Kari Basso, PhD

Kari, please tell us a little bit about yourself to start.

My name is Kari Basso, formally Kari Green-Church. I am currently the Director of the Mass Spectrometry Research and Education Center at University of Florida and previously I was the Director of the Mass Spectrometry and Proteomics Facility at The Ohio State University (1999 – 2014). I have been doing mass spectrometry-based research since graduate school at Louisiana State University where I was trained under Dr. Patrick Limbach I primarily started in proteomics research, and we still do a lot of proteomics. Since 2004 I have been working with lipids as well, I have loved the challenge of learning something completely new and different! I am married with 4 boys ranging from 25 years to 14 years old. I have a hobby farm where I tend a 1,100 sqft garden, 25 chickens and 4 dogs. The hobby farm is White Dog Farm, Inc and we produce hot sauces, jams, salsa, and eggs that we sell at local markets on weekends.

Being at a core facility, you must see a lot of different sample types. Do you have a favorite or least favorite sample type to analyze?

Polymers and MALDI are the most challenging (frustrating…). I love being in a core mass spec lab because we see so many different types of samples and projects. I think I have about done a project with every organism there is, we’ve even extracted proteins from the surface of rocks!

Tell us more about your first encounter with lipids in a mass-spec research setting.

I first started working with lipids in 2004 in a collaboration with Dr.’s Jason and Kelly Nichols at OSU College of Optometry looking at lipids from human tear samples. Back then, MS analysis of lipids was extremely challenging because there were no established mass spec methods, no reverse phase LC-MS methods for lipids, mostly it was direct infusion and all data had to be analyzed by hand because there were no databases to search. For example, LipidMaps was just developing. Our research goal was to identify tear lipids, which at the time were completely unknown, by hand. Some of this early work was groundbreaking work in the Optometry world as they only knew what general classes of lipids were in tears based on TLC. It was very exciting to be a contributor in the discovery of 100’s and 100’s of lipids in the tears. For example, it was completely unknown that a major fraction of tear lipids are wax esters. We studied lipids in comparison of normal patients to dry eye patients to better understand the tear film composition and how to repair it as a therapy for dry eye. We also did corresponding proteomic studies where the Left eye collection was for lipids (L for lipids) and right eye was for proteins (R for pRoteins).

What is something unique about working with lipids as opposed to other biomolecules?

Lipids are extremely challenging as they are so complex with all the isomers and pos/neg polarities. Proteomics is a breeze compared to lipids. Lipids oxidize, they have different solubilities, the have different preferential ionization modes, the solvents are not compatible with plastics, and they are extremely hard on LC columns with the harsh mobile phases. I used to joke in the early 2000s that no one did lipid mass spec because it was just too hard. People would try and then give up. Thankfully we kept at it and every year things got a little easier with break through research from other groups and databases like LipidMaps and Avanti Research providing quality standards!

What do you consider to be the greatest breakthrough in “omics” technology in the last decade?

The bioinformatic tools. Even with proteomics, so much data analysis in the early days was by hand because databases etc. just simply didn’t exist! The current bioinformatic software available from many sources is amazing to work with. However, it is also dangerous if in the hands of people that do not know how to properly use the tools.

A lot of researchers want to identify as many compounds as they possibly can in their samples. What are your thoughts on multi-omics from a mass spectrometry perspective? Do you recommend splitting samples to have a separate run for the lipidome, proteome, small metabolites, etc?

We get this request quite frequently! It is a powerful analysis if one has all the instruments, bioinformatic tools, and funds to do it! Experimental design is a critical step to ensure quality results. While it is fun to be able to do a multi-omics approach, there is also nothing wrong with focusing on a particular system to really study in-depth the chemistry of the organism.

In your most recent publication, you discuss the application of a Sulfo-Phospho-Vanillin Assay (SPVA) to normalize lipids in LC-MS analysis. A) Could you explain to the audience the importance of normalizing lipids in these studies and B) Could you give an overview of the SPVA method and how it can be used in LC-MS studies

Coming from quantitative proteomics research and using Bradford assays to normalize protein concentration, I was quite bothered that lipidomics did not have an equivalent method for lipid normalization. By accident I came across a paper that used SPVA just to measure total lipid for tear samples (no downstream mass spec, they just were measuring total lipid concentration) and I thought WOW! We could potentially use SPVA to normalize lipids for quantitative lipidomics instead of some arbitrary protein concentration measurement. From there the challenge was on to adapt the published SPVA methods to small sample volumes typical for mass spec applications and trying to better understand the SPVA mechanism.

We are developing a new method to normalize lipid concentration for quantitative lipidomic approaches. Now researchers can measure the total lipid concentration of their samples and then take the appropriate volume of sample to load the equivalent concentration of each sample. Currently most protocols measure total protein concentration and multiply by 4 to estimate lipid concentrations. We are unable to find any publication why this correlation applies or how it was determined to be so. Is the lipid: protein ratio always 4:1? In every species? In every disease state? Our manuscripts in preparation delve into this deeper and will show the problems with using protein estimation for lipid concentrations.

If you had to give one piece of advice to someone diving into lipidomics research, what would you most want to tell them?

Go slow, be thoughtful in experimental design and sample prep. Look at the raw data, don’t just depend on what the bioinformatics spits out, run quality controls!

Finally, I could not do this all by myself. I am extremely grateful and proud of Dr. Laura Bailey and Manasi Kamat who have been the pioneers in developing the SPVA for small volumes and for quantitative lipidomics. There are so many failed experiments along the way in finally succeeding. Being tenacious pays off. Dr. Jianzhong Chen and Dr. Kelly Nichols were my amazing collaborators at OSU for all the exciting lipid and protein tear work we did. My word cloud sums it up best. The size of the name is proportional to the number of years I have collaborated with them! I am very grateful for all these collaborative interactions over the years!