Are you certain that the peptide, when part of a protein in cells, does not undergo any post-translational modification that would affect its retention time?

I'd have real doubts that you are in fact monitoring the peptide you think you are. Every heavy peptide I've ever used completely overlaps with the native form, as it should. I'm assuming you can do MS/MS of the 38.8 min peptide and match that against a protein database? You need to investigate this and ensure you have the right peptide.

"Gradient" implies that you re using a HPLC separation. But this HPLC system works on polarity of the analytes, when separated on a column. I do not believe that there is much difference between polarity of your labelled and unlabelled peptides.  

Dear Juhura

You do not really give much information and there are many parameters that can affect what you are describing.

Taking the difference in the retention time between the two peptides (endogenous verses synthetic) I would suggest that they are not the same peptide. The retention times are way different.

How did you come to the "best peak" conclusion? Is this based on peptide mass? What criteria are you using to assess the different peaks?

How are you monitoring the eluted peptides, UV or mass spectrometry? The elution profile of peptides can be very complex and the question is how did you do your sample preparation? How did you get rid of the other compounds in your matrix like nucleic acids, lipids, carbohydrates, proteins, salts etc

Did you run the synthetic peptide as a standard in a separate injection or was it spiked into the cell lysate and if so at which point of the sample preparation?

How are you processing and enriching your sample? Is the cell lysis by detergent or mechanical means? Do you add protease inhibitors for the different classes of proteases?

Is the peptide prone to disulphide bridge formation, in which case there could be intra peptide bridges forming, and these could change the retention time of the peptide but this can be prevented by acetamidation.

If using mass spectrometry to monitor the eluent, then the mass should differ by the number of heavy isotope atoms added during the synthesis of the peptide as the precursor mass. If running MS/MS the fragmentation should be the same with the same mass offset seen for the precursor up to the labelled amino acid(s) after which the masses should exactly coincide.

If the peptide from the whole cell digest is eluting early it would imply a more polar peptide, which could point to a truncated peptide due to cleavage by a proteolytic enzyme during the lysis and sample preparation. This is an often overlooked aspect of sample preparation and when using cells or tissue can be a major complication.

If you have enough heavy peptide you could try adding it to your sample before your cell lysis and check if it also shifts in retention time, but this is a rather expensive experiment, taking the cost of isotope labelled peptides into consideration.

Fragmentation of longer peptides can be quite good depending on the amino acid sequence and ionisation state during mass spectrometry. 

More information could help trouble shoot your problem.

Good luck.

As you suspected, 10 min retention time shift is unacceptable in chromatography. When you work with a labeled peptide you are not significantly altering the physicochemical properties of the original peptide. Therefore, the changes should be minimal. From the isotope labeling peptide and last part of the description it appears that you are perhaps using mass spectrometry to determine the peptides (accurate mass?) . Here are a few additional questions:

How many times did you perform this experiment? How is your instrument validated? Did you run any known standards to validate the instrument performance? (This is crucial)

Did you run your labeled peptide without the matrix (no spiking experiment)?

What is your mass resolution? IMHO MS/MS will be suggestive and helpful in this case.

Good luck.

I agree with Chakraborty.  Labelled peptide should not be substantially different from unlabelled peptide in physiochemical properties.  ie: HPLC retention times.