Associate Professor, Department of Genetics and Pathology
Name: Caroline Kampf
Title: Associate Professor, Department of Genetics and Pathology, Uppsala University
Professional background: 2003-present, group leader, tissue microarrays, Human Protein Atlas project, Uppsala University; 2002-2003, laboratory technician, Affibody, Bromma, Sweden.
Education: 2001 — PhD, Dept. of Medical Cell Biology, Uppsala University; 1999 — MPhil, Dept. of Medical Cell Biology, Uppsala University; 1991 — MSc, biomedical sciences, Uppsala University.
The Human Protein Atlas is a free archive of histological images from sections of human tissue created by researchers at Uppsala University. Launched last year at the Human Proteome Organization meeting, the Atlas, accessible at www.proteinatlas.org, currently offers users the ability to view information and images related to 718 different antibodies. An updated version of the atlas will be released at the HUPO World Congress in October.
Critical to the formation of the atlas has been its tissue microarray module, a team of lab technicians at the Swedish Human Proteome Resource's Uppsala center that uses array technology to run the experiments that result in the information and images in the database.
Caroline Kampf, a former Affibody lab technician, was selected to head the TMA module for the Protein Atlas Project in 2003. She provided an overview of her group's role in the project last month at Cambridge Healthtech Institute's GOT Summit in Boston. BioArray News caught up with her this month once she returned to her lab in Uppsala.
Can you describe how you wound up in charge of the tissue microarray module for the Human Protein Atlas Project?
Directly after my PhD thesis, I started to work for a company called Affibody as a lab technician. Affibody was part of a collaboration proceeding from the human proteome resource project, a pilot project on chromosome 21.
During that time I met Professor Matthias Uhlén, one of the founders of Affibody and now the head of the [Human Protein Atlas] project and Associate Professor Fredrik Pontén, site director of the Uppsala site. I was offered the position as one of the team leaders for one of the modules when they decided to do this project at a larger scale and in a non-profit way.
I was working with tissue microarrays and immunohistochemistry at Affibody. This [module] was suitable [for me] since I got my PhD in the Department of Cell Biology [at Uppsala University], where I was teaching anatomy and histology.
How does the TMA module fit into the 'pathway' for adding antibodies to the Atlas?
We produce monospecific antibodies at the Royal Institute of Technology [in Stockholm], and when we receive the antibodies at the Rudbeck laboratory in Uppsala, many tissues are required to do a comprehensive study of protein expression in human cells and organs. The TMAs make this possible [and] are used to visualize the protein expression with the antibodies that the project generates. Afterwards, we digitalize the immunohistochemically stained tissue microarrays and the images are annotated and sent off to the Atlas.
What kind of instruments and materials are you using to print your arrays?
To produce the arrays we have both an automatic, ATA-27 and a manual tissue arrayer from Beecher Instruments. We have been using them since the beginning of the project.
For slides we are using normal saline-treated Super Frost Plus slides from Menzel-Glaser in Germany.
Where do you get your samples and how do fixed formalin, paraffin-embedded samples affect your work?
The samples are surgical specimens from the Department of Pathology, Uppsala University Hospital archives, selected and [made] anonymous before they are included in our TMA production.
What I can say about the FFPE issue — which also came up at the conference — is that this is the way that most tissues that are coming to the pathology department are stored. They are formalin-fixed and they are embedded in paraffin. This is the resource of tissue that we have.
We are aware of the issue using FFPE tissues. However, this is a high-throughput project, and the only way to do this is to utilize the almost unlimited resources of archive material available to us.
With the methodology we use, we are trying to minimize any problematic effects FFPE has on the results. But of course, it's a high-throughput project. It will use a lot of tissue. We [in the project] are using the same protocols and are using different antigen retrieval methods to optimize the results as best as we can with the tissue that we have.
What are some of the challenges you have faced in the project?
The hardest thing is not the methods, but the high scale of producing many arrays and antibodies. There is a challenge to take care of so much data and so much material every day.
But right now we are very satisfied with the setup and everything is working really nicely.
During the conference you said that you were holding 'annotation jamborees' where you invite researchers to Uppsala to annotate the results of your arrays …
We held [an annotation jamboree] once in October 2004 to get a start of the first phase of annotation and evaluation of the immunohistochemically stained slides. We invited approximately 35 pathologists from the Nordic countries, and 25 of them came.
We started with the jamboree to get pathologists interested and advertise and market the project. Many pathologists from that jamboree got very interested in continuing this project. We do annotation every day.
We have developed our own in-house annotation software, so the pathologists can now sit in their rooms in Malmo, Stockholm, or Oslo, or wherever they are and they don't have to come to us.
We are discussing having more jamborees because it's a nice get-together and you get to discuss a lot of science. It is also an efficient way of getting many annotations done and uniform evaluations. But on an everyday basis we do it online.
When can we expect another Human Protein Atlas upgrade?
We are going to upgrade the Atlas at the Human Proteome Organization World Congress in [Long Beach, Calif.] at the end of October this year. We actually had our first release at the HUPO meeting in Munich, Germany, in August 2005. We will upgrade the Atlas annually in conjunction with the HUPO meetings. The release will include the antibodies that we have added until HUPO as well as some special features. By October the plan is to have 1,500 antibodies in the database, so that's 700 or 800 more than were released last year.
In the future, we also want to include our cell studies in the release as an extra feature in the protein atlas. Our intention is to include a cell line array that has also been stained with the same antibodies as the tissue. Because a lot of research is done on cell lines, we thought it could be of interest for scientists. The cell array is also a complement to the tissue microarrays, representing tissues and tumors that are difficult to get hold of.