Number of Metals Per Antibody

[nameerabaig]

Hi all,

Do you determine the number of metal atoms per antibody for custom conjugated antibodies? If so what procedure do you use and what is the typical number of metal atoms per antibody that you observe? Literature and Fluidigm says that there ought to be 100-200 metal atoms per antibody but, I am observing a wide range depending on the antibody (as low as 30 and as high as 105). Does anyone have any input on this? Thank you!

[mleipold]

Hi Nameera,

In short: in my experience, different antibodies (or, more generally, probes) label with different amounts of metal. This is generally reproducible for a given *clone*, which means that a given antibody clone may have more metal (more signal) or less metal (less signal) than a different antibody clone, even targeting the same marker and using the same polymer+metal.

Assuming 100% loading of the MAXPAR polymers the amount of metal signal that you see is based on the *distribution* of the number of polymers covalently attached to the antibody molecules in your stock solution. There's not likely to be an exact amount (eg, not likely to have 3 and only 3 X8 per antibody, on every single antibody molecule).

This can be seen in the "smear" on the SDS-PAGE gels:
Fig 4 and S1: https://doi.org/10.1002/cyto.a.23283
Fig 1: https://doi.org/10.1021/acs.molpharmaceut.7b01000

I've also heard of people HPLC-purifying their conjugates to tighten this up (and also remove any unreacted/free polymer or antibody).


Mike

[JennyChuang]

Can I ask how to calculate after acquiring the data?

[CRStevens]

I just realized this thread was in here and wanted to put my 2 cents into the conversation as we have extensive experience with this method.
As Mike has mentioned with a single clone of Ab the consistency of conjugation is pretty good. Usually somewhere in the 80-100 metal atoms/ab range for us.
How we perform the quantification of metal per antibody is upon finishing conjugation we know our final concentration of Ab is 500ug/mL based on nanodrop A280 measurements followed by dilution. We then will take 5uL of this Ab prep and digest it in 495uL of 2% HCL acid at 60C overnight. After this we then further dilute that digest in 4500uL of 2% nitric acid. From here we use a standard solution from Inorganic ventures as a standard curve on an ICP-OES instrument to determine ppb of our metal in our digested Ab solution. The result will be something like 40ish ppb. We then calculate metal atoms per antibody based on a worksheet that DVS used to have on its website. If I can find the original file, I'll upload it.

Hypothetically you could run this standard solution and also your digested Ab on the CyTOF as well if you don't have access to an ICP-OES or ICP-MS, but it might not be as sensitive.
It really is about converting your ppb to metal atoms per mL, then dividing metal atoms per mL by Ab/mL.

-Chad

[mleipold]

Hi Chad and Jenny,

Chad: your method reminds me of the Multielement ICP-MS standard that was used in the 2015 Tricot et al paper (and which Stanford HIMC and others have also used to characterize our instruments):
Tricot paper: https://doi.org/10.1002/cyto.a.22648
thread: viewtopic.php?f=7&t=942&p=2783

I agree, the transmission efficiency of a CyTOF is probably lower than a classic ICP-MS, but it seems sufficient for the purpose as long as you use an element standard to establish your instrument's performance. Especially if your main goal is comparing performance between clones/preps on a single instrument, you should be able to do that fairly quantitatively.


Mike

[CRStevens]

Mike you are very right the Tricott et al. paper is a great reference for this subject. I also agree for general use the CyTOF would most likely be a sufficient substitute for an ICP-OES, unfortunately I just can't say it for certain as we haven't tested it in-house. I will say if you have a general idea and continue to use that same Ab/clone, then I'd feel comfortable with a range +/- 10 metal atoms from that average for metal atoms per ab calculation.

I believe the recent paper in chemical reviews talks about this subject as well:
https://pubs.acs.org/doi/10.1021/acs.chemrev.2c00350