Carina Wollnik
Department for Theoretical Biology at the University of Bonn

About me:

During the first semesters of my biology studies at the University of Bonn, I discovered professor Wolfgang Alt's group and participated in my first seminar called "Mechanismen und Modelle der Zellmigration" in the winter semester of 2006/07. The interdisciplinary work fascinated me. So many seminars and some practical courses followed. Although I had the chance to learn more about other disciplines in biology, my main interest remained in understanding cell migration.
In 2011 I had the chance to write my diploma thesis about single cell migration at professor Alt's group. The diploma thesis title is "Quantifying lamella and body dynamics in two-dimensional cell motility" and all experimental work has been done at the Forschungszentrum Juelich, Institute of Complex Systems (ICS-7 (former IBN-4), Biomechanics group).

After finishing my diploma thesis in professor Alt's group, I wanted to focus on cytoskeletal behaviour. Started in march 2012, I am very glad to do my PhD in the group of Dr. Rehfeldt at the Third Institute of Physics - Biophysics in Göttingen.

Diploma thesis:

To understand how a cell migrates, we concern ourselvers with single cell tracking of normal human epidermal kerationcytes (skin cells). We track the cell outline and the cell body region, to determine how the migration path is influenced by the lamella. The lamella is located in between the cell outline and the cell body region.
Cells bear some similarity with a fried egg. The central body region would be the yolk part of the egg, including the nucleus and the cell organelles. The flat lamella region would be the white of the egg part, containing cell cytoskeleton components and associated proteins.

For tracking the cell outline as well as the cell body region, we use a so-called stochastic chain algorithm. The steps in cell tracking, done in this thesis, are:

  1. Time lapse movies are recorded with a Zeiss inverted fluorescence microscope, from 400 to 1000 pictures in total. Two types of pictures were recorded simultaneously:

    a) Phase contrast pictures.
    In these kind of pictures, characteristic bright halos occur, according to the local variation of mass density. Specifically, there is a fairly small halo around the cell, which enables us to detect the cell outline by by means of a so-called stochastic chain algorithm (see c, d; blue cell outline in d ).

    b) Fluorescence pictures.
    In a cell, there are different so-called cytoskeletal structures to maintain the cell shape, to enable cell migration and also to help transporting molecules or organells. One of these structures are the microtubuli, small tubes, composed of filaments of tubulin molecules. Microtubuli are mainly located in the cell body region with extensions into the lamella.

    So we transfected human epidermal keratinocytes with pEYFP-tubulin to stain the cell body region by means of laser-stimulated fluorescence (see a ).
    During transfection, bacterial DNA is transferred into the cell. In this bacterial DNA (also called vector, pE), the genetic code for a combined molecule of yellow fluorescent protein (YFP) and tubulin are inserted. The cell is able to read this DNA and subsequently produce pEYFP-tubulin molecules.

    As well as the shape of the lamella, the regions of adhesive spots (or adhesion sites) around the cell might give another hint towards the migration direction. So we also transfected cells with pEGFP-zyxin. Thereby, a green fluorescent protein (GFP) is coupled to zyxin. Zyxin is a small protein located at so-called focal adhesion sites, which form the adhesive spots the cell uses to make contacts to the substrate it is migrating on.

  2. Tracking the cell outline by a stochastic chain algorithm:

    a) Finding the cell's center.
    We find the cell center by thresholding: the interior cell body usually has very bright spots because of the aforementioned halo effects. In this way, the cell center can be calculated as the center of mass of these bright spots, and is located in the cell body region.

    b) Transformation to polar coordinates and gradient generation.
    The phase contrast picture is transformed into polar coordinates around the cell center. Subsequently, starting at the cell center, a pixel-based brightness gradient is generated, of which then only the positive values are retained in the further analysis.

    c) Stochastic chain algorithm.
    In a specified distance from the cell center, we place multiple interpolating points, which jointly comprise the stochastic chain around the cell. Apart from white noise, there are three force-like contributions which affect the motion of those points:

    Once the cell outline is reconstructed from the halo, we are able to analyze how the cell shape changes during cell migration.

  3. Tracking the cell body.
    At first we tried the body/lamella distinction by looking at another small halo around the cell body, which is higher in z-direction as compared to the lamella. Unfortunately, at the left and right hand side of the migrating cell, the halo disappeared from time to time or was too small to clearly distinguish between halo and surrounding. The very bright and stable signal in the tubulin-fluorecence pictures at the cell body area, enables us detect the shape of the cell body very easily, again using the stochastic chain algorithm (see b. In c, the detected cell body outline is drawn onto the phase contrast picture.
After tracking the cell surrounding as well as the cell body, the region in between those two regions can be defined as the lamella.

The other main topic was to see, how the focal adhesion sites influence the cell migration direction. GFP-marked adhesion sites, by transfection with pE-GFP-zyxin do have a bright signal (see f; corresponding phase contrast picture in e ), so a threshold was used to distinguish between signal from the adhesion sites and background signal (result in g ). The background signal occurs in this case mainly because zyxin is so-called core shuttle protein and will appear in the cytoplasm as well as at focal adhesions.

As well as the center of mass of the whole adhesion site area, also the center of mass of each detectable focal adhesion site was calculated.
To see how the amount of focal adhesion sites influence the migration direction in comparison to some very strong adhesion sites, directors, pointing from the cell's center of mass towards the summed up adhesion sites center of mass, weighted with either the main brightness of each focal adhesion or the size of the adhesion site were calculated.

Also I gratefully thank the Volkswagen-Stiftung for supporting this project.

Conferences and workshops (poster contributions):