1: \section*{Materials and Methods}
2:
3: \paragraph{Cell Cultures.}
4:
5: Cultures of C6, U87 gliomas, and 3T3 fibroblast cell lines were grown in Minimum Essential Medium (MEM;
6: Sigma) supplemented with 10\% fetal calf serum (FCS; Gibco), 4 mM glutamine and
7: 40 $\mu$g/ml gentamycin in humidified air atmosphere containing 5\% CO$_2$, at 37
8: $^{\circ}$C. Primary fibroblasts were isolated from the bodies of 13 day old
9: mouse embryos and were grown to confluency in DMEM (Sigma) containing 10\% FCS.
10: Prior to the time-lapse experiments cells were passaged over the micropatterned
11: surfaces of 35 mm plastic petri dishes (Greiner) at a density of
12: $1-1.5\cdot10^4$ cell/cm$^2$. In some experiments the stripe pattern was
13: printed on a 25 mm glass coverslip, which was then placed into a petri dish.
14:
15: \paragraph{Microscopy.}
16:
17: Time-lapse recordings were performed on a computer-controlled Leica DM IRB
18: inverted microscope equipped with 10x, 20x objectives and an Olympus Camedia
19: 4040z digital camera. Cell cultures were kept at 37 $^{\circ}$C in humidified
20: 5\% CO$_2$ atmosphere within a custom-made incubator attached to the microscope
21: stage \citep{hebal00,CRRL02}. Phase contrast images were acquired in every 5
22: min for 2-3 days.
23:
24: \paragraph{Micropatterning.}
25:
26:
27: Chemically structured substrates containing 10-20 $\mu$m wide stripes
28: permitting and restricting cell attachment were prepared by microcontact
29: printing. Adhesive surfaces were coated with either 40 $\mu$g/ml fibronectin
30: with 20 $\mu$g/ml Alexa488-labeled fibrinogen (Molecular Probes), or 50
31: $\mu$g/ml FITC-labeled poly-l-lysine (SIGMA). Blocking of protein/cell
32: attachment was achieved by a subsequent covering the surfaces with
33: poly-l-lysine-g-poly(ethylene glycol) (PLL-PEG) \citep{HMV01}. For further
34: details see \citep{csucs}.
35:
36: \paragraph{Drugs.}
37:
38: Specific drugs (SIGMA) were used to affect various cytoskeletal proteins.
39: Microtubule (MT) disruption was induced by 20-100 nM vinblastine, MT dynamics was shifted
40: towards polymerization by 10-100 nM taxol \citep{taxol,MTdrugs}.
41: Cytoplasmic dyneins were inhibited
42: by 5 $\mu$M sodium-orthovanadate \citep{MBW00},
43: kinesins by 100 $\mu$M 5'-adenylylimidodiphosphate (AMP-PNP) \citep{algakinezin}.
44: F-actin was influenced by 100-500 nM cytochalasin-D \citep{cytochalasinD}.
45: Myosin-II activity was inhibited by the myosin light-chain kinase (MLCK) inhibitor ML-7 at
46: concentrations of 1 and 10 $\mu$M \citep{ML-7}. Reagents were added
47: to the culture medium after one day long time-lapse recording in normal medium.
48: Drug effectiveness was tested by standard viability assays \citep{MTT}. Briefly, C6 cells
49: treated for 24 hour with various concentrations of drugs were incubated with
50: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, 25 mg/ml). 2 hours
51: later cells and formazan crystals were dissolved in acidic (0.08 M HCl) isopropanol
52: and transferred onto 96 well plates. Optical density (OD) was determined at a
53: measuring wavelenght of 570 nm against 630 nm as reference using an SLT 210 ELISA-reader.
54: We administered the drugs in non-toxic concentrations in the time-lapse experiments. The use of vanadate in cell culture is not straightforward because the permeability of the cell membrane to vanadate can be problematic. We used vanadate in a concentration of 5 $\mu$M , as a dose, which still resulted in a reduction of overall cell viability. Time-lapse recordings did not show alterations in the rate of cell decay, cellular morphology or in the nuclear motility of
55: cells upon treatment with 5 $\mu$M vanadate.
56:
57:
58: \paragraph{Immunocytochemistry.}
59:
60:
61: Cells grown on micropatterned glass coverslips were fixed immediately following
62: the termination of the time-lapse recording (4\% paraformaldehyde in PBS for 20
63: min at room temperature ). Membranes of fixed cells were permeabilized by
64: treatment with Triton X-100 (5 min, 0,1\% v/v in PBS). Non-specific antibody
65: binding was blocked by incubation with 5\% FCS in PBS at room temperature, for
66: 1 hour. Antibodies to $\alpha$-tubulin (mouse; ExBio, Praha) and
67: $\gamma$-tubulin (rabbit; Sigma) were used at dilutions of 1 to 1000-5000,
68: respectively. Secondary antibodies to $\alpha$-tubulin were Cy3 (1:3000,
69: Jackson) or Alexa488 labeled anti-mouse IgG-s (1:1000, Molecular Probes).
70: $\gamma$-tubulin was visualised by 1.5 hour incubation with biotin-conjugated
71: anti-rabbit IgG (1:1000, Vector) followed by 1 hour incubation with fluorescent
72: avidin-TRITC (1:750, Sigma). The immunolabeled sample was compared to the last
73: frame of the obtained recording. By locating the same cells in both images, we
74: were able to identify the relationship between nuclear motion and the
75: distribution of labeled protein complexes. Thus, each immunostained frame
76: contained an assembly of cells with nuclei stopped at various stages of their
77: motion.
78:
79:
80: \paragraph{Data processing and analysis.}
81:
82: Cell nucleus positions together with the two process ends were determined
83: manually in each time-lapse frame using a custom-written software. The
84: procedure resulted in the nucleus position data $x_i(t)$, which denotes the
85: location of the nucleus of cell $i$ along the stripe at a time point $t$. The
86: duration of each nuclear migration period was determined as the elapsed time
87: between the peaks of $x_i(t)$.
88: Velocity of the nucleus was calculated from its net displacement in a 30
89: minute-long time interval as $v_i(t)=[x_i(t+\tau/2)-x_i(t-\tau/2)]/\tau$, where
90: $\tau=30$min.
91: The nuclear migration activity during a time period $T$ was characterized by
92: the average magnitude of nuclear velocities as
93: $V_T=\langle \vert v_i(t) \vert \rangle_{i,t\in T}$,
94: where the average $\langle \rangle_{i,t\in T}$ is calculated over all nuclei
95: $i$ which were oscillating within the time period $T$.
96: Drug effects were characterized by comparing nuclear migration activity of the
97: first day (control) to that of the second day (drug exposed) as
98: $V_{D2}/V_{D1}$.
99: To establish the statistical significance of the differences, velocity
100: data was sampled at 1h intervals. The resulting velocity magnitudes
101: (absolute values), characterizing the cell population before and after the
102: drug exposure, were compared by two-sample t-tests.
103:
104: \paragraph{Online supplemental material.}
105:
106: Video 1 displays the motility of C6 cells on the surface of 20 $\mu$m wide
107: Alexa488-labeled fibronectin-fibrinogen stripes. Video 2
108: shows the reversive nuclear motion in a single C6 cell on a similar stripe. Frame size: 270x60 $\mu$m$^2$. Video 3 (frame size: 400x115 $\mu$m$^2$), 4 (frame size: 250x50 $\mu$m$^2$), and 5 (frame size: 630x480 $\mu$m$^2$) show the effect of 30 nM taxol (administered at 16 h), 20 nM vinblastine (administered at 9 h), and 500 nM cytochalasin D (administered at 12 h) respectively on auto-reverse nuclear migration in C6 cells cultured on 20 $\mu$m wide stripes.
109: