Prof. Hanno Steen

An interview with Professor Hanno Steen

Professor Hanno Steen is an Associate Professor of Pathology at Harvard Medical School and Director of Proteomics at Boston Children’s Hospital. He received his PhD in Biochemistry and Molecular Biology from the University of Southern Denmark working in the laboratory of Matthias Mann on the mass spectrometric analysis of protein modifications. His research focuses among other things on development of novel and/or improved mass spectrometric and proteomic methods for the analysis of complex protein mixtures for the discovery and validation of disease markers in pediatric and neurological diseases.

What are some of the important factors in your work?

Working in clinical proteomics there is a focus on high quality data for high throughput analysis. This quality focus means how we handle clinical material is crucial. In our lab, we mostly use LC-MS based analytical methodologies. We also try very hard to avoid fractionation in order to assure the throughput necessary for clinical proteomics.

What demands do you place on sampling?

For us, sampling is extremely important. It’s one of the major challenges when working with clinical material. Your end results can only be as good as your initial samples. Here, the term “garbage in, garbage out” very much applies.
You also need to know how to analyze samples with very high reproducibility. You need to be sure that the differences detected are a true reflection of the underlying biology and not an artefact of the sample preparation. For instance, there can be huge differences between samples simply based on how long they were kept at room temperature before being prepared, or, how quickly they were frozen. When it comes to clinical samples, it can be particularly difficult to control the time from excision until the sample is placed in liquid nitrogen.

Do you have any examples you can share?

One of the most prominent examples was when we started working with pancreatic fluid, which is a highly concentrated protease solution. We could see immediate degradation if we did not handle the samples properly.

In my experience, another application where sampling is a decisive factor is phosphorylation analyses. Here, it’s crucial to control for phosphatase and kinase activity.

How does your lab deal with sampling issues?

We do a lot of optimization, so we can establish robust SOPs – standard operating procedures. The lab also needs to be able to handle the appropriate number of samples, which means having different requirements on the protocol depending on the number of samples to be analyzed. We also carry out both technical and biological repeats, so we can get an idea of the technical and biological variability.

So your sampling is very controlled?

As with all SOPs the goal is to control things as tightly as possible. It makes a huge difference whether it takes one minute or ten minutes to get a sample from an organism until it is prepared.
We rely on the surgeon, the person extracting the biopsy, to carry out sampling in a standardized and reproducible way. Obviously, the patient’s welfare comes first, but that should not prevent researchers from trying to work closely with surgeons to establish an optimal procedure. Clinical samples are very precious and we want to get as much information from them as possible. 

Does the scientific community focus enough on sample preparation?

No, I don’t think so. In fact, sample quality and sample preparation are the skeletons in the closet of the proteomics community. Much more needs to be done, especially in the clinical proteomics arena.
Funding agencies need to start realizing how critical sampling is. For instance, the NIH should insist more on reproducible sample handling.

Analysis instrumentation has undergone huge advancement in recent years. More sophisticated instruments are generating more and more data. However, it’s important to remember, data is not the same as information or knowledge.