DMPK Insights #16: HR-LC/MS: A Key Consideration for Environmental Fate and Plant/Livestock Metabolism Studies

Introduction

Hello and welcome to this Pharmaron podcast, part of our DMPK Insights podcast series. We are excited to present a special vodcast episode today within our DMPK Insights series. If you are listening to the audio only version, and would like to see the extra content provided in the vodcast, you can visit our website at www.pharmaron.com/knowledge-center/podcasts/ to catch this and all of our other series episodes.

As a brief introduction, I’m Mark Simmonds, the team leader for environmental metabolism here at Pharmaron. My colleague, Dr. Dylan Williams, and I would like to present the main considerations for identification of metabolism or breakdown products generated within our environmental fate and metabolism studies. Dylan is the team leader for our metabolite identification department.

As I said, I’m Mark Simmonds, and I’ll be talking about the challenges, issues, and considerations from the environmental fate and metabolism point of view. And then you’ll hear from my colleague, Dr. Williams, on the issues and solutions from a high resolution mass spectrometry or HRMS standpoint.

Presentation Outline

As a brief outline of what I would like to present today, I’ll discuss the longstanding requirements for identification and elucidation of metabolites and degradates generated within our regulatory studies. I’ll discuss the many challenges we are facing in the identification of said metabolites or degradates. What tools we have at our disposal and their potential pitfalls. I’ll discuss the secondary techniques and methodologies we can use to aid in the metabolite identification and other techniques available to us to help in this endeavor.

Regulatory Requirements for Metabolite Identification

So for environmental fate studies and metabolism studies for that, the guidelines have been the same for many years, and for those in the field, my apologies for now, stating the obvious. Most regulations, OECD, US-EPA call for the same with major metabolites being classed as any at or over 10% of the applied radioactivity and require identification or confirmation.

Additionally, those metabolites occurring at 5% of applied radioactivity or more or two consecutive time points, or approaching 5% at the end of the study with no decline will also require identification. I do remember a time when only the 10% rule existed and still does in some of the OECD guidelines, but the 5% criteria are outlined in the EU regulation number 283/2013.

Characterization Requirements

The table here and on the next slide shows the requirements for characterization of the metabolites formed with some of the emphasis around the quantities already identified. As we can see, these are a little open to interpretation with terms like “if straightforward” or “case-by-case” or “significant attempts.”

I’ll explore this a little more, but the less than 10% TRR, total radioactive residue, and less than 0.1 mg per kilogram or the greater than 10% TRR and greater than 0.05 mg per kg criteria are fairly straightforward.

Identification vs. Characterization

To follow on, there still seems some ambiguity or room for interpretation, with respect to the terms identification and characterization for plant and livestock metabolism studies. It has long since been my approach that the identification be reserved for the so-called extractable residue, i.e., activity recovered from the initial ambient extractions using polar and nonpolar solvents, plus activity recovered from the initial assault on the bound residues with mild acids, bases, or enzymes which can potentially release metabolites or conjugates and still be reasonably analyzed.

The fact is, the harder you’ve had to try to get the activity out of your matrix, the harder it is to identify, and as such, I believe further investigation of the bound residue after this, falls into the characterize category, where you are only attempting to prove incorporation into natural products and smaller carbon moieties.

Following stronger acid and base digests, you don’t really know to what extent you’ve hydrolyzed or altered the recovered activity. As stated in the OECD 503 guideline, characterization refers to the elucidation of the general nature or characteristics of the radioactive residue; short term metabolite identification.

My reason for stating this is, that as a CRO Pharmaron are experiencing more and more requests or insistence on further analysis of the activity that we believe falls into the latter category.

Experimental Challenges

If we move on to experimental challenges based within our E-fate and metabolism studies, one question often asked by my team is where have all the easy compounds gone? Many of these challenges are linked or at least have similar issues.

Application Rates

Application rates—they seem to be falling as companies find increased efficacy while reducing the quantities applied, leading to low level sample analysis or high factor concentration.

Volatility

Volatility—either the test compound itself, or in the formation of volatile organic metabolites. This in turn leads to difficulties in concentration or trapping and failure to retain the VOCs. Liquid-liquid partition or SPE can help, but often you still need to concentrate, either to get your activity into the right solvent or if there’s insufficient activity, for robust analysis. Repeat injection and fraction collecting either into 96 well-plates, or LSC valve for direct counting can also help.

Solubility

Solubility—low water solubility, or lipophilic compounds, or the sticky compounds as we like to call them, those in which you are forced into dosing at low levels as you’re dosing into a water phase, for example, typically half the aqueous solubility.

Matrices

And finally onto matrices, usually more of an issue with the metabolism studies with a variety of crops and tissue samples, each with their own specific issue. Oilseed rape or liver samples, for example, but also for the less routine extraction solvents such as EDTA, surfactants, buffers, etc., which can add to analytical woes.

Available Tools and Solutions

So what tools do we have at our disposal? As stated before, many of these challenges are linked and have similar solutions.

Addressing Application Rates

Application rates are what they are. We just have to deal with it. Hopefully not combined with volatility. So concentration is an option. Top count, stop flow. HPLC analysis are options for low level analysis as is accelerator mass spectrometry or ‘AMS’ test item.

Handling Volatility

Volatility—we can use a variety of keepers for concentrating DMSO, ethylene glycol, for example, or additional trapping solutions, polyurethane foam bungs ‘ORBO’ tubes to try to retain the volatile organics and hopefully analyze and identify them.

Solubility Solutions

Liquid-liquid partition is an option. Would never be my preference. As you’re potentially partitioning into volatile solvent and you need to concentrate or use normal phase HPLC. Solid phase extraction can help, but you still may often need to concentrate as stated previously. Collection into well plates needs to be with or into scintillant, i.e. wet top count. As the plates can’t be dried, and this limits the number of injections you can make into a single plate.

Solubility, low water solubility resigned to the usual restraints of low level analysis. As with lower application rates, typically concentration steps are adopted, partition, SPE, evaporation, et cetera.

Matrix Cleanup

And matrices, here we would use typical cleanup techniques, partition QuEChERS, SPE, in an attempt to make a cleaner, more concentrated sample for analysis, secondary analysis confirmation. And the final statement on the slide method development, cost, time and money, and can be frustrating.

Secondary Confirmation Methods

Looking at potential secondary confirmation methods. At Pharmaron unless something different is requested by our sponsors, our primary analytical technique is reverse phase HPLC. What we then use as our secondary technique is very much down to whether it’s a straightforward confirmation of metabolites or whether there are known metabolites and degradates to identify.

Available Secondary Methods

HPLC Alternative Methods: You can use an alternative HPLC methods for confirmation, such as normal versus reverse phase. This seems less favored.

Thin Layer Chromatography (TLC): This is good for confirmation with supplied references that can’t be used for unknown elucidation. Volatile compounds also pose a problem as these are potentially lost from the plate during development or drying. One dimensional TLC can also show poor selectivity and variability. Two dimensional TLC is more powerful, but also more time consuming when restricted to a single sample per plate. Again, this technique is unsuitable for unknowns.

Nuclear Magnetic Resonance (NMR): NMR is a very powerful technique for definitive positional identification, but you need a lot of material and a very clean, pure, isolated sample. And that tends to rule it out as a viable technique for environmental fate and metabolism studies.

High Resolution Mass Spectrometry (HRMS): Preferred instrument of choice, favored by most sponsors, and naturally compatible with our primary method of analysis.

Additional Analytical Approaches

As we can see on the next slide, we have some options to help us with the analysis. Clean up SPE and concentration have already been discussed. All can help provide a cleaner, improved sample to aid the analysis.

Exaggerated Rates

Exaggerated rates 2x, 5x or even 10x normal rate to aid the metabolite ID. Perhaps more of a dated approach. I’ve personally used this in the past, but with limited success. Finding that the exaggerated rate somehow changes the metabolism profile and you don’t necessarily see the same metabolites generated or at the same levels, if at all.

Stable Isotopes

Stable isotopes used to enhance the signal ratios for HRMS, we’ve had some good experience in using 13C and 15N, but I’ll leave that to the expert. Dr. Williams, over to you for the more technical bit.

---

High Resolution Mass Spectrometry Approach

Dr. Dylan Williams: Thank you. Thanks for that introduction, Mark.

This is a brief outline of what I’m gonna present today. Firstly, I’m gonna discuss how we go about performing metabolite identification using HRMS (High Resolution Mass Spectrometry), briefly talking about the instrumentation, and focusing on the different experiments that can be performed.

And we’ll then discuss some of the challenges posed by these studies and how we can overcome them, including how the use of stable labelled isotopes can aid with identification. Finally, I will discuss some case studies plant and livestock metabolism studies.

HRMS Instrumentation Overview

There are a variety of high resolution mass spectrometry instrumentation on the market from a range of manufacturers, each with their own characteristics. Not all manufacturers instruments are included on this slide.

Thermo Instruments: As examples from Thermo, Orbitrap Exploris and the IQ-X tribrid series – with mass resolution of up to 1 million positive ion and negative ion polarity switching. In the case of the IQ-X MSn capability, that is the ability to perform fragmentation of fragment ions.

These instruments offer two fragmentation options, CID, which is Collision Induced Dissociation, and HCD, Higher Energy Collisional Dissociation. These instruments can also be coupled to an Ultra-Violet Photodissociation source or UVPD source to produce an alternate fragmentation mechanism.

Sciex Instruments: From Sciex we have instruments such as the ZenoTof 7600 – a quadrupole – collision cell Zeno trap – Time of Flight arrangement with EAD. That is Electron Activated Dissociation. In addition to CID, this instrument combines faster scanning with high resolution and an alternative fragmentation technology, EAD source.

Waters Instruments: Waters have amongst their fleet a Xevo MRT and Cyclic IMS, the Xevo MRT combines faster scanning with high resolution using a Multi Reflecting ToF, and the Cyclic IMS combines cyclic ion mobility separation with a high performance Time of Flight Mass Spectrometer both systems with MS to the E acquisition of MS/MS data on everything throughout the run.

As technology advances, instruments become smaller, more powerful, and often an increased range of fragmentation mechanisms, which in turn enable a wide range of molecules to be analyzed by high resolution mass spectrometry. For biologists, it is now possible to use high resolution mass spectrometry instruments to analyze a single cell.

Analytical Workflow

In our labs, we follow the process shown here. The sample is loaded onto the analytical column with a typical flow rate of 1 mL/min. Post column, the eluent is split 1:5 so 1 part in 5 of the flow enters the high resolution mass spectrometer, and 4 parts in 5 enters the radiodetector, be that a β-RAM, for when the sample contains high levels of radioactivity, or more typical is that the flow is split to the fraction collector, especially for samples containing low levels of radioactivity.

Data Acquisition Strategy

Our usual approach is to acquire full scan MS data, ideally in both positive and negative ion mode at the same time, together with some MS/MS fragmentation experiment on specific target masses. Note that not all instruments can acquire both positive and negative simultaneously, and a separate injection of each polarity can be made.

Our methods often contain an element of data-dependent or data-directed acquisition where MS/MS scans are triggered if a signal reaches a threshold above background. This works well where the samples are relatively clean and sensitivity is not an issue.

For example, for the first run you target as much as you can without compromising scan time, and once you have that data, you can compare it to the radiochromatogram and make sure you have identified the regions that you need.

More often than not, you’ll need a second injection to target specific masses to obtain their fragmentation pattern or on masses you may have found in the full scan data that have the potential to be metabolites or sometimes just to get better data if you need to clearer picture on how a metabolite is fragmenting by selecting different collision energies.

TopCount Technology for Low Radioactivity Samples

For samples containing low levels of radioactivity, the radiochromatogram is obtained using TopCount technology. Eluent from the column is collected as multiple timed fractions into specialized 96-well plates across the entire length of the gradient.

These plates are evaporated to dryness using a sample concentrator, which forces the radioactivity in the fraction to come into contact with a solid scintillant at the bottom of the plate. This plate is then placed on a TopCount instrument and each of the 96-wells is counted typically between 4 and 15 minutes each.

The instrument converts the photons of light emitted from the interaction of the radioactivity with the solid scintillant into counts per minute. This data is then imported into a chromatographic data system to generate a radiochromatogram which is evaluated to obtain quantitative data.

Data Analysis and Structure Proposal

The high resolution mass spectrometry data, which was acquired at the same time is aligned with a radiochromatogram, the MS data corresponding to the major radioactive regions are investigated and structures proposed.

The gold standard is confirmation against a supplied reference standard where the retention time, elemental formula and fragmentation pattern alignment of your component agrees with the authentic standard. If a standard is not available, then we perform metabolite identification, and propose a structure based on the elemental formula, isotope pattern and interpretation of the fragmentation spectra.

At this point, it may be possible to synthesize a standard of the proposed structure to definitively confirm the assignment.

Analytical Challenges in HRMS

Now we move on to some of the analytical challenges we come across.

Mobile Phase Composition

Firstly, mobile phase composition. It is important to use mobile phases compatible with LC-MS analysis such as the use of volatile buffers, which include formic acid, ammonium acetate/formate/bicarbonate.

The use of phosphate buffers are not an option for LC-MS as they cause ion suppression and cake the source in a white powdery material. There are also issues with TFA (trifluoroacetic acid) being a highly efficient scavenger of negative charge, similarly TEA (triethylamine) in positive ion where these additives in the mobile phase compete for ionization with your drug-related material and can lead to significantly reduced ionization for your components of interest.

For this reason, met-ID scientists tend to shy away from using these additives.

Sample Solvent Composition Effects

The second challenge we come across is sample solvent composition effects. Where the sample radioactivity is very low there’s a temptation to inject the larger volumes, and if a sample also contains a high proportion of organic solvent, then this can lead to split peaks or elution at the solvent front, i.e. the component of interest is effectively unretained in your column.

Therefore, careful thought is needed on sample composition versus injection volume, and at the same time taking into account the polarity of your component of interest. Improving the sample clean-up and concentration may be preferable in this case.

Extremely Polar Components

Extremely polar components provide another challenge. These are usually a small mass to charge ratio and may elute on the solvent front. Isolation and analysis using orthogonal column chemistries such as Hypercarb, PrimeSep, ANP, HILLIC, although the later requires the sample composition to be in high organic, have all been successfully used in our laboratories to either retain or identify these components.

We’ve had success with Hypercarb, a porous graphite column, and with ANP, aqueous normal phase columns, and also with some of the columns where you can control the charge and the stationary phase, such as SCHERZO columns.

Isolation Experiments

These columns seem to work well because they can still use water as the weak solvent, which is more compatible with the isolate that you have collected from the reverse phase method. With an isolation experiment, you fraction collect a particular region of interest. This is normally on the solvent front where it’s almost impossible to identify any one metabolite.

That isolate is then re-analysed on an orthogonal HPLC system, which is more likely to retain very polar components. You normally have to concentrate up the isolate which can be a reasonable volume depending on the size of the collection window, and the sample might need further clean-up since the solvent front does contain a lot of endogenous material.

Low Abundant/Difficult to Ionize Molecules

Low abundant/difficult to ionize molecules, such as ketones, pose a challenge as well as with small molecules containing no other distinguishing elements other than carbon, hydrogen, nitrogen, and oxygen. This is where use of data processing software such as MDF Filtering are useful.

Mass Defect Filtering

MDF or mass defect filtering is one of the most useful techniques for stripping out background noise, and this is something you can only do with a high resolution, accurate mass instrument. The simplest way to describe it is how far is the accurate mass from the nominal mass?

Carbon is the only atom that has an exact mono isotropic mass of 12. Oxygen is ever so slightly less than 16, and nitrogen is slightly more than 14. These differences all add up to give the mass defect.

In the example here, the structure on the left has a defect of 138 mDa and the one on the right about 61 mDa, and there are software tools available that can filter the MS full scan data to show only components with a particular mass defect.

The way this works is that you tell the software the parent structure and what metabolites might be expected because any modification of the structure will change the mass defect. What tends to be the case is that the mass defect will sit there on a continuum, but within a specific range, and you can isolate any MS data that sits on that range making any drug-related components much more obvious.

This is a really useful and powerful tool, but to get the best out of it, you need to know what metabolites might be formed.

Screening Tools

There are also some useful screening tools that search through the data for you. I’m going to use the example of Compound Discoverer here since it’s the one we use at Rushden and is specific to Thermo instruments.

Here you give the software the parent structure and a list of potential biotransformations, and it’ll search through the data and come back with a list of potential hits that you can further investigate. This can be really good at finding small peaks against the background noise that can be quite difficult to spot by eye.

Radiolabel Positioning Strategy

The question of where to place the radiolabel is often asked? Take the fictional molecule shown here. It could be expected that this molecule would cleave in half as the oxygen or nitrogen atom, in which case it would be prudent to radiolabel the molecule in the respective halves of the molecule in order to fully follow the fate of the metabolism of your molecule.

Typically, a study will be performed with the radiolabel placed in one half of the molecule and re-run with the radiolabel in the second half of the molecule. Our expert team of radiochemists at our site in Cardiff have decades of experience in synthesizing compounds with radiolabel in various positions. Fate studies are often run with two or sometimes three labels to fully characterize the metabolism.

Stable Isotope Labeling

The introduction of 13-carbon or 15-nitrogen to enhance the isotope ratios in the mass spectrometer can make drug-related metabolites easier to detect in the full scan spectra.

Example with 13C2, 15N Labeling

As an example, let’s look at the same fictional molecule we met on the previous slide. The first dose is made up of a blend of the unlabelled parent compound, the 14C labelled parent compound in methyl imidazole ring in this case, stable isotope labelled parent compound 13C2, 15N – again in the methyl imidazole ring.

The spectrum shown here is an example of how the isotope pattern could look depending on the formulation. The key point here is the visible +2Da peak at a mass to charge ratio of 304 corresponding to the 14C-isotope and the +3Da peak at the mass to charge ratio of 305 corresponding to the 13C2, 15N stable label isotope.

Therefore, in the high resolution full scan data, you would expect to see a distinctive isotope pattern, which is very useful for confirming if the mass you are looking at is indeed drug-related. The pattern observed can be tailored depending upon the ratio of the unlabelled, the 14C-labelled and the stable isotope labelled material used in the dose formulation.

Example with 13C6 Labeling

Now let’s look at the introduction of 13C6 stable labelled isotope to enhance isotope ratios in the MS full scan. If we take the same fictional molecule and prepare a dose made up of a blend of the unlabelled parent compound, 14C labelled parent compound, universally labelled in the benzene ring in this case and a stable isotope labelled (13C6) parent compound.

This is an example of how the isotope pattern could look depending on the formulation. The key point here is the visible +6Da peak at a mass to charge ratio of 307 corresponding to the 13C6 stable labelled isotope. As is typical for a 14C-Universally labelled benzene ring, you would also expect to observe ions at +10 Da at a mass to charge ratio of 311 and +12 Da, at a mass to charge ratio of 313, corresponding to 14C5 and 14C6 isotopes.

Note that as these isotopes of low abundance, they may not be visible in the full scan data of metabolites, which themselves are of a low abundance or are not easily as ionized. Therefore, in the high resolution full scan data, you would expect to see this distinctive isotope pattern.

Benefits of Stable Isotope Labeling

Stable labelled isotopes are very useful if your metabolite of interest has a simple elemental formula containing carbon, hydrogen, oxygen, or carbon, hydrogen, nitrogen, oxygen, and is of low molecular weight or where your metabolite of interest may even correspond to an endogenous component in the sample.

The use of stable label isotopes is a powerful tool in your metabolite identification strategy and is often underutilized.

Case Studies

Finally, I would now share with you three case studies where high resolution mass spectrometry was used to identify the component of interest.

Case Study 1: Plant Metabolism Study

First, lets look at an example from a plant metabolism study. In this example, the crop was dosed with three different labelled parent molecules – labelled one, two, and three. Using radiodetection, two of the labels showed the presence of the same radioactive component.

By examining the high resolution full scan and fragmentation data, the metabolite structure was proposed as the acid metabolite. A standard of this metabolite was synthesized and injected alongside the sample. The metabolite was definitively assigned by comparison of the retention time, elemental formula and fragmentation pattern of the metabolite in the sample with that of the authentic reference standard.

In general, the main differences we observe in plant metabolism studies, as compared to animal studies is the formation of sugar related conjugates such as glucose, pentose, and malonyl glucose.

Case Study 2: Photolysis Study

The second case study looks at an example from a Photolysis study. The parent molecule contains a single 14C label in the Pyrimidine moiety, with a radiolable position as indicated by the arrow. Degradation resulted in cleavage of the pyrimidine moiety which resulted in the formation of an amide metabolite, which was unlabelled, and an unstable low molecular weight radiolabelled degradant.

High resolution mass spectrometry detected both the 12C and 14C masses in the negative ion full scan although due to the low mass, no product ions were obtained. Potential low molecular weight reference standards were purchased or custom synthesized in an attempt to assign a definitive structure.

However, whilst these were able to rule-out possible alternative structures, it was not possible to synthesize the proposed degradant of interest for definitive confirmation in this case.

Case Study 3: Livestock Metabolism Study – Morpholine Ring

And finally, let’s look at an example of morpholine ring metabolism from a livestock metabolism study. For this study, the 14C radiolabel was placed in the morpholine ring. Metabolites resulting from ring-opening of the morpholine whilst still attached to the majority of the parent molecule were detected.

In addition, metabolites where the morpholine ring itself has cleaved from the parent molecule and undergone extensive metabolism resulting in a large number of small molecule metabolites, C2 to C4 in chain length including incorporation into biochemical pathways were detected.

The morpholine ring and its small molecule metabolites eluted at the solvent front on our C18 reverse phase system. An orthogonal column chemistry was required to retain separate the morpholine metabolites after isolation on this reverse phase system.

Summary

To summarize, we have discussed the requirements for identification within regulatory studies, including the difference between identification and characterization. We discussed the experimental challenges be that low application rates, volatile degradants or metabolites, low solubility and difficult matrices, which require clean-up techniques, or concentration of large volumes of sample due to lower application rates and we discussed the tools at our disposal to overcome these challenges.

We discussed methods of for secondary confirmation, such as TLC or use of normal phase chromatography, and how LC-MS is favored by most clients especially the gold standard of confirmation against the reference standard.

We discussed the analytical challenges observed by LC-MS, such as mobile phase composition, solvent effects, and extremely polar metabolites, and how these challenges may be overcome using isolation techniques, alternate column chemistries, and use of software to aid identification. And finally, we discussed how dosing at an exaggerated rate or use of stable label isotope can enhance the MS signal for detection of metabolites.

Thank you for listening.

Closing Remarks

Mark Simmonds: Thank you, Dylan. It’s been a fascinating and insightful presentation. Thank you very much for your time.

Thank you all for listening to this special vodcast episode of Pharmaron DMPK Insight series. If you missed the video extras, please visit our website at www.pharmaron.com/knowledge-center/podcasts to view the video recording and read the transcript.

We would like to remind you that our DMPK webinar series is also available on demand covering a variety of key questions related to DMPK science in drug discovery and development. Stay tuned for more podcasts in our Pharmaron DMPK Insight series.

Thank you and bye for now.