Protocol Online logo
Top : New Forum Archives (2009-): : Immunology

unreliable results with LAL kit - (Aug/17/2010 )

I'm having a lot of trouble with the LAL kit to test for endotoxin. The most pressing issue right now is that the EU/mL of the samples are increasing with dilution. I did a 2-fold dilution series from 10X to 1280X. I assumed that the 1280x dilution would be essentially zero but it's not. The Vmax and OD values fall on the low end of the linear part of the std curve, so when multiplied by the dilution the EU/ml are very high. Do I just need to dilute my samples even further or is there something else going on?

-kellywelly-

I have see no one respond for a while to your questions. You use a "LAL kit"...so you purchased from a manufacturer. Companies have tech support departments. Have you tried contacting them as they know all the intricacies of their products?

-sgt4boston-

I have contacted tech support and they are no help. He told me to do it over and when I asked if there was anything else I could do, he said that he thinks doing it over will fix the problem. <_<

-kellywelly-

Although I am not familiar with this kit Iwill try to help.
You should only be running dilutions if the concentration of the samples fall outside of the linear range of the curve and higher than the highest standard in the kit.
Samples should dilute linearly within the range of the dose response curve.
You do not multiply the dilution factor x the signal; to obtain concentration you multiply the dilution factor X the concentration observed of the diluted sample.

If your undiluted samples fall on the curve there is no need to dilute.
If they are > the highest standard; if >>> do serial 10 fold dilutions.
Check the matrix used for dilutions it should be the same as the 0 standard (avoid matrix effects).

To test your dilutions serially dilute BOTH one of your high samples AND the highest kit standard...dilute them in parallel and they should BOTH be linear.

Hope this helps.

-sgt4boston-

Thanks for your response. I'm sorry that I did not explain myself better before.

You should only be running dilutions if the concentration of the samples fall outside of the linear range of the curve and higher than the highest standard in the kit. Yes, this is true of our undiluted samples.
Samples should dilute linearly within the range of the dose response curve. This is my problem. The samples increase with dilution instead of decreasing.
You do not multiply the dilution factor x the signal; to obtain concentration you multiply the dilution factor X the concentration observed of the diluted sample. I did multiply dilution factor by concentration. Sorry that was not clear before.

If your undiluted samples fall on the curve there is no need to dilute. The undiluted samples are completely off the curve from the start of the assay.
If they are > the highest standard; if >>> do serial 10 fold dilutions. I would do this except my samples are not decreasing but increasing. I think there is something else going on but I'm not sure what it is.
Check the matrix used for dilutions it should be the same as the 0 standard (avoid matrix effects). I am not sure what you mean by the matrix. Everything is diluted in endotoxin-free water.

To test your dilutions serially dilute BOTH one of your high samples AND the highest kit standard...dilute them in parallel and they should BOTH be linear. This is what I did. The standard curve is as it should be but my samples are not. There seems to be MORE endotoxin as I dilute the sample not less.

-kellywelly-

OK. I think I have more info now. 1.Matrix Effect: Are the standards a set or are you provided a concentrate that you dilute? If a set comes with the kit then your high samples must be diluted with the "0" (not the assay buffer or diluent)as the matrix of the standards and samples must 'match'. As the sample becomes further and further dilute you begin to change the matrix of the sample to that of the material used for dilution and the difference between the 'sample' and the standard becomes more dramatic. If the kit is 'sensitive' to matrix differences you see issues just like what you are observing. You said the curve is ok...so you must be making dilutions to create the curve? To see that your technique is correct and matrix(ces) are ok making the same dilutions of the highest standard and one of your high samples in parallel will demonstrate linearity and parallelism.
2. Linearity of the samples and concentration. You were doing 2 fold dilutions...make 10 fold and go out multiple logs; perhaps you are not going far enough (ie your samples are highly concentrated and you need to dilute further)You may also be seeing some type of 'hook effect'.
3. Heterophile abs. I do not know the sample type you are running but human/animal samples may have heterophile abs which require a completely different approach.
3. Could there be some contamination of the solution used to dilute the samples: is it the same as the standards?

Although I do not perform this test these are the best ideas I have.

It is sad that the kit mfg can not help; you paid for the kit...are their other suppliers? If kits are roughtly the same...you could contact another manufacturer and provide details ...maybe they can assist.

I would not use that supplier again as all they can say is redo the test...ie BUY MORE OF OUR PRODUCT. Please share their name with this forum so others can avoid them.

Good luck.

-sgt4boston-