You should increase the concentration of your template of apply preamplification step before Real-Time reaction in order to lower Ct value of the reference gene. Otherwise, you cannot use it as reference.
Also, you can change your reference gene. Since the appropriate reference gene may change according to your sample or tissue type.
I absolutely agree with Sercan - try to find another relevant gene that can be used as a reference. I would not do preamplification for reference gene if it's not single cell PCR, increasing the template sounds better.
In short, you can't. I agree with the others in that you should change your reference gene and increasing the concentration of your template. Just out of interest, how many times have you tried this and did you use primers of your own design?
It is important how much RNA you use. I typically use 50 ng per real-time RT-PCR using. It is actually easy to get that amount of RNA. If the number of Ct shown on your reference or your target is more than 30sh, I would not feel confident about the results. The number of Ct for the reference should be around 20sh, and most of the housekeeping genes behave like this. If the number of Ct is really that low (such as 37 cycle), I would bet the PCR primers might not be great and you need to re-design the primers.
Especially if you are unsure of an appropriate reference gene, it may be wise to run several reference genes. That way, if one fails you hopefully have at least one good reference, or you can use the mean of several as your baseline.
your relative gene expression is better to be a gene which is constitutive expressed like a GADPDH or Actin.
usually these genes are very highly expressed.
You need to titrate your template cDNA to test for efficiency of replication and to know the concentration of template need for your future experiments.
Thank you all for your inputs. I would like to add few points here. I have undifferentiated stem cells and I am differentiating them towards different lineage. Now our genes are lineage specific genes those are not getting expressed by undifferentiated cells at all. So as per published data they compare relative expression/ fold change with respect to undifferentiated cells. So how to go about it. Our primers are working absolutely fine. As differentiated cells gives us good ct value. we also run housekeeping gene. But how to interpret our data?
You read this paper. Here you will find how to calculate fold changes using ddct method.
http://www.gene-quantification.de/livak-2001.pdf
Second, the ct value of your housekeeping gene should not be more than 20. you can not use ct value of 37 to determine the any change or comparison study. Either you change the primer sets or increase the template concentration.
Go through Bustin et al., 2009, MIQE guidelines. this will be very useful for you to carry out the experiment correctly.