first you should try a reference/control gene/primers such as actin. if it is ok, then try different concentrations of cDNA.

good luck

You can also run your mRNA soon after isolation on gel, there should be appearing two bands on the gel, if it's just a smear then there is problem with your mRNA isolation.. Once you confirm that your mRNA is ok, then you should make first strand cDNA and use it further in your experiment..

Some possible reasons:

1. The efficiency of cDNA synthesis is low.

2. The PCR primers are not working properly.

first I would check that your primers are exonic as cdna has no intronic sequence. Then I would amplify a housekeeping gene as Elvan says  just to make sure that you do have cDNA  present in amplifiable amounts. Because your genome is now small you can lower the annealing temperature of your reaction without too many non specific amplification problems. You can run more cycles of pcr or design 2 nested primers inside your primers and re-amplify the product of the first round pcr with 10-20 cycles of nested pcr. Take particular care to add EtBr to the checker gel.

Hi Lulu

to add to the above considered answers:

1. Make sure you use 1ug of total RNA to make your cDNA and then initially validate with housekeeping primers like b actin or GapDH. This will verify that that your cDNA is in good working order and thus any down stream problems with others GOI primers are likely there fore a function of those primers and/or copy number of you gene and not an artefact of poor quality cDNA; I do this in fact each time I make a new batch of cDNA

2. Design 2 sets of primers to your target of interest: Not all primer pairs work and some work but generate secondary products as well even with apparently optimised PCR cycling conditions. When designing primers check for things like hair pin structures within a given forward and reverse primer and also linear intra primer annealing as well as F-R inter primer annealing, i.e. primer dimers

I find the following IDT program useful in that regard:

https://www.idtdna.com/calc/analyser

3. Apart from designing primers to exon exon boundaries as Paul has suggested (to specifically target cDNA and not genomic DNA which will undermine the efficiency of your specific RT PCR reaction) also consider potential secondary structure within the target sequence itself: Is my target excessively GC rich potentially causing secondary structure and thus limiting the availability of linear single stranded template and thus undermining my PCR or consisting of regions with more than 4 x G/C residues together, which will cause polymerases to stop and then fall away from the replicating sequence, there again reducing amplification efficiency

M Fold has been used historically to determine such potential structural complications in target sequences and IDT also has its own version:

https://www.idtdna.com/UNAFold?

http://unafold.rna.albany.edu/?q=mfold

If secondary structure in target sequence is potentially unavoidable try adding 0.5M betaine to your PCR reaction

4. Equally as important if not more important are PCR conditions themselves, even when all of the above has been carefully considered; In particular

A. Design primers as above but make sure their respective Tm's differ by no more than 2C

B. Regarding Tm of primers and actual annealing temp deployed during PCR, for new primers initially try a PCR annealing gradient from Tm-3C to Tm+2C: In truth most modern highly processive polymerases like phusion will amplify specifically and efficiently @ the Tm itself but sometimes when dealing with difficult target sequences are simply because all primers to an extent are a law unto themselves, i.e. there are confounding/kinetic factors that cannot always be known, dropping the annealing temp to Tm-3 to Tm-5 (no lower) can act as a trouble shooting aid

C. Determine Tm of your primer pairs in the context of your actual PCR conditions: Intrinsic factors linked to the PCR reactions mix itself, e.g Mg concentration, primer concentration, dNTP concentration will have local affects on duplex stability and hence actual real time Tm values

Again I find the tool offered by IDT useful in this regard

D. Where at all possible keep your PCR amplicons below 1KB: Shorter amplicons simply amplify more efficiently and are less likely to incur secondary structure regions/problems in your target sequence. What is more if I was experiencing problems and didn't require longer amplicons than I would personally make such amplicons < 500bp 

To bring all of the above together  primer BLAST is a very useful primer design tool:

http://www.ncbi.nlm.nih.gov/tools/primer-blast/

E. Finally, an initial denaturation cycle of 2-5 minutes @ 94-96C can render your target sequence more single stranded for commencement of PCR which will have a big impact on PCR efficiency and hence visualisation of PCR amplicons    

Thank you so much for the answer and suggestion :)

I have used reference gene (actin, tubulin and ubiquitin) for PCR but the reference gene also not appear on electrophoresis gel, the primer that i used for PCR just have a little self dimer, even i have reisolated RNA and remake cDNA for two batch but the bind is still not appear on electrophoresis gel.

How if i use the cDNA for qPCR to make sure that the reaction is work? Is that will be ok?