Q & A

Looking for Answers in Yeast

Anuj Kumar, LSI Research Assistant Professor and renowned Detroit Tigers fan, is a biologist who uses genomics and high-throughput approaches to understand the basic mechanisms of cell's operations in the bakers' yeast Saccharomyces cerevisiae. His new database Organelle DB tracks protein locations in cells from 150 organisms.

You use yeast in your research. Can you explain how yeast works as a model system?

Yeast is a classic model system and one of the most common. Yeast is a single-celled organism and it never becomes anything more. If you look at each of our individual cells and you compare those to a yeast cell, they are not as dissimilar as you might think. So, the information we learn about proteins within a yeast cell is actually relevant to the similar proteins we have in our cells.

What are the advantages for studying the proteins in yeast?

Actually, it would be great to do all your studies on human cell lines, but in practice it is difficult. We don't have some of the tools available to do those studies in human cells, and many experiments that we can do in yeast, involving very precise DNA manipulations, are not easy to perform in human cells. The idea behind any model system is that it makes that test simpler, you can study a protein or cell process to a greater degree and the information that you find is relevant to what is going on in our cells.

How are yeast and humans cells related?

We know a lot about yeast, and, for the most part, this information is quite relevant to the molecular mechanisms at play in a human cell. Evolutionarily you know the proteins are going to be slightly different, but actually the majority of them do seem to have a similar protein in humans, so you can usually infer something about function.

So how similar is yeast DNA to human DNA?

There are a certain percentage of genes in humans where an evolutionarily related protein is present in yeast. Or considered from the other direction, there are approximately 6,000 proteins in yeast and roughly 2/3 of those proteins are similar to human proteins. In those cases, if you look at the DNA encoding the proteins, or even better, the amino acids that makeup the proteins, I'd guess that maybe 30% of the amino acids would be identical or very similar. With proteins, the amino acid sequence dictates, or contributes to, shape, and shape equals function—in many cases, the human and yeast proteins are very similar.

Your research involves genomics, can you describe what genomics is?

I try to understand how a cell grows and all of its functions. There are different approaches to understand how that cell works. An individual lab may study a particular protein within a cell and learn a lot about that protein. In my lab, we collectively take all that knowledge from many individual studies and begin to build an understanding of how the cell works. We can also take the approach of stepping back and doing a broad survey, not looking in incredible detail at a particular protein, but try to look at properties related to all the proteins. That is the idea behind genomics—doing studies that would encompass a larger set of genes. Techniques in genomics examine the entire set of genes in hopes of gaining some understanding that may not be evident from examining a single protein in isolation.

Can you explain how you apply genomic approaches to understand cell biology in yeast?

I use the analogy of the cell being like a factory. So, in a cell, the genes make proteins, and the proteins carry out the functions in the cell. Think of the proteins as factory workers. There are many ways you could try to understand how that factory functions. You could try to study each factory worker individually and find out as much as you can about each person—where that person was born, is he strong, weak, etc...try to figure out what that worker does. Or, you could implement a broader survey of the factory. In the cell for instance, you might want to know when genes are expressed, when are they active? So in the analogy of the factory, you would be looking at a particular time point, let's say the night shift. You want to see what factory workers are present then—are they active and doing something that gives you an indication of their function? If you studied an individual factory worker, you may or may not identify a worker on the night shift. In fact, maybe we don't even know there is a night shift because no one has happened upon one of those workers. By taking a large-scale survey approach, you may not identify every detail about an individual worker, but, by the same token, you may be able to at least identify that the night shift happens.

Is there a particular focus of your research that directly involves human consequences?

I focus some of my studies toward a particular process where the yeast cell actually undergoes a change in its growth form under conditions of stress (nitrogen deprivation). Some strains of yeast will actually form chains of elongated and connected cells, as opposed to its normal growth form where the cells are basically isolated spheres. Scientists are interested in this because there are pathogenic fungi that also undergo this change in growth form, and these fungi can be very harmful, causing meningitis and potentially lethal infections in immunocompromised individuals. It seems that if this shift in growth form is halted in pathogenic fungi, the strains are no longer virulent. So, the genes mediating this change in growth form could make very good drug targets. We use the budding yeast as a model for the pathogenic fungi in hopes of identifying the genes involved in this process.

You have also developed a new database about proteins, Organelle DB, which was recently written about in the journal Science. Can you describe the database?

The database grew out of my interest in trying to understand protein localization, where the proteins are exactly in the cell. There is information to be gained from understanding where a protein is located. Using the factory analogy, we can guess something about what a factory worker does if the person is always found in the mailroom. The idea of the database was to try and present as much knowledge as is available describing the localization of proteins. So, Organelle DB catalogs the subcellular location of approximately 30,000 proteins in 150 different organisms. I think it's a nice resource that will hopefully be useful to a wide audience. The database is fully searchable and freely accessible online at http://organelledb.lsi.umich.edu/

back to top