panel of six snoRNAs/snRNAs  are often used as reference genes or endogenous controls giving an average measure of relatively stable expression levels across tissues and cell types (BMC Mol Biol. 2008 Aug 21;9:76).

I hope these information will help.

Start first with a luciferase reporter gene to verify that seed regions are functional

Hi

Thanks a lot for your great comments. (Dear Justin, unfortunately, it's not possible to check protein levels of the miRNA targets).

As you suggested,I think, I should check pathways for finding feedback loop and also find an affecting  transcription factor that regulates my genes.

Best, Maryam