Application Note

Label-Free/Live Adherent Cell Con�uenceBrightfield

ObjectiveIn this application note, an automated method for detection of the area occupied by unlabelled, adherent and living CHO cells in 96-well plates, using the MIAS®-2 truVIEW microscopy reader (figure 1), is described. Automated whole-well image capture and image analysis using eaZYX®

imaging software showed accurate detection of the area covered by CHO cells from seeding densities ranging from 2000 to 24000 cells per well, corresponding with 10% and 70% confluence, respectively.

IntroductionThe MIAS-2 is a compact fully automated versatile microscopy reader combining 5 brightfield with up to 8 fluorescent modes. For this application the brightfield mode is selected. Using cell seeding experiments with the adherent cell line CHO as an example, this application note describes how the detection of the area covered by the cells in plates can be executed fully automated and in a fully non-invasive manner at plate cycle times that match high content screening requirements.

Materials/MethodsCHO cells (ATCC CCL-61) were grown in Ham’s F12 medium (Invitrogen, Cat.No. 21765-029) in 175 cm2 T-flasks. The density of the cell suspension was calibrated using a Neubauer counting chamber immediately prior to cell seeding in plates (1). 2000 to 28000 CHO cells were seeded in 200 µl medium into 96 well plates (Greiner, CellStar, Cat.No. 655180). The plates were incubated overnight in the incubator (95% RH, 37 ºC, 5% CO2) allowing cells to adhere to, and to spread at the bottom of the well. Prior to scanning with MIAS-2, plates were equilibrated to room temperature for 15 minutes. Whole well, 5 x 5 tiled images were captured using a 5x objective and 1.0x Optovar settings. Brightfield light settings were adjusted for proper illumination at the centre of the wells. A specific confluence application for adherent cells was selected for analysis. Whole well overviews with image

.)2( erawtfos gnigami-XYZae gnisu detaerc erew syalrevo tluser sisylanaData tables were imported in a spreadsheet program (MS-Excel) and graphics were prepared. The images shown in figures were JPEG compressed and contrast adjusted prior publication.

Automated Assessment of the Area Occupied by Unlabelled Living CHO Cells in 96 well plates using the MIAS®-2 truVIEW Microscopy Reader and eaZYX® Imaging Software

Bieke Govaerts, Leen Geuens, Marc Moeremans, Kris Ver Donck & Johan Geysen

Figure 1. The MIAS-2 reader equipped with eaZYX IMAGING software.

Figure 2. Graphical presentation of the percentage of the total well area versus the actual cell seeding density (n=6). Linear regression analysis results in a correlation coefficient of 0.9944.

R2 = 0.9944

0

25

50

75

100

0 10 20 30

Cell seeding density (x1000)

% o

f tot

al w

ell a

rea

BA

DC

BA

DC

ResultsThe total area, covered by the cells is automatically calculated with a specific routine of the eaZYX software. In Figure 2, data is expressed as percentage versus total well area, in other words the percentage confluence. For this adherent CHO cell line, the curve reaches a plateau at a seeding density of 24000 cells/well which corresponds to ±70% confluence. Linear regression analysis of the data in the 2000-24000 cell seeding density range results in a correlation coefficient of 0.9944, which clearly demonstrates the linearity of the method. In Figure 3 is shown that despite the variable backgrounds, the overlay image analysis result shows robust area identification from the centre to the perimeter of the well.Figure 4 shows the correlation between the MIAS-2 application and a blind visual score using computer images of the scanned samples. The visual score, average of data obtained by 3 researchers, results in somewhat lower values as compared to the results obtained with MIAS-2 and eaZYX. The image capture plate cycle time was ~20 min. for whole well scanning and ~5 min. per plate when a single image in the center of the well was captured.

DiscussionThe results indicate that the eaZYX imaging software can assess, in an accurate way, the area covered by unlabelled CHO cells, at ~10% up to ~70% confluency, using brightfield images obtained with the MIAS-2 microscopy reader at low magnification. This procedure is extremely valuable allowing to monitor when the optimal cell confluence for initiating experiments has been reached, e.g. the degree of cell confluence in cell lines may influence the activation of transcription factors such as NF-kB. Acutally, there is no automated alternative because no genuine surrogate endpoint exist for confluence. As compared to visual estimation this automated method results in a substantial reduction in hands-on time.

References1. Freshney R. Ian, 1987, Culture of Animal Cells: A Manual of Basic Technique, Second

Edition, 227-229.2. Van Osta Peter, Geusebroek Jan-Marc, Ver Donck Kris, Bols Luc, Geysen Johan, ter

Haar Romeny Bart M., 2002, The principles of scale space applied to structure and colour in light microscopy, Proceedings of the Royal Microscopical Society, 37 (3): 161-166.

Figure 3. View of a whole well with, in red, the overlay of the image analysis result. The cell seeding density of this well was 16000 cells which corresponds with approximately 50% coverage of the total well area.

Figure 4. Graphical presentation of the correlation between the image analysis result versus the visual score using computer images of the scanned samples.

Digilab, Inc.84 October Hill Road

Holliston, MA 01746 USA

Phone: (508) 893-3130 Toll Free: (800) 935-8007 Fax: (508) 893-8011 E-Mail: infor@digilabglobal.com

ANE.0641.002.01

BA

DC

BA

DC