Hello Nasra,

Each tissue has specificities according to gene expression. The problem will be if the Ct variantions occur among the different samples of the same tissue.

Use triplicates (or more) of your housekeeping gene for each sample, preferencially in the same qPCR run of your target gene samples, and analyze different tissues separately.

Another suggestion: test de different samples your housekeeping gene, because for certain tissues, other housekeeping genes might be more appropriated.

Ideally, you should determine appropriate reference genes de novo for each new comparative scenario: genes X Y and Z might be appropriate for healthy liver tissue, but not appropriate for healthy AND diseased liver tissue, for instance.

This is especially critical for comparing multiple tissues, where many commonly used reference genes might well be inappropriate even if they appear stable in the tissues individually: a gene could be an entirely stable reference in fat, and also an entirely stable reference in muscle, but if the expression in fat is (stably) ten-fold higher than in muscle, it makes for a poor reference gene when comparing the two tissues.

You should take a panel of possible reference genes (ideally 8-12 candidates) and a representative panel of samples (say 3-5 cDNAs from each tissue) and determine the expression of each candidate gene in each sample, then run these results through one or more algorithms (such as geNorm or Normfinder) to determine the most appropriate reference genes for comparing multiple tissues.

Bonus points: if you do this comprehensively and your analysis is solid, this is a publishable finding.

• As mentioned by Amanda, it is primarily a copy number issue
• That is tissue specific expression levels
• It is also salient to note that different preps (i.e. biological replicates) derived from the same tissue (but different starting RNA preps) can also vary by a cycle because of differences in the efficincy of your reverse transcription linked to your starting RNA prep
• This is why when you use triplicates as stated by Amanda they should be techncial replicates, i.e. 3 samples derived from the same cDNA prep and not biological replicates: that is different RNA extractions derived from the same tissue/cell sample