The use of mass spectrometry to aid ADC development

Our speaker Chris Nortcliffe, Mass Spectrometry Manager at our Deeside site, discusses how Sterling uses mass spectrometry in junction with our traditional techniques to analyse antibody drug conjugates (ADCs) and other biologics to get the best quality data on these molecules.

Today, I’d like to talk to everyone about the use of mass spectrometry (MS), to aid in the development of antibody drug conjugates (ADCs). And how here at Sterling, we use mass spectrometry in conjunction with other traditional techniques to analyse ADCs and other biologics to get the real best quality data on these kinds of molecules.

First off, I’d like to talk about the biologics field in general, and where that’s heading in the market right now. And we’re seeing a real growth in the field going from traditional mAb IgG formats into new architectures.

Now some of these are heading towards smaller features such as just variable domains, things like VNARs, Fab fragments, even down to peptides and cyclic peptides, just taking the very small variable binding regions.

On the other side, we’re going towards larger more heterogeneous structures, such as bispecifics, tetra bodies, other such formats that have different protein chains that have different challenges for the analytics of how we analyse these molecules. And on top of all these molecules, we’ve got the addition of drug conjugates of various kinds coming through that add an extra layer of complexity to these molecules. So, analysing these can present different challenges going forward.

Where we fit in with this at Sterling, at our site at Deeside, is that we’re a CDMO and we specialise in the development and analytics of these protein drug conjugates. We have a lot of expertise in a variety of linkers and toxins, so looking at a variety of protein scaffolds, including mAbs and other IgGs; and the smaller things I talked about, like mini bodies, and nano bodies, down to peptide, things like that, with a variety of linkers, including cysteine and lysine links to a lot of toxins that are seen coming through the clinic such as exatcans, PBDs and more traditional things like auristatins. So we have a wealth of expertise, analysing all these kinds of molecules, and conjugating them going forward.

The particular challenge of ADCs I’d like to talk about today, is going through the partial reduction approach making ADCs. So with the partial reduction approach, we take our IgG of interest and we do a reduction to break apart the disulfide bonds into their free cysteines, and then we conjugate our payload molecule to these.

This approach is quite well established in the field as it gives a very reliable, robust drug antibody ratio in even numbers such as zero, two, four, six, and eight. Because of the breaking apart of the disulfide chains, when we conjugate these drugs on, we break apart the covalent interacts of the molecule, and the molecules are held together by non-covalent interactions, now there can be a couple of challenges in working out exactly what your DAR ratio is for these molecules.

One technique for the analysis of ADCs is hydrophobic interaction chromatography, or HIC. Now, HIC is a non-denaturing technique that relies upon the interaction of our ADC for the stationary phase based upon the hydrophobicity. On the left hand side, we can see a HIC trace with separation of the zero, two, four, six and eight drug conjugations, and we can take the peak area of these peaks to get a DAR ratio.

Now, HIC relies on the absorbance of a molecule via UV so we’re seeing and inferring what each peak is rather than being a direct measurement. Relying on inferring of peaks, some molecules might not separate very clearly, or where there are overlapping peaks, it can be difficult for some molecules to get a DAR from this technique.

Another technique we use a lot here at Deeside, is reverse phases analysis often using a PLRP column. Now again, we’re doing UV analysis with these molecules, but now we’ve separated these molecules into their heavy and light chains, with separation based on their light chains and heavy chains with each conjugation in turn.

So here we have a trastuzumab conjugate and we can see the light chain and heavy chain separating out from the other species. Now, this technique can be quantitative determining the DAR of each chain and combining them together, however, it also relies on the inferring of peaks; it can struggle if things overlap, or if there isn’t clear definition, if there are any unexpected peaks that appear, and you can’t be sure what they are. And for some peaks, such as the H3 peak down here, it can be difficult to know where a peak begins and ends.

Now how can mass spec fit in with these techniques? So, mass spectrometry is a very well established technique in the pharma and biopharma industry, working with both purity and characterisation of small and large molecules.

In contrast to the optical techniques, I’ve talked about previously, it provides specific information on the ID of species via their mass to charge ratio. So we can use this to get a lot of information on the molecules we’re looking at.

On the top here, I’ve got a light chain and a heavy chain of a protein, and we can work out things like glycan modifications on the heavy chain, we can ensure we have any conjugations or effectors on there. On the peptide level on the bottom, we use digestion as to work out the particular location of every amino acid in a peptide. And therefore, we can work out exactly where a conjugation has occurred, or any of the modifications such as oxidation or deamidation. And the mass spec we have here at Deeside is a SCIEX X500B. And we use this at line in the lab on a variety of our conjugation projects.

When analysing ADCs compared to just mAbs or other proteins, there can be increased difficulty due to the heterogeneity added to the molecule with drug conjugation, combined with that inherent in the protein. And a lot of optical technique can struggle to reveal the information that mass spectrometry can provide.

One example is lysine conjugation, when you have a lysine conjugation, you get a more stochastic level of binding that if you get this Gaussian envelope of conjugates, then mass spectrometry can be really useful for finding a DAR ratio.

Another example is ring opening, a lot of payloads have a maleimide present in them that can be ring opened to prevent reverse Michael’s reactions. And this small mass shift can’t really be measured by optical techniques, whereas mass spectrometry can be really useful to work out the level of ring opening on various payloads.

The next part of the presentation will look at a case study of various payload molecules combined to a trastuzumab scaffold and how analysis via mass spectrometry can compare with optical techniques.

So, we selected three drugs with different levels of hydrophobicity, that have been well established in ADCs. On the bottom here, we’ve got MMAE, which is an auristatin, commonly known as the vedotin, which is a reasonably hydrophobic molecule present in a commercially available ADC.

Going up, we have MMAF, which also mafodotin. Again, in a commercial ADC slightly less hydrophobic, you can see you’ve got a similar thought payload to the MMAE, they’re both auristatins. And at the top, we’ve got amanitin, now amanitin looks very different in the warhead region; a much more peptidic in nature. Amanitin derives from the death cap mushroom, and it is currently in development for an ADC, and much more hydrophilic than the other two molecule we’re looking at. So, these three covers quite a wide range of hydrophobicity that is present in ADCs.

Now we analysed these conjugates on a well-established HIC method to work out the DAR ratio. For all these molecules we took a trastuzumab scaffold and reduced it before doing the conjugations. So, they should all have the same DAR ratio as they came from the same original stock.

Looking at the yellow trace first, vcMMAE, we can see very clear separation between the two, four, six and eight peaks on this HIC trace. We get a very quantitative DAR ratio for this molecule.

Going next to the purple trace, mcMMAF, we can see the zero, two, four, we’ve got some separation though not baseline, but when we get into the six and the eight, we’re getting a lot of smearing the peaks are overlapping a lot, you really can’t get a quantitative DAR ratio for that molecule.

And then looking at the pink trace, amanitin, we’re getting a lot of smearing and overlap even on the low DAR species. This HIC method really isn’t separating those species out. So overall, for MMAE we could use this HIC method to get a DAR, but for the other two molecules, a HIC would really struggle to get a DAR for these species.

Now, we took these three molecules have been run previously on a generic HIC method and ran them on a generic mass spectrometry method. Before we ran them on the MS we did a partial digestion with IdeS followed by reduction to make it into light chain Fc and Fd regions.

Now, on the traces here we can see clear separation in the mass dimension the reconstructed spectra for the light chain, the Fd, the Fc region, and any drug conjugations that are present and then we have three very different hydrophobicity’s here, and they all ionise well enough for us to get a measure of all the drug species present. And we get a DAR molecule for all the species involved.

Now previously amanitin was very poor on the hydrophobic interaction chromatography because it’s very peptidic and has lot of ionisable points in the warhead, it actually worked very well on the mass spectrometry. So, what can be a problem molecule for HIC can be a really good molecule for mass spectrometry, but it shows how mass spectrometry under generic conditions can be really useful for getting DAR ratios in these kinds of species.

Here we have a table of a number of different conjugates we’ve produced on this particular scaffold, with either a DAR of one to two, 2.8, 3.7, 4.5, six and a high conjugation species. Now again, we worked out the DAR via hydrophobic interaction chromatography (HIC) for the MMAE species, and because we produced all these species in the same manner, they should all have the same DAR. Now comparing the LCMS DAR, the HIC DAR, we can see all across the table, a very good agreement with a small margin of error across all these molecules across all levels of conjugation. This shows how a mass spectrometry method, and a generic mass spectrometry method, can be really useful in a lot of molecules.

One disagreement here in the bottom corner, of the amanitin DAR, reading a DAR eight for the LCMS versus DAR 7.3, is actually an analytical challenge due to conjugation; it can very difficult get a high number of conjugates on an antibody, but amanitin conjugates very well, it shows how when you have an amanitin conjugate compared to MMAE and MMAF, we just get a higher level of conjugation.

The other technique I talked about with reverse phase of PLRP. Now on the data here we’ve got a conjugation with SG 3249. For these PLRP traces here we can see the light chain and the conjugated the heavy chain, the L1 and H0 are overlapping. And as they’re overlapping it can be difficult to work out how much of one species contributes to this peak, therefore we can’t get a quantitative DAR for this molecule.

In addition, the H3 peak is very low, and so we still have a lot of ambiguity around whether the H3 is present or not. Again, going for a generic PLRP method to a generic mass spectrometry method on the bottom, with the light chain FC and FD regions, we can see all the conjugate species very clearly, and we get a DAR for this species that matches other techniques quite well.

The last technique I want to talk about is native mass spectrometry analysis using size exclusion chromatography, or SEC. Now most mass spec is taking place under reverse phase conditions with denature buffers such as acetonitrile and levels of acid. And this unfolds a protein into its denatured state. With a lot of surfaces accessible sites it can charge, so here on the left hand side we’ve got a spectra for a mAb with between 30 and 60 positive charges in an unfolded denatured state.

Native analysis seems to interact more in the gas phase in a more folded state keeping its non-covalent interactions bound to each other, keeping a protein in its each folded state as it would be in solution. On the right hand side, we can see the same protein but now ionised via native SEC mass spec, you can see a much narrower charge state distribution as molecules are now folded as it would be in solution. And only surface accessible sites can become ionised, they’ve got a narrower charge state distribution.

What this means is we can get an analysis of molecules in the same way we can with HIC, the molecule being intact, rather than individual chains after reduction and IdeS.

So, here’s how the native mass spectrometry looks for an ADC rather than just for a mAb. So here we can see on the left hand side the raw charge state distribution of the molecule. Now compared to the general mAb we saw in the previous slides, things look more complicated because now as well as glycans, we also have drug conjugate present on our molecule.

On the right hand side, we do the reconstruction molecule, we’ve taken the glycans off with PNGase F. So here we can see a very reminiscent of spectra of what we’d see via HIC with a zero, two, four, six and eight peaks being very clearly visible. So, analytically it seems very similar in the data. Looking at the raw numbers for this via HIC we’ve got a DAR of 3.98 and via mass spec we’ve got a DAR of 4.03, we’re getting very good agreement in the two techniques for a reasonably hydrophobic payload. Now native mass spec could be very useful things like bispecifics or other molecules we need to get an intact DAR ratio rather than breaking apart into light and heavy chain.

Tying all this together, Sterling has a deep expertise in the conjugation and analysis of a variety of next generation ADCs. We work with numerous commercial partners on protein and ADCs across various levels of developmental pipeline, from discovery all the way up to preclinical and beyond. This broad understanding combined with a specialised lab to handle cytotoxic material makes Sterling a proven partner for analysis of these biomolecules. We have a mass spectrometer inside our lab, we do analysis at line with all our chemists at various levels of the analytical process, we can get really rapid answers using mass spectrometry and other techniques I mentioned. It is this combined use of mass spectrometry with traditional chromatographic techniques help bring a full understanding to any biopharmaceutical challenge.

With that, I thank everyone for listening to listen to me today. If you have any further questions, please visit our website and get in touch.