LemnaTec ForzaHTS detects and quantifies the mobility of small organisms in microtiter plates. The plates are placed into the LemnaTec positioning system. Starting the image acquisition and analysis programs the user chooses a test program according to the used organism, that automatically activates the suiting methods for acquisition and analysis. During the acquisition process the plates are identified by their barcode followed by the positioning of the camera into the scanning position. There the CCD camera samples the required images in chosen intervals. Then the camera is positioned at the next position for either barcode reading or image acquisition. In this way images off all plates are digitised successively. Even while taking the pictures the analysis starts in the background. As a first product a representational image of the movements is calculated from the single pictures. This new image is optionally stored with its original picture. In the next step of analysis, all movements in the wells are quantified and classified by previously defined classes of mobility. The result is a distribution of mobility for every well, stored in the database, together with the original pictures and raw-data. In the images and in the table plotted from the database certain, predefined hits can be highlighted. All controls in the tests are undergoing the same procedure so a continuous quality control of the tests is possible.

Stored images and data

The image representing the movements includes all information on mobility for quantitative evaluation and any later re-analysis. From the original picture remarkable findings in the results can be re-constructed easily. This provides a maximum of data-validity. So können z. B. auf den Originalbildern die Inokulation mit Eiern, besondere Verfärbungen oder Markierungen für nicht auswertbare Wells detektiert werden. All data are stored according to GLP and made available to database: Images, methods of analysis employed (number of images, time between images, classes of mobility) and the parameters of the mobility (distribution of the mobility- classes, classification as living, influenced, immobile) the identification of each well (barcode ID, name of the well) and the GLP relevant data about the test (time, location, user).

Resolution and choice of image frames

The user specifies which area of the plate will be acquired as one image. From a plate with 96 wells and with the organisms tested here, 4x6 or 2x3 wells in every picture are an appropriate choice. With smaller frames (2x3 wells) the optical resolution increases whereas the time needed for the image acquisition is multiplied. To alter the resolution in general, only the camera's lens is changed and the configuration file for the analysis software has to be reconfigured. All images here were 2x3 well cut-outs.

Duration of scanning

When the tested organisms move rapidly or set each other into movements due to high amounts of organisms in rather small wells, the time between image acquisitions may be reduced. Thus reducing the overall time of scanning without a reduction in the quality of the classification. A time gap of 5 seconds between image acquisitions yields rigid data free of statistical dropouts. In the case of the organisms used for this test, it may be reduced to one second only.

Results of picture analysis

The mosquito-larva on the plates were scanned by the ForzaHTS system being simultaneously identified via barcodes or data matrices. The scanning time in this case was between one and 5 seconds. Pic. 1 shows images of 5-second sequences taken as control and an other series after a test of 2 hours with active substances. The pictures show evident differences in the size of the organisms because of the culture being not synchronized. Looking at a picture series in sequence, more distinctions are noticed in the organisms movements concerning either speed and regularity. Clouding and colourings caused by the test-substances would impair a differentiated observation by a human evaluator especially in well 2a (substance: surfactant).

Picture 1: Original images of the wells. Left: control 1-6 , right: test with active substances (2h). Active substances, wells: 2a: surfactant, 2b: control 7, 2c: control 8 , 2d: ethanol 5% , 2e: ethanol 50%, 2f: nicotine

A mobility index is introduced to quantify the movements of the test organisms. The index is scaled to be decoupled of difficult to stabilise factors like number of individuals and their size. Scaling relies on the organisms number and size and on a mathematical model of movement. The threshold for the detection of influenced organisms have to be set by a set of controls once for every test design and can be statically applied on an arbitrary number of tests. The scaled mobility indices determined by the LemnaTec ForzaHTS system are displayed in table 1. All controls are well over 3000 mobility units. In the case of incubated organisms the highest value for mobility is 650, indicating a lucid decrease in their activity. This attenuation of mobility can also be recognized by an observer regarding the sequence of scanned images (movie-like). If a prolonged time of scanning is applied on the organisms treated with nicotine, most frequent and weak movements are noticed by an observer. These very weak mobility is detected by the ForzaHTS system and is rated with 62 mobility units.

Table 1: Integrated scaled mobilities

The mobility of small organisms can be detected with a high dynamic using the ForzaHTS software. It is extensively independent of the species of test organisms, their number, size or patterns of mobility. A false-coloured illustration of the mobility detected by the ForzaHTS software is shown in Picture 2. The activities of the test organisms were assigned to 6 classes of mobility. The classes 'very slow' up to 'very fast' were detected in all wells in the beginning of the test. Without any further analysis it can be estimated, that the classification of the wells shows some individual distinctions. This is founded in two factors: At first, the number and size of organisms analysed is varied widely. Second, the movements in direction perpendicular to the plane of observation is not detected.

As can be verified by viewing the scanned pictures movie-like, after two hours of incubation no mobilities in wells a and e (active substances surfactant and ethanol 50%) are detectable anymore. In contrary, the test objects in the control wells are still active and all classes of mobility are detected. In well d and f (ethanol 5% and nicotine) the mobility has widely ceased. Some isolated 'normal' movements were detected in well

Picture 2: False-coloured illustration of various mobilities of the tested organisms (Wells identical to picture 1).

In this way the LemnaTec ForzaHTS system reliably differentiates mobile from immobile organisms on a 96-well plate within seconds. It is not hindered by clouded or tainted samples and large individual differences in the test- populations.

Statistical evaluation of the image analysis

The mobility indices of the controls and the tests with active substances after 30 mins. are shown in Picture 3. A threshold at 3000 units can be defined based on the mobility indices of the controls. As a result all values exceeding this mobility are determined to be uninfluenced. The fluctuation of the control mobilities is founded by individual distinctions of the organisms.

Picture 3: Scans of five secs. of a control series (left) and a test with active substances (right) after 30 mins.. The threshold for the detectable effect on the mobility is defined to a value of 3000 units.

The mobility of the test organisms treated with ethanol 5% over 30 mins. is well below the threshold, whereas nicotine decreased the mobility of the test organisms even more. The organisms intoxicated with surfactent and ethanol 50% are immobile.

The results of the electronic image analysis can be verified by regarding the scanned series of images as a movie. Even a slightly diminished mobility like with 5% ethanol after 30 mins. test time can be estimated this way. Nevertheless a precise quantification or a reproducible result can not be obtained by a human observer.

Picture 4: Time depended effects on the mobility of the test organisms

In picture 4 the effects of the active substances on the mobility of the test organisms are shown in dependency of time. It is obvious, that the effects of nicotine and alcohol are not yet maximal. Anyhow they fall well below the threshold determined by the control series analysed in beforehand. By conducting time dependent measurements e.g. profiles of active substance effects can be gained.

Resume

With LemnaTec's ForzaHTS small organisms reactions on active substances can be reliably and effetely determined. The method of detection is highly significant and enables a rigidly classification of mobilities as desired in HTS tests. Disturbing factors like colourings and clouding don't have negative influences on the test results qualities. A scanning time of as less as 5 seconds yields valid data, free of statistical dropouts due to random twitching movements of small organisms without clearly patterned motions. As such the test is extensively independent of the species used. By scaling the mobility index gets independent even of the number and size of the organisms in the test. He results contain much more information then just the classification in moving and immobile organisms. Beyond this, slowed or even excited organisms can be recognised. By quantifying the mobilities even changes in the patterns of mobility can be detected. Testing the mobility in dependence of time, profiles of substances effects can be obtained and test optimisations are carried out. In combination with the Scanalyzer HTS up to 100 plates à 96 wells are evaluated in one batch. The extensive documentation of the images and data grants the option for data mining, reanalysis, quality management and GLP conformity.