No, not at all. matrix of human serum & non-human serum samples are different.

Thanks Dr. Yadav,

The matrix is not the matter. For, example, esteroidal hormones are analyzed by human kits, commonly. Any way, I would like to know if these kits are not suitable for non-human species, how they give us values?!

The only sure way, is if you have purified non-human antigen you want to test to do a calibration (0concentration) curve.

But you can always test whether the assay itself will work in your matrix by doing a calibration curve with different amounts of the human antigen added to whatever the final material is you want to use. If you get the same slope as with a calibration curve for the human antigen in optimal buffer, it shows that the assay system itself will work with your sample type. Whether it will detect your non-human antigen, though?

Dear Buelow,

Thanks for your feedback. Yes, it detects. For example, my colleague measured fish plasma GH, Prolactin and Insulin using human diagnostic kits. The values were GH= 4-8 ng/ml; prolactin = 0.4-0.8 ng/ml and insulin was = 3-10 micro IU/ml. This is exactly what confused me! If the kits do not work for non-human samples, what are these values? These values come from ELISA method, probably. In ELISA methods, the hormones should react with antigen that kits work.

You can't generate a real standard curve because you lack the purified protein from your specific organism. You can report the numbers from your non-human samples as "human GH / prolactin / insulin equivalents"  but its not a direct measurement anymore, you could be detecting something else entirely. Its possible to show  linear competition with the human standards to infer you are measuring what you say but it may not satisfy some other scientists (or reviewers) good luck!

If kit is not based on human antigenic or epitopic specific then it may be used, still be sure about false positive results.

Protein molecules- including protein hormones- from different species have differences in molecular structure; including type and number of amino acids, and also 3-dimensional structure. These differences result in different antigenic properties.

As all immunoassay methods (including ELISA) produced to assay human hormones use antibodies directed to the human-specific molecular structure, therefore if we use a human kit to assay animal specimens, the antibodies in the kit won’t react with animal hormone the same way as human specimens. The difference is sometimes so profound that human specimen doesn’t give any result with animal specimen (for example human insulin ELISA kit would give zero results with rat specimens).

Though using human kits for animal specimens may give some numbers (as your colleague has got some results), but the numbers are not necessarily correct. To get correct results, you have to use the kits that are specially produced for the interested animal specimen.

Any way, if you are obliged to use human kits, you have to perform linearity and repeatability experiences using animal specimens; these are the least validation studies. If the linearity and repeatability is acceptable, then you can report the results the way Edward Michelini has pointed; i.e. as “human equivalent” results, not as ‘correct’ results.

If there is any help I can do please call me at (011)42237081-2 (Sina laboratory, Qaemshahr).

HI there,

All  of the people above have mentioned right things. First of all it is very unlikely that a human specific kit will react with any other species, specially fish. However, in some rare circumstances that may be possible (for few related species) but you have to optimize several things before you can be sure that the cross re-activity is genuine and you are not just getting non-specific binding. In your case, the best way is to get the non-animal species purified antigen (in your case from fish, as Sven mentioned) and generate a standard curve (or ODs) against the human kit standards. In addition, you have to run some non-related proteins from fish to eliminate the possibility that the coated protein does not cross react with any other protein from fish (non-specific binding) but your protein of interest. Even after that you can only report the results as "human equivalent" as suggested above.

Good luck.

Hi,

It depends of biochemical characteristic of the assay. For example, immunologically based tests using antibodies may not work properly if the epitope is not the same in human and animal samples creating false positive (in rare occasions) and mainly negatives. Immunological test based on cell function may not be enough specific and the “chain reaction” activation typical for this response is not usually correlated (linear preferred) so there will not be present a typical dose-response curve except you have previously tested or calibrated.

If the target is hormones you may choose different type of assays: determining its presence in producer glands or their release into blood and you may also detect their presence (or expression/release ratio) through enzymatic activity/ physiological response in target tissue/organ/system where you can compare basal and after stimuli response.

Enzymatic determination kits may work more properly except in cases where there is a Km or reactant specificity variations which is less probable to introduce data errors than in immunological assays.

Problems could be solved if you do a calibration assay with values obtained from known quantities in other related test and those of the same sample with this new use of the test in fish or animal samples in general. Specially valid if there is a linear regression with a good r2 score (I don´t recommend you scores less than 0.9 because is the result of a calibration curve based in a previous / similar method which have already its own deviation, which could be added totally or partially in the new assay calibration curve). It could be certainly more expensive but you should evaluate the cost/benefits balance.