Methods for Rapid Metabolite Classification

Alexandra P Platt Author

07/27/2021 Added

17 Plays

Description

This research is working on method development for LC-MS metabolomics research.

Searchable Transcript

Search:

[00:00:00.570]Hello, my name is Alex Platt
[00:00:02.700]and today I will be sharing with you
[00:00:04.150]my research this Summer, Determining a Method
[00:00:06.940]for Rapid Metabolite Classification.
[00:00:09.980]One of the reasons we're interested in this
[00:00:11.880]is to help fight bacterial infections.
[00:00:14.550]These infections are a global cause of death
[00:00:17.860]more so because these bacteria have started developing
[00:00:20.900]an antibiotic resistance.
[00:00:23.080]We are hoping that with metabolomics,
[00:00:25.430]the study of small molecules involved in metabolism,
[00:00:28.700]we will hopefully be able to understand
[00:00:30.330]the mechanism by which these bacteria
[00:00:32.400]are resisting antibiotics by comparing the metabolome
[00:00:35.140]of susceptible versus antibiotic resistant bacteria.
[00:00:42.810]One of the current challenges with metabolomics
[00:00:45.070]is the easy and accurate identification of metabolites
[00:00:48.270]from these complex cellular extracts.
[00:00:50.700]So because of that, we've been working to develop a method
[00:00:53.980]using LCMS or liquid chromatography-mass spectrometry.
[00:00:58.510]This way we can identify metabolites
[00:01:01.220]based on master charge ratios, retention times,
[00:01:03.800]and fragmentation patterns, and hopefully overall,
[00:01:08.080]make an in-house chemical library of the spectra
[00:01:11.540]to help automate the process.
[00:01:13.770]The general methods behind metabolomics
[00:01:15.960]start with metabolite extraction, for this project,
[00:01:18.890]we were using amino acids of molecules from the TCA cycle.
[00:01:23.090]From there, these molecules go through
[00:01:24.490]liquid chromatography which helps to separate
[00:01:27.230]the metabolites so we can examine them individually
[00:01:31.680]then to mass spectrometry, where they're ionized,
[00:01:34.620]and we can find the exact mass.
[00:01:37.240]From there, we do tandem mass spec,
[00:01:39.370]which gives us fragmentation patterns for each molecule,
[00:01:43.340]and from there, we can upload the spectra
[00:01:45.290]into different software and automate the process.
[00:01:48.710]However, we're having problems
[00:01:50.150]in the liquid chromatography-mass spectrometry
[00:01:52.170]portion of these methods.
[00:01:53.550]And so I've been spending my Summer trying to optimize
[00:01:55.880]these portions of the overall goal.
[00:02:00.220]One of the first things we wanted to look at
[00:02:01.800]were different columns.
[00:02:03.130]There are various factors and the first one we decided
[00:02:05.800]to look at was the length of a column.
[00:02:08.180]So here on the left, we have a 50 millimeter column
[00:02:10.590]and on the right, a 100 millimeter.
[00:02:13.590]The 100 millimeter column gives us better separation
[00:02:17.180]as we don't have all of our metabolites
[00:02:19.900]correlating at the very beginning.
[00:02:22.900]In the 50 millimeter column on the left,
[00:02:25.820]despite the fact that it was a 35 minute method,
[00:02:28.550]none of our metabolites alluded past two minutes.
[00:02:31.370]Whereas we got separation all the way to 11 minutes
[00:02:33.870]on the 100 millimeter column on the same length of method.
[00:02:39.100]We also wanted to compare with the 150 millimeter column,
[00:02:42.400]which again gave even better separation
[00:02:44.610]between even these peaks that were very similar
[00:02:48.650]retention times at the beginning.
[00:02:50.950]Because of this, we opted to go
[00:02:52.100]with the 150 millimeter column length.
[00:02:55.470]We also wanted to look at the type of a column.
[00:02:57.670]There's two main types of column, reverse phase,
[00:03:00.410]which is shown on the right and HLIC or hydrophilic column,
[00:03:03.730]shown on the left.
[00:03:05.300]From the HLIC column, the peak shapes
[00:03:08.080]are less than desirable, most of them are bi-modal
[00:03:11.680]and almost all of them are very broad.
[00:03:13.680]Because of this, and the better peak shapes
[00:03:16.110]given from a reverse phase column,
[00:03:18.130]our final column we ended with was a 150 millimeter
[00:03:20.990]reverse phase column.
[00:03:23.210]We also wanted to examine different mass spec parameters,
[00:03:26.840]such as the electrospray ionization voltage.
[00:03:30.280]We could say that this was an issue
[00:03:31.870]when we tried to do tandem mass spec.
[00:03:33.970]Here at the top, the tandem mass spec spectrum
[00:03:37.840]for Phenylalanine targeting a master charge of 166
[00:03:41.660]and on the bottom, targeting Threonine
[00:03:43.730]or a mass to charge of 120.
[00:03:46.500]However, with the highlighted peaks,
[00:03:48.440]we can see that these spectrum are very similar
[00:03:50.630]showing that instead of fragmenting Threonine at the bottom,
[00:03:54.960]we're actually further fragmenting Phenylalanine.
[00:03:59.050]And this shows that this Phenylalanine fragment
[00:04:03.130]is making it through the quadrupole selection,
[00:04:05.500]showing that it's happening during that initial ionization,
[00:04:08.670]rather than further on when we are trying to fragment
[00:04:11.870]these molecules.
[00:04:13.530]This is further supported on our top spectra
[00:04:16.170]because we are missing our precursor ion of 166.
[00:04:20.440]Usually, we would have some of the precursor ion remaining,
[00:04:23.540]but for here because so much of Phenylalanine
[00:04:25.940]has already fragmented, there's a minimal amount
[00:04:29.400]of precursor when we began actually targeting
[00:04:31.990]that master charge, which is why it doesn't show up here
[00:04:34.640]and is again, suggesting that in-source fragmentation.
[00:04:38.940]In order to combat this, we started changing around
[00:04:42.340]the voltages applied during the electrospray ionization,
[00:04:46.790]specifically the capillary voltage
[00:04:48.540]in the sample cone voltage.
[00:04:50.760]Here on the left, we have a graph showing the peak area
[00:04:54.030]or general intensity of Phenylalanine and on the right,
[00:04:57.700]it is shown for the fragment.
[00:04:59.750]What we wanna see here as high intensity for phenylalanine
[00:05:02.750]and lower intensity for that fragment,
[00:05:05.470]which is shown at the lower capillary voltages
[00:05:08.970]about 0.5 kilovolts.
[00:05:11.930]We also took a look at the ratio
[00:05:14.490]between the phenylalanine precursor and fragment.
[00:05:17.720]Because we want the precursor to be so much more intense
[00:05:20.020]than the fragment, we're looking for this ratio
[00:05:21.990]to be as high as possible, which is achieved
[00:05:24.330]at this green triangle at capillary voltage of 0.5 kilovolts
[00:05:28.420]and a sample cone of 20 volts.
[00:05:30.690]Because of this, this is where we determined
[00:05:33.870]to be the best factors for our mass spec.
[00:05:38.880]And this is seen again here,
[00:05:40.430]you can see where we started at the top left
[00:05:42.590]at a capillary voltage of three kilovolts
[00:05:45.300]and sample kind of 40 volts, the fragment,
[00:05:47.650]a significantly more intense than that precursor.
[00:05:50.630]But as we alter these settings,
[00:05:52.030]we can get it so that the signals are about the same
[00:05:54.530]and then the precursor is finally more intense
[00:05:57.010]than that fragment, which is what we want to see.
[00:06:00.650]Our final LC methods that we settled on
[00:06:02.750]and had a column temperature of 30 degrees Celsius,
[00:06:05.830]a flow rate of 40 microliters per minute,
[00:06:08.830]we were using a 150 millimeters C18 reverse phase column,
[00:06:13.150]and this gradient with mobile phases water,
[00:06:15.350]and as seen on that trial.
[00:06:17.810]For the mass spec portion of this,
[00:06:19.330]our final ESI settings had a capillary voltage
[00:06:22.750]of 0.5. Kilovolts, a sample cone voltage of 20 volts,
[00:06:26.570]and a source offset voltage of 80 volts.
[00:06:29.970]From this project, we were able to determine
[00:06:31.950]that longer columns are best from a tablet separation
[00:06:35.030]and reverse phase columns give a better peak shape.
[00:06:38.220]In the future, we would like to alter the LC gradient
[00:06:41.350]and shoot for a shorter method to be more efficient
[00:06:43.900]in our separation while still getting that same degree
[00:06:46.780]of separation.
[00:06:48.560]We were also able to determine that current methods
[00:06:50.960]are leading to in-source fragmentation of phenylalanine,
[00:06:54.060]at least partially due to high sample cone
[00:06:56.440]and capillary voltages.
[00:06:58.740]As we continue changing these voltages,
[00:07:00.680]we want to monitor larger metabolites,
[00:07:03.120]such as lipids to make sure that they are still ionizing
[00:07:06.160]and that we can detect them at these very low voltages.
[00:07:09.690]We would also like to examine other factors
[00:07:11.640]that are possibly contributing to interest fragmentation,
[00:07:14.240]such as that source offset about voltage,
[00:07:17.210]as well as the temperature of our system.
[00:07:20.740]For this project, I would like to thank the NSF,
[00:07:23.020]NIH, and Nebraska CIBC as well as the University
[00:07:26.080]of Nebraska, Lincoln and Truman State University.
[00:07:29.470]I would also like to thank my mentors, Isis and Alana,
[00:07:33.030]as well as my advisor, Dr. Powers, thank you.

The screen size you are trying to search captions on is too small!

You can always jump over to MediaHub and check it out there.

Comments

0 Comments

Methods for Rapid Metabolite Classification

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments