Human Proteome Maps

The sequencing of the human genome represented a game changing moment for science, but it didn’t provide the whole story. The human proteome, which is still not fully explored, represents a wealth of information. Identifying which proteins are being expressed at a given time, in what tissues and in what volume, could provide completely new insights into disease conditions and aid drug discovery. It’s not hard to see why so many would want to be involved in this important research challenge.

In May of this year, two teams published drafts of the human proteome in Nature; researchers at Johns Hopkins University and the Technische Universität München (TUM) (1, 2). The Pathologist spoke with Bernhard Kuster, lead author of the Munich team, about the progress of their research and the most important findings. Michael Tress of the Spanish Cancer Research Center queries the quality of the data and offers a counter opinion here.

The sequencing of the human genome represented a game changing moment for science, but it didn’t provide the whole story. The human proteome, which is still not fully explored, represents a wealth of information. Identifying which proteins are being expressed at a given time, in what tissues and in what volume, could provide completely new insights into disease conditions and aid drug discovery. It’s not hard to see why so many would want to be involved in this important research challenge.

In May of this year, two teams published drafts of the human proteome in Nature; researchers at Johns Hopkins University and the Technische Universität München (TUM) (1, 2). The Pathologist spoke with Bernhard Kuster, lead author of the Munich team, about the progress of their research and the most important findings. Michael Tress of the Spanish Cancer Research Center queries the quality of the data and offers a counter opinion here.

We’ve Come a Long Way Together

In an interview with The Pathologist, Bernhard Kuster, whose team has so far cataloged over 18,000 proteins, explained the next important steps for the mapping project.

“We have now joined forces with the Johns Hopkins team to take the research one stage further. By compiling all of our data into one central source, the ProteomicsDB, and partnering with others, such as the Human Protein Atlas [a Swedish team that is working to develop antibodies against all proteins; see sidebar, “The Human Protein Atlas”], our hope is to eventually provide as broad coverage as possible of the human proteome,” says Kuster.

Clearly, the research efforts so far have generated huge amounts of data – data which needs to be accessed and used by the scientific community at large. Both proteome maps are available online (through the ProteomicsDB and Johns Hopkins Human Protein Map) and application programming interface (API) access has been enabled for the TUM database, which allows computers to “talk” to the database.

“Non-coding” coding regions

One of the most interesting findings from Kuster’s work so far (mirrored by the findings of the Johns Hopkins group) was the discovery that some regions of the genome previously thought to be non-coding do actually code for protein. “This is especially significant as it implies that we don’t yet fully understand which DNA regions encode for proteins. I believe our findings are only the beginning; I suspect we will find a lot more ‘non-coding’ regions that have functions we aren’t yet aware of. We do not yet know what biological significance these proteins will have, but uncovering their functions is an interesting future task for us,” explains Kuster.

Missing proteins and our diminishing sense of smell

On the flip side are the so called “missing” proteins, that is, proteins thought to exist that weren’t found during the course of the study. “There are several explanations for this,” Kuster says, “the first is that the current technology simply isn’t able to detect them. Another is that they are expressed in tissues we haven’t yet looked at. The third, and possibly most interesting, is the hypothesis of “obsolete” genes. During the course of our work, we discovered that many olfactory G protein-coupled receptors (GPCRs) were missing, and in much higher proportions in comparison to other protein families. This pointed to the possibility that it was more than a technical problem or a case of examining the wrong tissue type.” Added to this theory is the work of other geneticists who have proposed that many olfactory GPCRs are no longer functional (4). We also know that humans have lost a lot of their sense of smell compared with other animals in which these proteins are active (dogs and truffle pigs being two good examples). “Even though this finding may not have far reaching clinical implications, it is nonetheless extremely interesting from a scientific perspective, and will also help in the annotation of the human genome,” adds Kuster.

The Human Protein Atlas

• Established in 2003
• Set up by a Swedish team, headed by Mathias Uhlén
• Key aim is to generate an antibody against every human protein
• Information on over 21,000 antibodies has been collected to date, targeting proteins from more than 16,600 genes
• Overall objective is to have a first version of the proteome by 2015, and a curated version by 2020

The quest for 100 percent

It is clear that mapping over 90 percent of the proteome is a significant advancement for proteomics and biomedical research in general, but Kuster believes there is still some way to go: “While the majority of the proteome is now mapped, the last 10 percent is still missing, and it may transpire that getting that last 10 percent turns out to be ten times harder than getting the first 90! It is very difficult to say when, or if ever, we will be able to claim to have a complete map.” He admits the idea of a ‘complete human proteome’ is rather a philosophical one: “We have currently set out to find one protein product per gene. But we all know that a single gene can have many protein products, perhaps even hundreds or thousands. We are still very far away from covering every variant of every protein. Despite this, we have come a long way and we are learning more than ever before.”

Next steps

As well as continuing their current work on the human proteome, Kuster and his team also want to work with diseased tissues (most of their data is currently taken from healthy tissue) in order to gain more information on protein expression in different contexts. They hope to begin similar efforts for the mouse (an important disease model), the rat (an important toxicological model), other animal species, and plants (which could prove valuable for the food industry).

Working to map the human proteome is important in and of itself, but Kuster predicts that it will also help to progress clinical research and development by supporting the discovery of new molecular disease markers, or by tracking the progress of drug treatment. “There’s a clear translational aspect to our work, although these developments will obviously arrive further in the future,” he adds.

Despite the volume of work ahead, Kuster is happy with the progress made so far. “In terms of my hopes for this project, I’m pleased to say that we are already ahead of expectations, mainly because of the excellent technology that we have at our disposal (and its ability to analyze huge volumes of data), collaborations with our academic partners, and also because of the donation of data from fellow scientists. One of the best aspects of the scientific community is the spirit of collaboration. Many people are willing to share their discoveries to provide different pieces of the puzzle, and by doing this we are able to do so much more than we could alone. It is this emphasis on collaboration and this willingness to freely share information that I find truly heartening in scientific research; without it, we wouldn’t be where we are today,” he concludes.

Professor Bernhard Kuster is the chair of proteomics and bioanalytics at the Technische Universität München (TUM), Germany.