Final Report Summary - SMW (Single Molecule Workstation)
Executive Summary:
Light and atomic force microscopy are complementary techniques for the study of live cells. With light microscopy, the interior of living cell is accessible for observation via fluorescence emission, and when light microscopy is combined with optical tweezers, molecular interactions can be observed even inside an intact cell. In contrast, AFM provides resolution on the cell surface topography at the nanometre scale and allows for studying mechanical interactions of membrane components. With SMW, all conceivable functions of these complementary technologies have been integrated into a single and unified Single Molecule Workstation (SMW) concept. The microscopy hardware to be developed has been combined with a single, highly intuitive graphical user interface (GUI) integrating all functional modules of the system, exchanging and controlling measurement protocols in real time and in highly synchronised fashion. The application of this SMW instrument beyond this project will enable major advances in life science research by delivering new insights into the molecular biology of living cells and complex biomolecular processes, which are yet inaccessible.
The main objective of the project was to create a flexible, modular and user-friendly workstation combining the three most important microscopic techniques ultra-sensitive inverted light microscopy (ILM), atomic force microscopy (AFM) for studying cell topography and molecular recognition imaging, optical tweezers (OT) for the detection of ultra-low forces and additional photo-thermal nano-spectroscopy (PTNS) to allow for the chemical characterisation of cellular material. These tasks have been worked on by the SMEs and the industrial partner of the consortium. On a second level, the SMW instrument has been applied for the study of the correlation between structure and function of living cells, with exemplary applications in molecular immunology and basic cancer research as performed by the academic partners.
The project started in December 2008 and the second period ended in November 2011. After the significant advances in the first reporting period (design of the mechanical, optical and electronic concept and software for the combination of AFM and ILM, implementation of SLM technology, establishment of test systems for cancer detection and immunology) a prototype of a combi AFM-ILM instrument with integrated SLM was built in the second reporting period. Its functionality was intensively tested and the observed interferences and problems with light efficiency could be successfully solved. Furthermore, the PTNS technology has been established and tested on living cells for the characterisation of cancer characteristics. After the optimisation of the SMW setup, it has been validated by the academic partners in a variety of experiments. The application of the system to study cancer cell characteristics and immunological T cell activation both by AFM and fluorescence microscopy techniques yielded interesting results that prove the capability of the developed SMW.
Results will be published in due time on the project website www.till-id.com/smw.html where also contact details and information of the Coordinator and the consortium can be found.
Project Context and Objectives:
Light- and atomic force microscopy (AFM) are complementary technologies for studying live cells. While light microscopy renders the interior of a living cell accessible for observation, AFM permits nanometre resolution on cell surfaces and allows not only viewing the cell surface topography, but is also applicable to study mechanical interactions of single membrane components. Light microscopy combined with optical tweezers can detect molecular interactions even inside an intact cell. The aim of our project thus was to integrate all conceivable functions of these complementary technologies into a single unified “single molecule workstation” concept. A single, highly intuitive graphical user interface (GUI) that integrates all functional modules of the system should exchange and control measurement protocols in real time and in a highly synchronised fashion. To achieve these goals it was not sufficient to take two separate, already completed technology approaches and try to merge them. Instead, the consortium needed to design the methodologies with the aim of a joint functionality in mind.
In terms of the technological aspect, the main goal of the project was to develop an innovative single molecule workstation (SMW) combining the most important microscopic techniques into one workstation:
• Inverted light microscopy (ILM)
• Atomic force microscopy (AFM)
• Optical tweezers (OT)
• Photo-thermal nano-spectroscopy (PTNS)
Thus, in the first phase of the project, the objectives of the technological concept development mostly were in the foreground:
1. Electronic and mechanic integration of AFM and ILM including system specification, noise characterisation, interface definitions, electronic controller design and integration and software design and architecture.
2. Design of a unified and intuitive user interface including customer workflow analysis, streamlining sample handling, general controlling calibrations and intelligent control interface enabling ease of use of the setup.
3. Integration of spatial light modulator (SLM) for beam shaping, phase contrast enhancement and optical tweezers for dynamic phase shaping, shaping the spatial intensity pattern in the specimen plane, holographic optical tweezers and position detection system
With this integrated workstation, a new quality level in the study of the molecular biology of living cells and biomolecular processes can be reached, enabling major advances in life sciences.
The strategic objective of the project thus was to create a flexible, modular and user-friendly microscope, which achieves a performance far beyond the current state-of-the-art.
Consequently, after the technological concept development phase, the objectives of validating the integrated system capabilities by scientific applications (cancer research and molecular immunology) were relevant in the second phase:
1. Correlated measurements by simultaneous topography and optical imaging with synchronised high-resolution topography imaging and fluorescence imaging with single molecule sensitivity.
2. Enabling space- and time-correlated cellular stimulation and fluorescence imaging with controlled cellular stimulation by positioning functionalised AFM-tips and monitoring cellular responses in real time via single molecule fluorescence.
3. Simultaneous AFM and fluorescence spectral imaging with higher harmonics topography and elasticity imaging combined with multi-wavelength fluorescence excitation and detection, localised on-site force spectroscopy with full fluorescence spectral read-out and fluorescence-guided force spectroscopy.
Project Results:
Work package 1: SMW system specification refinement and requirement analysis
Work package leader: Ferry Kienberger (Agilent)
Partners involved: TID, TILL, Anasys, Arivis, Agilent, IFJ PAN, HEPAG, JKU
Summary
The compatibility of an Agilent AFM with the TILL Photonics fluorescence microscope (inverted light microscope, ILM) is investigated with respect to noise issues, stability measures, mechanical interface definition, and optical as well as electronic implementation. Performance measures and systems specifications for the integrated single molecule workstation are setup, based on requirements analyses from cancer biology and molecular immunology, substantiated by market studies and polls to life science researchers.
Task 1.1: Requirement analysis for mechanical and electronic components
For studying single molecule processes at the cellular level there are various requirements that researchers expect from the combination of AFM and ILM. For a combined setup of AFM and ILM, including also potential optical tweezers (OT) and photo-thermal nanospectroscopy (PTNS) functionality, we did a market survey asking various researchers (including the bio-application partners from the SMW consortium, but also external life-science researchers). Three classes of researchers were addressed: (cell) biologists, biophysicists, lab managers. Many of them use the AFM as stand-alone, not on the ILM. The combined AFM/ILM has the following requirements, divided into two categories: MUSTS: (i) noise very important for biophysicists (in the 10 pN range and below; for OT below few pN); (ii) environmental control and fluid exchange is important; (iii) operational robustness: I sit down, turn on the instrument, and get the data I want; (iv) Ergonomics: easy to work with; (v) Repeatability: get the same results time after time. WANTS: (i) Automation is good but experienced users, particularly biophysicists, want to have freedom for advanced experiments; (ii) Biologists do not like writing software code, but want simplicity.
While the electrical components from the stand-alone AFM and ILM meet corresponding specifications easily, the mechanical components were investigated in more details to cover the force noise sensitivity as requested from the researchers. In particular we applied iterative CAD-modelling and high precision mechanical fabrication to get an optimal AFM-ILM interface (cf. WP2).
Task 1.2: Requirement analysis regarding mechanical stability and noise sources
Particularly for cancer biology and molecular immunology, as well as other life-science research areas, the noise level is the most important parameter for the combined AFM/ILM, according to the market survey. Stand-alone AFM noise is typically below 10 pN (with OT noise typically below few pN), with molecular forces typically ranging between 10 pN and 300 pN. Accordingly, the mechanical interface definition of AFM and ILM needs to be well defined to realize proper stability measures and reasonable low force noise of the combined setup.
Below figure shows a selected CAD model (computer aided modeling) for the mechanical interface (the quickslide) of the Agilent 5500 AFM and the TILL Photonics iMIC. The design was guided by the requirement of lowest possible mechanical noise, with design and construction know-how from related Agilent Technologies products. Thereby the 5500 AFM head is mounted on the iMIC body via its three main screws on the quick-slide. The three AFM-head screws are responsible for the coarse z-approach of the AFM-cantilever to the sample, with the sample mounted on the integrated stage of the light microscope. This mode is called free standing operating mode FSOM. This design allows having full operational functionality of the AFM and the ILM in the combined setup.
Fig. 1.1: CAD model of the mechanical interface (quickslide) of the combined 5500 Agilent AFM and the iMIC from TILL Photonics. The quickslide (left panel) is mounted on the iMIC body (blue part), and the AFM is put on top of the quickslide (red part) in free standing mode operation FSMO. This allows for full operational freedom of the individual microscopes.
Task 1.3: Performance specifications for optical and mechanical components
The first prototype combines the Agilent 5500 AFM with the TILL Photonics iMIC (the ILM). The previous part showed the design of the mechanical interface (the quickslide), which allows simply adding and removing the AFM from the ILM, without any compromise on the performance of the individual microscopes. Similarly as for the mechanical integration, we designed electronic and optical integration without compromising the individual microscopes performances. The stand-alone electronics from the individual microscopes is used and termed as ‘AFM-Controller’ and ‘iMIC-Controller’. Both controllers interact properly with the mechanics, optics, and software interfaces, as they do in their regular stand-alone settings. Since the AFM is simply put on top of the ILM, and the optical elements of the ILM are below the integrated stage, there is no compromise on the optical performance in the ILM. The same is true for OT and PTNS functionality.
Work package 2: Interface definitions and software concept
Work package leader: Stefanie Graf (TID)
Partners involved: TID, TILL, Anasys, Arivis, Agilent, IFJ PAN, HEPAG, JKU
Summary
Based on the requirements specified in WP1, the mechanical interfaces between the AFM and the ILM were defined and mechanically built in a way to allow for the lowest possible interferences between the two microscope systems. Furthermore, a software concept has been designed to enable the appropriate communication between and the synchronisation of all components.
Task 2.1: Mechanical definition of ILM-AFM interface
In the first phase of the project, the technology partners for the light microscope and AFM (TID, TILL, Agilent) defined the mechanical interface of the combined ILM and AFM platform based on the specifications elaborated in WP1. This interface “quickslide”, cf. WP1 Fig. 2.1) was mechanically built and the AFM 5500 was mounted on the quickslide on top of the light microscope body (iMIC). All investigations on the compatibility of the Agilent AFM 5500 and the TILL iMIC regarding noise, stability and interface definitions brought satisfactory results, thus the first working prototype could be used for further technology integration.
In the second reporting period, the second working prototype of a combi ILM-AFM was built using the new TILL light microscope MORE™ and the Agilent AFM 6000. Consequently, for this platform, another interface had to be designed and constructed to connect the ILM and the AFM.
In a next step, a design for the electronic and optical integration was developed. Concerning the electronics, an AFM controller and an ILM controller will interact properly with the mechanics, optics and software interfaces as they do in their stand-alone settings. Regarding the optical concept, there is no compromise on the optical performance of the iMIC ILM, because the AFM is simply mounted on top of the ILM and the optical elements (optical paths) are below the integrated stage (cf. WP1). The same holds true for optical tweezers (OT) and PTNS functionality. Anasys defined the mechanical requirements for the integration of the PTNS sensor into the AFM stage and HEPAG defined the requirements for the opto-mechanical integration of the SLM in the light microscope platform.
Fig. 2.1: Mechanical interface (quickslide) for putting the AFM 5500 on top of the iMIC. Left: prototype of the quickslide, right: 5500 AFM (black) put onto the quickslide (gray).
Task 2.2: Development of software concept for user interface
TILL and Arivis - together with Agilent – have developed concepts for the appropriate communication between and the synchronisation of all components and their implementation into the software. Communication paths and data exchange routines between the two main parts of the SMW, the AFM (including PTNS) and the ILM (including OT), have been established. With this concept, calibration routines and workflows can be realised for proper experimental measurements. Moreover, workflows for fully automated measurements have been designed.
In another step, concepts for the combined graphical user interface (GUI) were defined and implemented (see WP3). The TILL Live Acquisition (LA) software and the Arivis Browser were adapted to meet the requirements of the combi AFM/ILM platform for proper fluorescence image data handling. They communicate with the AFM software Agilent PicoView and the AFM-related imaging data analysis software Agilent PicoImage (Fig. 2.2).
Fig. 2.2: The AFM software (Agilent PicoView Version 1.8) and the AFM-related imaging data analysis software (Agilent PicoImage) communicate with the Live Acqusition (LA) software of TILL Photonics and the arivis Browser.
Work package 3: Software and electronic design
Work package leader: Christian Götze (Arivis)
Partners involved: TILL, Arivis, Agilent, HEPAG
Summary
In order to access the full potential of the atomic force microscope (AFM) and inverted light microscope (ILM) combination, several electronic control parts and a software must be provided to access the AFM and ILM devices, control the single devices, correlate their coordinate systems and combine the resulting images. These tasks were worked on in WP3. As a result an integrated software system for control and analyze experiments had to be provided.
Task 3.1: AFM software design and electronic controller design
Agilent successfully completed the design of software required for the control of the AFM, based on Agilent’s AFM software PicoView and PicoImage, which provides the AFM elements within the GUI. This concept also includes the appropriate firmware and control electronics for the AFM part of the combi setup. The firmware of the AFM controller was designed to be able to generate TTL pulses whenever predefined events occur and the resulting TTL pulse is subsequently fed into the ILM controller, which will trigger the optical measurements. The synchronisation between AFM and ILM has to be very fast and reliable due to the ms time scale of biological reactions, which need to be followed with the setup.
Task 3.2: ILM software design and electronic controller design
TILL has completed the design of the software concept for controlling the ILM based on its Live Acquisition (LA) software. The new LA GUI concept provides access to all ILM elements and allows for setting up and executing real-time protocols. The firmware concept and the control electronics for the ILM have also been re-designed and the real-time capabilities of the Imaging Control Unit (ICU) could be successfully demonstrated. The ILM controller is used to operate the light microscope, camera and light source and is driven by the ILM driver software, which is also able to send commands to the AFM driver software via a communication module.
In addition, HEPAG provided software modules for OT and phase contrast measurements as well as a driver software for the SLM control unit to be integrated into the ILM platform and the appropriate elements for use in the combined GUI. HEPAG realized a real-time capable software platform for generating intensity patterns for photo-bleaching and photo-activation purposes.
Task 3.3: ILM and AFM software integration into combined user interface
For controlling both microscope devices, hardware connections between the instruments had to be developed. The electronic implementation bases on a TTL pulse protocol, generated by the AFM, triggering the ILM and controlling basic system tasks of joint optical experiments (see Figure 3.1).
Fig. 3.1: Schematic representation of device interfaces.
Basing on these hardware connections between the devices, a software framework was designed and implemented. The project partners preferred to extend and integrate the existing special software systems over re-implementing everything in one software in order to keep all the expert software features for the integrated system.
Thus, in the SMW context there are 4 different software systems for hardware control and image analysis: the TILL Photonics Live Acquisition software (TILL LA) for ILM control, the Agilent PicoView (PicoView) software for AFM control, the HOLOEYE software for controlling the spatial light modulator (SLM) and the arivis Browser software for image overlay, visualization and analysis. To integrate these applications they were connected as shown in Figure 3.2.
Fig. 3.2: High level scheme of software scenario.
For the scenarios specified in this project, the integrated system must be able to perform three different tasks:
• Calibrating the two devices to get a transformation between their coordinate systems,
• Positioning the AFM tip interactively and
• Acquiring ILM images and correlated AFM Regions of Interest (ROI).
All of these tasks have to be performed with two different AFM types – the Agilent 5500 AFM and the Agilent 6000 AFM. Both devices differ completely, in mechanical construction as well as in electronic interfaces. Thus, different calibration and operation modules had to be implemented.
Fig. 3.3: Experiment control UI in TILL LA software.
As a result of the software development the application partners could use a powerful AFM-ILM software for their experiments (refer to Figure 3 and 4). Each of the four integrated software packages has its clearly defined tasks in this setup:
• TILL LA controls the whole experiment. It performs the calibration procedure and is used as the control center to set ROIs and pass parameters and results. Furthermore, it controls all ILM parameters.
• Agilent PicoView is used to control more advanced parameters of the AFM. Moreover, AFM images can be pre-processed in this software.
• HOLOEYE SLM software controls the instruments Spatial Light Modulator
• Arivis Browser is the visualization and analysis tool. Combined image sets can be visualised in any combination, colour properties can be set and analysis procedures as segmentation can be applied.
Fig. 3.4: Combined software for controlling the integrated setup.
Conclusion
In this work package a stable, usable toolkit for everyday work with the combined setup was developed and implemented. The existing software systems were highly integrated and the typical workflows (calibration, point scan, ROI scan) were implemented for ease of use. As the evaluation work packages showed, electronic and software of the combined setup is highly usable, even for complex and sophisticated experiments.
Work package 4: Mechanical and optical integration
Work package leader: Sven Krüger (HEPAG)
Partners involved: TID, Agilent, HEPAG
Summary
Within the project time HOLOEYE has grown into an independent fabless SLM/microdisplay producer for photonic and industrial applications. That is, the application-oriented customization of the SLMs was enabled: cell gap, LC-Material, packaging, flex, interpixel gap, AR-coatings etc.
The implemented SLM was integrated into the MORE™ microscope platform of partner TILL Photonics. Furthermore real-time hardware-accelerated CGH calculations were developed (CUDA, Clearspeed), demonstrated and tested with biophotonic applications.
Optical tweezers with multiple traps and aberration correction, SI and FRAP were implemented and tested in the microscope platform.
Task 4.1-4.3: SLM mode for dynamic phase shaping, for shaping intensity patterns and for using dynamically altered structured illumination
Task 4.4: Optical tweezers functionality for force measurements
Introduction
The general goal in the project for HOLOEYE was the implementation of the liquid crystal on silicon (LCOS) spatial light modulator (SLM) technology for biophotonic applications. This comprised the development and production of the dedicated SLM and integration of the SLM into the commercial microscope platform.
The new SLM
The existing commercial LCOS phase SLMs are working mostly in a relatively limited spectral band. This does not fit commercial requirements for biophotonic applications, because the novel microscope products implement numerous of techniques and functionalities at different wavelength from blue to IR part of spectrum. The blue-green part of spectrum is well suitable for structural illumination, FRAP and TIRF techniques, which were implemented with the developed SLM within the project. The IR part was intended mostly for optical tweezers (OT) application. The red part of the spectrum was used for SLM-based phase contrast imaging.
Started with the single wavelength version of our phase-only SLM PLUTO NIR with HD resolution, which was compatible with the wavelength of 1064 nm only, we developed in the first reporting period and brought to the market a new phase-only broadband SLM PLUTO NIR-II, which was integrated into the MORE™ microscope platform of TILL Photonics. The SLM is specified for working in the 400-1064 nm spectral band with the full phase modulation over 2pi.
Further Customisation
Within the project time HOLOEYE has grown into an independent fabless SLM producer. This allowed further improvement of the SLM technology in the second reporting period, using application-oriented customisation, which comprises various options in addressing of the SLM and in design of the panel itself.
In the driving options a new technique was implemented, which decorrelates time-dependent flicker-noise, hence providing a significant reduction of this kind of noise. The fig. 4.1 show flicker noise with standard (left) and with improved (right) addressing (the amplitude mode was used for simplicity). As one can see, the flicker noise in this case is much lower (about 6 times) and the base frequency of the flicker is two times higher. The last fact is also advantageous, because in most cases higher-frequency flicker has lower affect to the performance of the system. The reduction of flicker noise, which is typical for all digitally addressed microdisplays, is important for certain applications, because it affects the performance of OTs, as well as the performance of SI/FRAP, especially when pulsed light source is needed.
The other approach for reduction of the flicker noise is the optimisation of the design of the panel itself. The flicker noise is dependent mostly on the addressing frequency, viscosity of the LC material used and the cell gap. That is why HOLOEYE produced additional versions of PLUTO NIR-II with different cell gaps in the range between 4 and 12 microns. The versions with different LC materials were also implemented.
To analyse phase temporal response in more general way, HOLOEYE implemented a high-speed interferometer, based on CCD-line camera, which allows measuring phase response of the panel at high frequencies (up to 32kHz).
The problem of the panel curvature also was addressed here. The curvature of the panel introduces aberrations into the system, which can be compensated with the panel itself. This software solution is based on the measurement of the wavefront from non-addressed panel (fig. 4.2 left). The addressing of the certain phase map compensates the curvature. The procedure can be made at low cost using a simple interferometric setup and the software, where phase map can be represented e.g. with Zernike polynoms. The interferogram of the compensated panel is shown at fig. 4.2 (right). One can see, that only the addressable area of the SLM is compensated. The real-time compensation can be achieved using developed Framework PhaseFrame (see below), where the static phase map will be subtracted from the holograms, addressed to the display.
Nevertheless a hardware solution is preferable here, so other approach was to improve design of the panel itself as well as the packaging process. This allows in general improving flatness of the panel (fig. 4.3). The performed analysis showed, that the thickness of the display coverglass affects the flatness of the panel. The thicker coverglass helps to keep panel flat, that is why the versions of NIR-II with the coverglass thickness of 1.1 mm were produced. Additional option was the usage of the ceramic stiffener instead of metall.
Further developments include improvements in light and diffraction efficiency of the SLM. Currently wafers with reduced inter-pixel gap (200 nm vs. 500 nm) are ordered and processed. The SLMs based on this wafers will be available in the near future.
Fig. 4.1: Flicker for standard (left) and for optimised (right) addressing.
CGH calculations towards biophotonic applications
During the second reporting period, HOLOEYE used part of resources to implement an advanced software package in order to help developers to realize own real-time algorithms for SI, FRAP and OT using GPU calculations.
Fig. 4.4: Screenshot of the sample program, written with the Framework: Real-time hologram calculation of the checker board pattern.
For advanced researchers HOLOEYE developed PhaseFrame Framework (Fig. 4.4) which enables calculation of holograms on the nVidia graphic card (GPU), using CUDA programming language. This Framework allows developers to implement quickly real-time FFT (for up to 3 SLMs simultaneously) using as input a still image, image sequence, video data or video stream from extern camera as well as data from frame buffer.
Integration of the SLM into the microscope platform
With the integration of SLM into the MORE™ microscope platform, the following techniques were enabled with the SLM: SI, SI TIRF, FRAP, OT.
OTs were demonstrated with multi-traps with help of the optical tweezers software developed by Institut für Technische Optik (ITO), Stuttgart. The software works in Labview environment and allows managing multiple optical traps real-time using GPU computations.
Fig. 4.5 shows two beads that are trapped in the MORE™ platform with the IR laser (825 nm). The size of each bead is 16,3 µm. During experiments also beads of other sizes were trapped (6µm, 10µm). The number of traps is limited to a large extent only with the power of the laser source.
Fig. 4.5: Two beads 16,3um are independently trapped with OT Software, image courtesy J. Madl JLU Linz.
At first, the quality of the trapping laser beam was suboptimal and we demonstrated the flexible aberration correction with SLM, which could improve the spot quality (fig. 4.6). The original non-corrected spot has a significant amount of astigmatism and coma (fig. 4.6 left). This was corrected with the optical tweezers software (fig. 4.6 right).
The realisation of the SI/FRAP/uncaging techniques is very similar with OTs, except of the laser used (480 nm). The calculation of the periodic pattern for SI is made usually with iterative Fourier transform algorithm (IFTA). At the fig. 4.7 one can see a reconstruction of the biological object, imaged with SI (colour represents the depth). For FRAP, the basic algorithms remain the same. Fig. 4.8 shows a FRAP image of the pattern with letters “BIZ”, that was generated at the sample plane onto a SIP chart. The SLM can be easily synchronized with the camera in order to make quantitative FRAP measurements. Structured illumination for TIRF, where SLM is located in the object-conjugated plane, requires block of the zero order. This can also be done with the SLM.
Task 4.5: CCD-based wide-field imaging with TIRF and focus hold system
Test experiments have been performed with the novel algorithm for structured illumination to further optimise image resolution and the signal-to-noise ratio. This new approach was favoured instead of the CCD-based imaging due to a better sectioning and image resolution.
The observed temperature variations in the ILM causing tiny changes in the output polarisation of the lasers and signal changes indistinguishable from changes of the focus position could be solved with an optimised TIRF and focus hold system. The focus sensor configuration has been also implemented into the acquisition software.
Work package 5: Hardware integration and proof of concept building
Work package leader: Rainer Daum (TILL)
Partners involved: TID, TILL, Agilent, HEPAG
Summary
In the course of the SMW project, two combi microscope setups have been mechanically built. The first workstation was made up of the TILL light microscope platform iMIC and the Agilent AFM 5500, which was fully functional and achieved satisfactory results in performance and stability tests. With this platform, simultaneous topography and recognition imaging was performed on biological samples. As soon as the novel light microscope platform MORE™ was available, the HOLOEYE SLM was implemented into this platform for multiple microscopic applications. This ILM/SLM platform was mechanically combined with the new Agilent AFM 6000 to a fully integrated, highly stable and versatile microscope platform.
Task 5.1: Implementation of SLM functionality into ILM-AFM setup
In a first step, TID and TILL have been working with HEPAG on the implementation of SLM functionalities into the light microscope platform. The SLM hardware provided by HEPAG was designed into the new TILL microscope platform and a first microscope prototype was built, in which the SLM as well as optical components for trans-illumination have been optically and mechanically implemented (Fig. 5.1). The SLM was equipped with the new “D”-type of phase modulator for a broad waveband (400-1100nm). This device has been characterized for use in OT (1064 nm) as well as in optical filter applications (630 nm).
Furthermore, the control electronics, a new DSP control-board, for the ILM were developed and the appropriate firmware for real-time control of all required ILM functions and their synchronisation with the SLM and AFM was created by TILL. HEPAG provided driver electronics for the SLM.
In order to use the same SLM both in epi-illumination mode (together with an IR-laser for optical tweezers, or with a visible laser as pattern generator for structured light or FRAP) and in transmitted light mode as filter element for contrast-enhanced viewing, a special trans-illumination mode had to be established, which strikes the sample at small angles only.
Fig. 5.1: 3D optical concept (left) and first mechanical prototype (right, turned by 90° to the left) of microscope platform with implemented SLM (upper left corner on the right image), two diode lasers and CCD camera
Agilent mechanically integrated full AFM functionalities on top of the ILM/SLM setup, including the possibility to run TREC measurements, force experiments and imaging (in contact and MAC mode). For the mechanical integration of AFM and ILM, the “normal” 4-objective-slider (movement in one direction) of the MORE™ platform had to be mechanically modified by TILL. The filter slide was equipped with the various filter sets (e.g. for Cy3- and Cy5-like dyes).
The combined ILM/AFM/SLM setup built of MORE™ light microscope and the Agilent AFM 6000 is shown in Fig. 5.2. The system was equipped with a 828 nm IR laser (for OT) and a TILL Oligochrome light source for fluorescence (with excitation filters for GFP, Cy3 and Cy5) and set up in an acoustic isolation chamber mounted on a massive stone plate.
Fig. 5.2: The SMW setup. The Agilent AFM 6000 is mounted on a TILL MORE™ stage. The HOLOEYE SLM and the 828nm laser are integrated underneath the stage (not visible here, see fig. 5.1). The setup is placed on an active dampening system. Everything is isolated acoustically.
With the SMW setup the HOLOEYE SLM application software can be used to Fourier-transform images and transmit this information to the SLM. As shown in the figure 5.3 that alignment is good enough to project images into the sample plane, at least with a 20x air objective.
Furthermore, the functionality of the AFM in the combi setup was intensively tested on fixed T24 cells and it worked properly using all required AFM modes (dynamic force imaging, force spectroscopy, TREC…. Combined dynamic imaging with both brightfield and fluorescence imaging was successfully performed using a 20x air objective (see fig. 5.4).
Fig. 5.4: Combined optical microscopy and dynamic AFM imaging on the SMW setup. The AFM scans were carried out at the marked section in the brightfield image. The AFM scans were performed in liquid.
Task 5.2: Noise level characterisation
For a fully functional ILM-AFM setup the noise level for any AFM measurement induced by the ILM is a critical point, which had to be characterised in detail. The ILM itself contains moving parts, such as filter cubes, objectives. It was clear from the beginning that there will be certain limitations for an AFM measurement in terms of what kind of movements or vibrations inside the ILM can be tolerated. For instance, moving the filter cube or changing the objective while the AFM tip is attached and performing a force-distance cycle in certainly impossible, but luckily this is not required by any real-world ILM-AFM applications.
Therefore the involved partners mainly focused on improving the mechanical interface of the ILM-AFM and the full integration of the voice-coil z-drive. Finally the damping characteristics of the ILM body (made of mineral cast) in combination whit the low-noise z-drive induced a noise level into the AFM measurements, which is sufficient to carry out force-distance cycle experiments while running a live z-stack (oil objective moves up & down while in contact with the sample cover slide). For most application the ILM objective can be set to a distinct focal position (no movement) enabling even AFM experiments requiring exceptional low noise levels. While the first test were still performed on the iMIC platform, more comprising data were recorded on the much more stable MORETM platform, which is of course the natural choice for a combined ILM-AFM setup.
Fig. 5.5: Simultaneous topography and recognition imaging; (a) signals coming from the maxima and minima of each sinusoidal cantilever oscillation are recorded as topography and recognition images in TREC. (b) A specific ligand, attached to an AFM tip via a flexible PEG tether, binds to a target molecule. (c) Simultaneous topography (left) and recognition (middle) images of CD1d-?-GalCer complexes using TCR modified AFM tips. Recognition sites are indicated by dark spots in the recognition image. The right panel shows the 3D topography superimposed with the distribution of receptor spots. The scan size was 1.1× 1.1 ?m2 with the height scale ranging from 0 to 0.28 ?m on the upper panel and the scan size was 3.2 × 3.2 ?m2 with the height scale ranging from 0 to 0.5 ?m on the bottom panel (from Duman et al., 2010).
A high signal-to-noise ratio is particularly important for simultaneous topography and recognition imaging (TREC) to explore cell shapes, membrane domains and an accurate localisation of receptor binding sites with a resolution of 5 nm. With the SMW setup, theses TREC measurements can be combined with fluorescence imaging to determine the overall expression level and the distribution of receptor sites on the cell surface. The signal-to-noise ratio achievable with the first prototype of the AFM-ILM-combi setup (AFM 5500/iMIC) was high with an apparent spot diameter of 25 nm. Results of TREC imaging experiments are shown in Fig. 5.5.
The noise levels were also measured for force-distance cycles in order to analyse the impact on force spectroscopy experiments and we achieved ~12pN peak-to-peak fluctuations in the measured force. At this range, biologically relevant receptor-ligand unbinding forces can be easily studied.
Work package 6: SMW test, validation and optimisation
Work package leader: David Grandy (Anasys)
Partners involved: TID, TILL, Anasys, Arivis, Agilent, IFJ PAN, HEPAG, JKU
Summary
This work package was the most important part of the SMW project, because it aimed at the testing and validation of the integrated SMW platform with all its functionalities (ILM, AFM, SLM/OT and PTNS) and the optimisation of noise levels and vibration, light efficiency, optical tweezers functionality and many more. Only after the technology partners have successfully finished the integration and test of the hardware and software of the SMW microscope, the application partners could start with the validation and the specification of additional requirements for their biological experiments.
Task 6.1: Verification and optimisation of ILM imaging performance
Before the integration of ILM and AFM, the optical imaging functionalities of the new ILM microscope from TILL have been tested with respect to high speed, low vibration and long-term performance with satisfactory results. In parallel the controlling and analysis software was also tested intensively. Due to some problems in the calibration process and imaging, the controlling software had to be partly redesigned (see also WP3). After re-implementation of the software, further tests were started, which showed a general improvement of the acquisition results. With the optimised ILM-AFM platform, a number of optical performance tests were performed and for instance, the measured point spread function is symmetrical and its dimensions fit to the expected value.
Task 6.2: Verification and optimisation of AFM performance
The combi AFM was tested using standard AFM imaging and force-distance cycle acquisition. High resolution topographical images of protein layers showed resolution of few nanometres while molecular recognition imaging showed single molecule detection sensitivity. After optimisation of the magnetic coil of the AFM, magnetic AC imaging MAC-mode showed proper performance in amplitude-frequency plots as well as phase-frequency plots.
As a biological test, JKU used the AFM 6000 system to study the distribution of GPI-GFP molecules on the T24 cell surface and obtained good correlation between the GFP fluorescence and the anti-GFP recognition signals. The accuracy of the overlay was proved by optical and topographical imaging of FITC beads on the same position. The specificity of the recognised sites was investigated by using amplitude block and by injecting free anti-GFP antibody molecules, both of which showing convincing results (Fig. 6.1).
Fig. 6.1: Overlay of optical and topographical images of FITC beads.
Further tests were performed (details in 2nd periodic report) and in conclusion, the combined AFM-optical setup has shown high accuracy for overlay of optical and AFM images. The good correlation between the fluorescence and the recognition images enhanced the reliability of both techniques and provided additional information of molecules on cell surface and within the cell.
Task 6.3: Test of PTNS functionality
Anasys decided to start with the characterisation of pharmaceutical drug morphology using PTNS as this is a sample system that we were very comfortable with. We hope to switch to looking at cells in the next phase. Shown below are some of the results obtained regarding characterisation of drug morphology using photothermal FTIR spectroscopy
Determining drug morphology is an important issue in the pharmaceutical industry. In this work the photothermal technique was used to investigate a model drug supplied by GSK. The PhotoThermal technique initially characterised tablets made purely of the drug in its amorphous and crystalline forms. Figure 6.2a shows the topographical image of each tablet surface whilst 6.2b shows the corresponding photothermal spectra. Whilst there are clear differences in the topography, it is not possible to state which surface represents each material. The spectrum of the amorphous material shows much broader peaks that that of the crystalline. This is a common observation with IR spectra of amorphous materials due to non-specific intramolecular interactions in this state. An investigation was then carried out into a mixed system of amorphous and crystalline materials. Figure 6.3a shows a topographical image with locations at which photothermal spectra (Figure 6.3b) were obtained. Point 1 shows a spectrum characteristic of a crystalline material whilst points 5 and 6 show that of an amorphous material. The remaining points, however, appear to show intermediate spectra. It is likely that these points are close to the interface between the two materials and that the thermal waves travelling through the materials will cause spectral mixing. Subsequent thermal analyses were carried out at these locations, using the same thermal probe, to distinguish between the materials, thus demonstrating the power of the technique to obtain information on the physical, chemical and thermal nature of the sample.
(a) Crystalline Amorphous (b)
Fig. 6.2: (a) AFM topography images (b) photothermal spectra of a crystalline and amorphous model drug.
Fig. 6.3: (a) AFM topography image showing locations at which photothermal spectra (b) were obtained.
Experimental results on cells
The procedure is described in detail in the 2nd periodic report. Four cell lines were prepared on glass slides before immobilisation on the probe. Once the particle was immobilised the tip was located at the focal point of the IR beam so that a spectrum could be measured (fig. 6.4). Spectra were acquired at a resolution of 8 cm-1 for a total of 1000 scans.
Fig. 6.4: Schematic digram of the process of immobilisation of cellular material on the tip of the probe.
The aim of this work was to compare cancerous cell lines with noncancerous lines. As mentioned previously there are four lines in total, two pairs each of homologous cell lines. Within each pair there was a cancerous and non cancerous cell line. To compare the spectra within each pair, the three individual spectra were averaged and overlaid (see detailed figures in 2nd periodic report and Fig. 6.5 below).
Concluding remarks
The results show that it is possible to obtain IR spectra of small amounts of cellular material with the PTMS technqiue. In accordance with previosuly reported work using conventional techniques, spectral signatures were observed for cancerous lines by comparing ratios of specific peaks. In these studies several measurements were made (at least ten) of each cell line to establish a true spectral signature. It was therefore agreed that many more repeat measurements would be carried out on fresh samples.
Task 6.4: Establishment of test system for cancer cell recognition
Two human bladder cell lines (HCV29 – non-malignant cancer of cell urether and T24 transitional cell cancer of bladder) were chosen as a model system used to test capability of AFM to identify cancerous cells. The studied bladder cells are different in many features - they have different morphology, actin filaments and microtubules organization, and N-cadherin expression.
The structure of cancerous T24 cells revealed quite different morphology as for example cell nucleus with irregular shape and cytoplasm less rich with organelles resulting in homogenous image of cell interior by TEM. The cell membrane is smoother, and there are fewer protrusions present on its sided. The observed differences in cell morphology of single cells were the criteria for choosing these cells as model system.
Task 6.5: Characterisation of chemical composition of cellular structures
The aim of this work was to compare cancerous cell lines with reference cell lines (non-malignant in case of bladder and less-differentiated cancer in case of melanoma). As mentioned previously there are four lines in total, two pairs each of homologous cell lines. Within each pair there was a cancerous and reference cell line.
a) b)
Fig. 6.5: a) The average IR spectra for studied cell lines: HCV29 – non-malignant transitional epithelium of urether (blue); T24 – transitional cell cancer of bladder (red); WM115 – melanoma cells from primary tumor site on skin (green), and WM266-4 – melanoma cells from metastasis to skin (magenta). b) Raw data with the fits. The ratio between peaks centred at 1035 cm-1 and 1087 cm-1 corresponds to the level of glycogen consumption.
Test measurements with Raman spectroscopy have been carried out in order to define the differences between the studied cells – human bladder (HCV29 and T24) and melanoma (WM115 & WM266-4) cell lines. The obtained results showed major difference within the range 1000-2000 cm-1 and 500-1400 cm-1 for bladder and melanoma cells. Afterwards, the IR spectra for the four studied cell lines were measured and the corresponding pairs of cells, i.e. HCV29 with T24 and WM115 with WM266-4, were compared.
To compare the spectra within each pair, the three individual spectra were averaged and overlaid (Fig. 6.5). Based on this data, the spectral signature of cancer progression was calculated. The R-value for non-malignant cells varies from 0.6 to 0.8. The R = 753 obtained for non-malignant HCV29 cells lies within the limits of this range. For malignant cells, the R-value should be equal to or smaller than 0.5 which corresponds well the obtained values for T24, WM115 and WM266-4 cells.
The results show that it is possible to obtain IR spectra of small amounts of cellular material with the PTMS technqiue. In accordance with previosuly reported work using conventional techniques, spectral signatures were observed for cancerous lines by comparing ratios of specific peaks.
Task 6.6: Test of trapping capabilities and force calibration system
After optimisation of the poor light efficiency of the OT system (cf. WP4 and WP5) and substantial increase of laser power, the trapping capabilities of the setup were tested with beads of different sizes and the results showed that the number of traps is limited only by the power of laser source.
Task 6.7: Test of AFM functionalities on living cells
The goal of this task was to established AFM functionalities by (1) characterisation of the cellular stiffness accompanied by imaging the cytoskeleton using optical microscope; (2) establishing the model for single molecule interactions; (3) elaboration of sample preparation; and (4) establishing test system for cancer recognition.
(1) Characterisation of the cellular stiffness (AFM) and actin filaments imaging (fluorescence microscope)
It is known that differences between non-malignant and malignant cells are larger than between normal and non-malignant cells. Therefore, the two human bladder cell lines were chosen as a model, reference system. They were non-malignant transitional epithelial cell of urether (HCV29) and transitional cell carcinoma of urine bladder (T24). These cells are characterised by distinct shape and distribution of actin filaments, enabling the correlation between the fluorescent images and surface topography measured by AFM (Fig. 6.6).
Fig. 6.6: (left) Morphology of human bladder (HCV29 & T24) cells visualised by AFM topography imaging and fluorescent staining of actin filaments by phalloidin labelled with Alexa Fluor 488. These changes were accompanied by distinct elastic properties (right).
(2) Model for single molecule interactions in cancerous cells
The measurement of N-cadherin expression in T24 and HCV29 cells using Western blot (left) and fluorescence microscopy (upper right corner) showed its various levels depending on the type of the studied cells (Fig. 6.7).
Fig. 6.7: Measurement of N-cadherin expression in T24 and HCV29 cells using Western blot (left) and fluorescence microscopy (upper right) and measurement of unbinding force using AFM (down right).
The larger N-cadherin expression was detected in cancerous T24 cells as compared to HCV29 ones. Both the abundance and unbinding force was determined delivering information on larger amount of N-cadherins and on larger unbinding force (above 2 fold) pointing out the formation of more stable complexes in cancerous cells. With this data, a model system basing on elasticity measurements and N-cadherin expression in human bladder cells has been developed.
(3) Sample preparation for AFM- based recognition measurements
Within this task, the detailed growth conditions and steps for sample preparation were intensively tested and defined both for the separate growth of cells and for the simultaneous growth of two cell lines. They are described in detail in the 2nd periodic report.
(4) Test system for cancer recognition
The observed distinct elastic properties of bladder cells together with morphological differences were the basis for cancer cell recognition studies. Each cell line was cultured both separately and as a mixture of two corresponding cell types i.e. HCV29 together with T24 cells.
Once a mixture of cells was prepared, the AFM-based elasticity measurements have been performed on unstained cells. For each single cell, the Young’s modulus was calculated showing the dependence of the Young’s modulus versus cell number ordered as a function of time. If there is no deviation from the mean, the obtained Young’s modulus is not affected by other external factors influencing the cell elasticity. The average elasticity of non-malignant HCV29 cells was 4-fold larger than that of cells from malignant cell cancer of bladder. The mixed cells showed two modulus values that overlapped with those obtained for each cell line separately. The obtained values of the elastic modulus were in good agreement with those measured earlier, so with this data a reliable test system for cancer recognition could be established.
Work package 7: Developing the SMW towards an accurate tool for cancer detection
Work package leader: Malgorzata Lekka (IFJ PAN)
Partners involved: IFJ PAN
Task 7.1: Cancer recognition using the SMW platform
The goal of this task was the characterisation of morphological properties of cancerous cells alone using both optical and fluorescence imaging. The cell morphology of studied cell lines (melanoma and bladder cells) has been characterized using optical and fluorescence microscopy. Two types of samples were studied. Those containing only one cell type (either non-malignant bladder HCV29 or malignant bladder T24 or not-metastatic melanoma WM115 or metastatic melanoma WM266-4 cells) or two corresponding cell types mixed (i.e. pairs of two bladder or two melanoma cell lines). Imaging was carried out as a function of time of co-existence of two cell types. Cells were visualised through staining of actin filaments using phalloidin labelled fluorescently with Alexa-Fluor 488. To verify the growth conditions the doubling time was determined for cells grown separately and in co-cultures.
A) B)
Fig. 7. 1: Morphology of human (A) bladder (HCV29 & T24) and melanoma (WM115 & WM266-4) cells visualized by AFM topography imaging and fluorescent staining of actin filaments by phalloidin labelled with Alexa Fluor 488.
Task 7.2: Quantitative description of cancer cell parameters
Quantitative description of the expression of cell surface molecules (distribution, interaction forces, binding/unbinding kinetics)
The expression of N-cadherin has been studied in bladder and melanoma cells in search of differences both in an unbinding force and distribution on a cell surface. The measurements of the unbinding force as a function of loading rate, performed for melanoma cells, showed distinct unbinding kinetics. The obtained TREC images showed the N-cadherin distribution directly on a surface of melanoma cells.
These results were partially presented in February 2011 at the XIII Linz Winter Workshop as a poster presentation “Unbinding properties of N-cadherin in Human Bladder Cancer”.
Determination of stiffness of cancerous cells via AFM und 2nd harmonics imaging
The AFM measurements of cancer cell elasticity using AFM are completed. The analysis of the stiffness dependence of as a function of the indentation depth showed the decrease the of the Young’s modulus with the increase of the indentation depth. However, the relation between melanoma cells (WM115 & WM266-4) remained the similar. Independently of the indentation depth, the WM266-4 are always more deformable as compared to cells originating from primary tumour site.
The 2nd harmonics imaging was performed using AFM-ILM-OT platform. The results showed that distribution of storage and loss modulus is more homogenous in melanoma cells coming from metastasis to skin (WM266-4), what was conformed by the fluorescent microscopy showing less differentiated actin cytoskeleton in these cells.
Quantification of cytoskeleton organisation (relative Young’s modulus value)
To quantify the observed alterations in cell elastic properties, the 2D-distribution maps of the Young’s modulus are created based on the AFM elasticity measurements. Both in histograms and in 2D-maps two fractions of the Young’s modulus were observed i.e. softer one corresponding to places where AFM probe indents only cell membrane and more rigid fraction related to probing the actin filaments underlying beneath cell membrane. The fraction of stiffer filaments was larger for non-malignant cells (HCV29) as compared to malignant counterpartner (T24), and also for melanoma cells from primary tumour site (WM115). The results were verified by 2nd harmonics imaging performed using SMW platform.
Quantification of correlation between the Young’s modulus and cancer malignancy
The metastatic cells independently on the cell studied are characterized by higher deformability (lower Young’s modulus) independently on the indentation depth dependence.
Visualisation of cytoskeleton by fluorescence imaging
Visualization of cell cytoskeleton (actin filaments and microtubules) were performed using both AFM and fluorescence microscopy in order to find the correlation with the state of cells (non-metastatic or metastatic one). Four different cell lines, human bladder (non-malignant HCV29 & malignant T24) and melanoma (non-metastatic WM115 & metastatic WM266-4), have been measured in physiological conditions. The differences between actin filaments organization in their cytoskeletons have been observed using AFM, and verified through fluorescence microscopy by staining them with fluorescently labelled phalloidin.
Work package 8: Application of SMW to the immunological process of T cell activation
Work package leader: Peter Hinterdorfer (JKU)
Partners involved: JKU
Task 8.1: Establishment of test system for T cell activation using AFM-mediated stimulation
For the experiments of T cell activation, both the biological systems and the measurement instruments have been established. Jurkat T cell (human cell line, TIB-152) can be maintained using normal cell culture procedure. Immobilization of T cells have been tested with bare glass slides, epoxy functionalised glass slides, and poly-L-lysine coated glass slides. T cells can be gently adsorbed on bare glass surface. It is possible to approach the AFM cantilever onto the living Jurkat cell and measure the force-distance curves for 1000 times, which supports the feasibility of T cell activation experiment. The above experiments were tested with two AFM systems: Agilent AFM5500 and AFM6000. The X and Y translation systems worked very well. The activation of the T cells is monitored by the fluorescence measurement of the calcium response in the cell. The optical microscope more from TILL is equipped with Fura-2 filter set, which can continuously switch between 340 nm and 380 nm for excitation. The software Live Acquisition from TILL provides the option of the online calculation of the ratio of the fluorescence signals from the two channels which allows the quick and sensitive monitoring of the change of calcium concentration in the cells.
Task 8.2: Characterisation of the interaction of CD3 with Jurkat T cells
For AFM imaging and force-distance curve measurements, fixed Jurkat cells were prepared. The fixed cell sample was at first measured with contact-mode AFM imaging in PBS by a soft cantilever with a spring constant of 10 pN/nm. From the topography (Figure 8.1) we can see that the Jurkat cells have a diameter of about 10 µm and the cell surface is not too rough. Therefore, cantilever tip functionalised with OKT3 was used to measure the simultaneous topography and recognition (TREC) images (Figure 8.2) on the top of the fixed Jurkat cell. The upper panel shows the topography (left) and recognition (right) images measured before block. From the recognition image, we can see that there are some nano-domains of CD3 molecules on the Jurkat cell surface. To examine the specificity of the recognition events, block experiment was performed by injection of free anti-CD3 antibody into the measurement solution. The images after the block are shown in the lower panel of Figure 8.2. The topography is still similar to that before block. However, most of the recognition signal is reduced.
We also used the OKT3 functionalised cantilever tip to measure the force-distance curves on the Jurkat cell. Two typical curves with binding events are shown in Figure 8.3(A and B). From measurements of more than 100 curves, the statistic distribution of the unbinding force between the OKT3 tip and the Jurkat cell can be obtained (Figure 8.3C).
Figure 8.3: (A, B) Typical force-distance curves measured on Jurkat cell with the tip functionalised with OKT3. (C) Distribution of unbinding force between OKT3 tip and the Jurkat cell measured with scanning range of 1 µm and scanning time of 8 s/cycle.
The force-distance curves were also measured with different force loading rate by using different scanning time and scanning range. It was found that the unbinding force is proportional to the logarithm of the force loading rate.
Task 8.3: Quantification of T cell triggering and response
The experiment configuration for the activation of living Jurkat cells is shown in Figure 8.4. The cantilever tip was treated with APTES, conjugated with NHS-PEG-aldehyde or NHS-PEG-acetal crosslinker, and covalently linked with anti-CD3 antibody (OKT3). Living Jurkat cells were loaded with Fura-2 before the activation experiment. During the experiment, the functionalised cantilever tip was used to continuously measure the force distance curves on a randomly selected cell. Meanwhile, the fluorescence images were continuously recorded at two excitation wavelengths: 340 nm and 380 nm. The ratio of the fluorescence signal from the two channels was calculated, which can sensitively monitor the change of the calcium concentration in the cell. One example of the measurement was shown in Figure 8.5. In panel A, the functionalised cantilever tip was localised above a Jurkat cell. Before force-distance curve measurement, the calcium concentration in this cell is as low as most of other cells, and after the force-distance curve measurement was started for about 3.5 min, the calcium concentration in this cell became higher. The results in Fig. 8.5 clearly demonstrated the increase of calcium concentration in this cell.
As a control experiment, we performed the measurement with a cantilever tip functionalised with non-specific goat IgG and none of the cells could be activated.
In conclusion, the single molecule workstation has been successfully used for the experiment of T cell activation by functionalised cantilever tip. The Jurkat cells are much more active in HBSS with Ca++ & Mg++ and with 10% FCS at 37°C than in PBS at room temperature. The cantilever tip functionalised with non-specific goat IgG cannot activate the Jurkat cell, but the tips with anti-CD3 antibody can. The recognition imaging revealed that there are nano-domains of CD3 molecules on the Jurkat cell. The force-distance curve measurements at different force loading rate allowed the calculation of energy landscape of the interaction between the anti-CD3 antibody tip and the Jurkat cell.
Figure 8.5: (A) A cantilever tip functionalised with OKT3 was approached to a Jurkat cell loaded with Fura-2. (B) Before force-distance curve measurement, the calcium concentration in the cell under the cantilever tip is low as indicated by the ratio of fluorescence signals excited by 340 nm and 380 nm. (C) After the force-distance curve measurements were started for 228 s, the calcium concentration in the cell under the tip increased significantly, which showing the activation of the T cell. (D) The Fura ratio signal of the cell measured by the tip indicates that the cell was activated about 3.5 min after the start of the force-distance curve measurement, and that the increase of calcium concentration lasted for about 70 s.
Additional Task: Study of High-Density-Lipoprotein particles (HDL) and the transfer of lipids to cells by the optical tweezers (OT)
Details of this task have been reported in the 2nd periodic report and deliverable D6.3. In conclusion, the optical tweezing capability of the SMW setup was applied for studying lipid transfer from HDL particles to living cells. We developed three different functionalisation strategies. All of them are straight forward and easy to implement without the need for sophisticated surface chemistry. The protocols described here should be applicable to a large number of biomolecules for various OT experiments. We found specific binding of HDL-coated beads to living cells with SRB1 whereas cells without SRB1 yielded no binding events. The interaction force even surpassed the maximal trapping force of the OT setup. Despite the strong binding we found no specific transfer of DiI from the HDL-coated beads to the cell. With respect to the AFM transfer experiments, this observation indicates that force plays an important role in the transfer mechanism. OT has a lower force range than AFM. Furthermore, in AFM the force is concentrated in a sharp tip, which increases the effective force applied between HDL particles and the plasma membrane. Our observations here indicate the importance of further studying this process in more detail.
Potential Impact:
1.4.1. Potential impact
Understanding how cells function requires characterising not only the molecular constituents of the cell, as afforded by the family of “-omics” disciplines, but also their spatio-temporal interplay. The two microscopes combined in the results of the project approach this goal from different sides, from the outside with maximal spatial resolution (AFM), and from the inside of the cell, with maximal temporal resolution. The possibility to gain overlapping information combined with two different approaches to initiate and modulate cellular responses, mechanically from outside and by light from the inside, opens a whole new dimension for characterizing and eventually understanding cell function.
1.4.2. Dissemination activities and exploitation of results
Beginning at the end of the first period, but mainly in the second period, the results of the SMW project were made available to the public in a variety of dissemination demonstration activities, such as the participation in scientific and industrial conferences and the publication of project results in scientific journals. The dissemination activities of the project period are listed in detail in Template A1, but the main activities were the following:
Conferences
- 3rd International Workshop HoloMet 2010 (June 13-16) on Perspectives of Optical Imaging and Metrology in Balatonfüred, Hungary
- XIII Linz Winter Workshop 2011 (February 4-7) in Linz, Austria
- AFM Biomed 2011 (August 23-27) in Paris, France
Publications
- Duman M. et al., “Improved localization of cellular membrane receptors using combined fluorescence microscopy and simultaneous topography and recognition imaging”, Nanotechnology, 2010, No. 21(11):115504
- Lekka M. et al., “Characterization of N-cadherin unbinding properties in non-malignant (HCV29) and malignant (T24) bladder cells”, Journal of Molecular Recognition, 2011, 24(5), 833-42
- Pogoda K. et al., “AFM depth-sensing analysis of cytoskeleton organization in fibroblasts based on AFM data”, Eur. Biophys. J. (Oct 2011)
- Woerdemann M. et al., “Controlling ghost traps in holographic optical tweezers”, Optics Letters, 36, 3657-9 (2011)
2. Use and dissemination of foreground
The research work carried out in SMW project aimed at developing an integrated ILM-AFM-OT platform with possible PTNS extension. Such tool can have a major impact for the biology community and hence it was extremely important to plan the dissemination activities for propagating the project results to the largest number of potential users.
Since the topics of the SMW project were not especially suited for non-specialised general audience, we also made an effort to reach a wider public than the scientific community, through press releases, disclosing the main achievements in an understandable way for European citizens and through presentations given on conferences and workshops for a more industrial audience.
Synergy with education levels was also foreseen towards the participating universities. The universities have integrated some of their SMW activities and result in the curricula, and several PhD students have been involved in SMW research activities and became very familiar with the SMW setup when performing the biological experiments. TID will also provide a microscope setup to the BioImaging Zentrum of the Ludwig-Maximilians-Universität München, with which it is closely collaborating for many years now. This ensures the technology transfer from industry to research and will help to foster academic research and education in biological and biophysical subjects.
Goal of the project was to develop new features for both the AFM- and the light microscope part of the instrumentation. One should be able to use them not only together, but also separately. Thus a number of features, developed in the course of the project, have meanwhile been turned into product features and are available as individual instrument modules from Agilent (AFM related) and TILL Photonics (light microscopy). There are more to come after the conclusion of the project, and the commercial success will thus not only rely on the success of a combined instrumentation, but also on the success of the two individual system modules.
List of Websites:
www.till-id.com/smw.html
Contact:
Dr. Rainer Uhl
rainer_uhl@me.com
Light and atomic force microscopy are complementary techniques for the study of live cells. With light microscopy, the interior of living cell is accessible for observation via fluorescence emission, and when light microscopy is combined with optical tweezers, molecular interactions can be observed even inside an intact cell. In contrast, AFM provides resolution on the cell surface topography at the nanometre scale and allows for studying mechanical interactions of membrane components. With SMW, all conceivable functions of these complementary technologies have been integrated into a single and unified Single Molecule Workstation (SMW) concept. The microscopy hardware to be developed has been combined with a single, highly intuitive graphical user interface (GUI) integrating all functional modules of the system, exchanging and controlling measurement protocols in real time and in highly synchronised fashion. The application of this SMW instrument beyond this project will enable major advances in life science research by delivering new insights into the molecular biology of living cells and complex biomolecular processes, which are yet inaccessible.
The main objective of the project was to create a flexible, modular and user-friendly workstation combining the three most important microscopic techniques ultra-sensitive inverted light microscopy (ILM), atomic force microscopy (AFM) for studying cell topography and molecular recognition imaging, optical tweezers (OT) for the detection of ultra-low forces and additional photo-thermal nano-spectroscopy (PTNS) to allow for the chemical characterisation of cellular material. These tasks have been worked on by the SMEs and the industrial partner of the consortium. On a second level, the SMW instrument has been applied for the study of the correlation between structure and function of living cells, with exemplary applications in molecular immunology and basic cancer research as performed by the academic partners.
The project started in December 2008 and the second period ended in November 2011. After the significant advances in the first reporting period (design of the mechanical, optical and electronic concept and software for the combination of AFM and ILM, implementation of SLM technology, establishment of test systems for cancer detection and immunology) a prototype of a combi AFM-ILM instrument with integrated SLM was built in the second reporting period. Its functionality was intensively tested and the observed interferences and problems with light efficiency could be successfully solved. Furthermore, the PTNS technology has been established and tested on living cells for the characterisation of cancer characteristics. After the optimisation of the SMW setup, it has been validated by the academic partners in a variety of experiments. The application of the system to study cancer cell characteristics and immunological T cell activation both by AFM and fluorescence microscopy techniques yielded interesting results that prove the capability of the developed SMW.
Results will be published in due time on the project website www.till-id.com/smw.html where also contact details and information of the Coordinator and the consortium can be found.
Project Context and Objectives:
Light- and atomic force microscopy (AFM) are complementary technologies for studying live cells. While light microscopy renders the interior of a living cell accessible for observation, AFM permits nanometre resolution on cell surfaces and allows not only viewing the cell surface topography, but is also applicable to study mechanical interactions of single membrane components. Light microscopy combined with optical tweezers can detect molecular interactions even inside an intact cell. The aim of our project thus was to integrate all conceivable functions of these complementary technologies into a single unified “single molecule workstation” concept. A single, highly intuitive graphical user interface (GUI) that integrates all functional modules of the system should exchange and control measurement protocols in real time and in a highly synchronised fashion. To achieve these goals it was not sufficient to take two separate, already completed technology approaches and try to merge them. Instead, the consortium needed to design the methodologies with the aim of a joint functionality in mind.
In terms of the technological aspect, the main goal of the project was to develop an innovative single molecule workstation (SMW) combining the most important microscopic techniques into one workstation:
• Inverted light microscopy (ILM)
• Atomic force microscopy (AFM)
• Optical tweezers (OT)
• Photo-thermal nano-spectroscopy (PTNS)
Thus, in the first phase of the project, the objectives of the technological concept development mostly were in the foreground:
1. Electronic and mechanic integration of AFM and ILM including system specification, noise characterisation, interface definitions, electronic controller design and integration and software design and architecture.
2. Design of a unified and intuitive user interface including customer workflow analysis, streamlining sample handling, general controlling calibrations and intelligent control interface enabling ease of use of the setup.
3. Integration of spatial light modulator (SLM) for beam shaping, phase contrast enhancement and optical tweezers for dynamic phase shaping, shaping the spatial intensity pattern in the specimen plane, holographic optical tweezers and position detection system
With this integrated workstation, a new quality level in the study of the molecular biology of living cells and biomolecular processes can be reached, enabling major advances in life sciences.
The strategic objective of the project thus was to create a flexible, modular and user-friendly microscope, which achieves a performance far beyond the current state-of-the-art.
Consequently, after the technological concept development phase, the objectives of validating the integrated system capabilities by scientific applications (cancer research and molecular immunology) were relevant in the second phase:
1. Correlated measurements by simultaneous topography and optical imaging with synchronised high-resolution topography imaging and fluorescence imaging with single molecule sensitivity.
2. Enabling space- and time-correlated cellular stimulation and fluorescence imaging with controlled cellular stimulation by positioning functionalised AFM-tips and monitoring cellular responses in real time via single molecule fluorescence.
3. Simultaneous AFM and fluorescence spectral imaging with higher harmonics topography and elasticity imaging combined with multi-wavelength fluorescence excitation and detection, localised on-site force spectroscopy with full fluorescence spectral read-out and fluorescence-guided force spectroscopy.
Project Results:
Work package 1: SMW system specification refinement and requirement analysis
Work package leader: Ferry Kienberger (Agilent)
Partners involved: TID, TILL, Anasys, Arivis, Agilent, IFJ PAN, HEPAG, JKU
Summary
The compatibility of an Agilent AFM with the TILL Photonics fluorescence microscope (inverted light microscope, ILM) is investigated with respect to noise issues, stability measures, mechanical interface definition, and optical as well as electronic implementation. Performance measures and systems specifications for the integrated single molecule workstation are setup, based on requirements analyses from cancer biology and molecular immunology, substantiated by market studies and polls to life science researchers.
Task 1.1: Requirement analysis for mechanical and electronic components
For studying single molecule processes at the cellular level there are various requirements that researchers expect from the combination of AFM and ILM. For a combined setup of AFM and ILM, including also potential optical tweezers (OT) and photo-thermal nanospectroscopy (PTNS) functionality, we did a market survey asking various researchers (including the bio-application partners from the SMW consortium, but also external life-science researchers). Three classes of researchers were addressed: (cell) biologists, biophysicists, lab managers. Many of them use the AFM as stand-alone, not on the ILM. The combined AFM/ILM has the following requirements, divided into two categories: MUSTS: (i) noise very important for biophysicists (in the 10 pN range and below; for OT below few pN); (ii) environmental control and fluid exchange is important; (iii) operational robustness: I sit down, turn on the instrument, and get the data I want; (iv) Ergonomics: easy to work with; (v) Repeatability: get the same results time after time. WANTS: (i) Automation is good but experienced users, particularly biophysicists, want to have freedom for advanced experiments; (ii) Biologists do not like writing software code, but want simplicity.
While the electrical components from the stand-alone AFM and ILM meet corresponding specifications easily, the mechanical components were investigated in more details to cover the force noise sensitivity as requested from the researchers. In particular we applied iterative CAD-modelling and high precision mechanical fabrication to get an optimal AFM-ILM interface (cf. WP2).
Task 1.2: Requirement analysis regarding mechanical stability and noise sources
Particularly for cancer biology and molecular immunology, as well as other life-science research areas, the noise level is the most important parameter for the combined AFM/ILM, according to the market survey. Stand-alone AFM noise is typically below 10 pN (with OT noise typically below few pN), with molecular forces typically ranging between 10 pN and 300 pN. Accordingly, the mechanical interface definition of AFM and ILM needs to be well defined to realize proper stability measures and reasonable low force noise of the combined setup.
Below figure shows a selected CAD model (computer aided modeling) for the mechanical interface (the quickslide) of the Agilent 5500 AFM and the TILL Photonics iMIC. The design was guided by the requirement of lowest possible mechanical noise, with design and construction know-how from related Agilent Technologies products. Thereby the 5500 AFM head is mounted on the iMIC body via its three main screws on the quick-slide. The three AFM-head screws are responsible for the coarse z-approach of the AFM-cantilever to the sample, with the sample mounted on the integrated stage of the light microscope. This mode is called free standing operating mode FSOM. This design allows having full operational functionality of the AFM and the ILM in the combined setup.
Fig. 1.1: CAD model of the mechanical interface (quickslide) of the combined 5500 Agilent AFM and the iMIC from TILL Photonics. The quickslide (left panel) is mounted on the iMIC body (blue part), and the AFM is put on top of the quickslide (red part) in free standing mode operation FSMO. This allows for full operational freedom of the individual microscopes.
Task 1.3: Performance specifications for optical and mechanical components
The first prototype combines the Agilent 5500 AFM with the TILL Photonics iMIC (the ILM). The previous part showed the design of the mechanical interface (the quickslide), which allows simply adding and removing the AFM from the ILM, without any compromise on the performance of the individual microscopes. Similarly as for the mechanical integration, we designed electronic and optical integration without compromising the individual microscopes performances. The stand-alone electronics from the individual microscopes is used and termed as ‘AFM-Controller’ and ‘iMIC-Controller’. Both controllers interact properly with the mechanics, optics, and software interfaces, as they do in their regular stand-alone settings. Since the AFM is simply put on top of the ILM, and the optical elements of the ILM are below the integrated stage, there is no compromise on the optical performance in the ILM. The same is true for OT and PTNS functionality.
Work package 2: Interface definitions and software concept
Work package leader: Stefanie Graf (TID)
Partners involved: TID, TILL, Anasys, Arivis, Agilent, IFJ PAN, HEPAG, JKU
Summary
Based on the requirements specified in WP1, the mechanical interfaces between the AFM and the ILM were defined and mechanically built in a way to allow for the lowest possible interferences between the two microscope systems. Furthermore, a software concept has been designed to enable the appropriate communication between and the synchronisation of all components.
Task 2.1: Mechanical definition of ILM-AFM interface
In the first phase of the project, the technology partners for the light microscope and AFM (TID, TILL, Agilent) defined the mechanical interface of the combined ILM and AFM platform based on the specifications elaborated in WP1. This interface “quickslide”, cf. WP1 Fig. 2.1) was mechanically built and the AFM 5500 was mounted on the quickslide on top of the light microscope body (iMIC). All investigations on the compatibility of the Agilent AFM 5500 and the TILL iMIC regarding noise, stability and interface definitions brought satisfactory results, thus the first working prototype could be used for further technology integration.
In the second reporting period, the second working prototype of a combi ILM-AFM was built using the new TILL light microscope MORE™ and the Agilent AFM 6000. Consequently, for this platform, another interface had to be designed and constructed to connect the ILM and the AFM.
In a next step, a design for the electronic and optical integration was developed. Concerning the electronics, an AFM controller and an ILM controller will interact properly with the mechanics, optics and software interfaces as they do in their stand-alone settings. Regarding the optical concept, there is no compromise on the optical performance of the iMIC ILM, because the AFM is simply mounted on top of the ILM and the optical elements (optical paths) are below the integrated stage (cf. WP1). The same holds true for optical tweezers (OT) and PTNS functionality. Anasys defined the mechanical requirements for the integration of the PTNS sensor into the AFM stage and HEPAG defined the requirements for the opto-mechanical integration of the SLM in the light microscope platform.
Fig. 2.1: Mechanical interface (quickslide) for putting the AFM 5500 on top of the iMIC. Left: prototype of the quickslide, right: 5500 AFM (black) put onto the quickslide (gray).
Task 2.2: Development of software concept for user interface
TILL and Arivis - together with Agilent – have developed concepts for the appropriate communication between and the synchronisation of all components and their implementation into the software. Communication paths and data exchange routines between the two main parts of the SMW, the AFM (including PTNS) and the ILM (including OT), have been established. With this concept, calibration routines and workflows can be realised for proper experimental measurements. Moreover, workflows for fully automated measurements have been designed.
In another step, concepts for the combined graphical user interface (GUI) were defined and implemented (see WP3). The TILL Live Acquisition (LA) software and the Arivis Browser were adapted to meet the requirements of the combi AFM/ILM platform for proper fluorescence image data handling. They communicate with the AFM software Agilent PicoView and the AFM-related imaging data analysis software Agilent PicoImage (Fig. 2.2).
Fig. 2.2: The AFM software (Agilent PicoView Version 1.8) and the AFM-related imaging data analysis software (Agilent PicoImage) communicate with the Live Acqusition (LA) software of TILL Photonics and the arivis Browser.
Work package 3: Software and electronic design
Work package leader: Christian Götze (Arivis)
Partners involved: TILL, Arivis, Agilent, HEPAG
Summary
In order to access the full potential of the atomic force microscope (AFM) and inverted light microscope (ILM) combination, several electronic control parts and a software must be provided to access the AFM and ILM devices, control the single devices, correlate their coordinate systems and combine the resulting images. These tasks were worked on in WP3. As a result an integrated software system for control and analyze experiments had to be provided.
Task 3.1: AFM software design and electronic controller design
Agilent successfully completed the design of software required for the control of the AFM, based on Agilent’s AFM software PicoView and PicoImage, which provides the AFM elements within the GUI. This concept also includes the appropriate firmware and control electronics for the AFM part of the combi setup. The firmware of the AFM controller was designed to be able to generate TTL pulses whenever predefined events occur and the resulting TTL pulse is subsequently fed into the ILM controller, which will trigger the optical measurements. The synchronisation between AFM and ILM has to be very fast and reliable due to the ms time scale of biological reactions, which need to be followed with the setup.
Task 3.2: ILM software design and electronic controller design
TILL has completed the design of the software concept for controlling the ILM based on its Live Acquisition (LA) software. The new LA GUI concept provides access to all ILM elements and allows for setting up and executing real-time protocols. The firmware concept and the control electronics for the ILM have also been re-designed and the real-time capabilities of the Imaging Control Unit (ICU) could be successfully demonstrated. The ILM controller is used to operate the light microscope, camera and light source and is driven by the ILM driver software, which is also able to send commands to the AFM driver software via a communication module.
In addition, HEPAG provided software modules for OT and phase contrast measurements as well as a driver software for the SLM control unit to be integrated into the ILM platform and the appropriate elements for use in the combined GUI. HEPAG realized a real-time capable software platform for generating intensity patterns for photo-bleaching and photo-activation purposes.
Task 3.3: ILM and AFM software integration into combined user interface
For controlling both microscope devices, hardware connections between the instruments had to be developed. The electronic implementation bases on a TTL pulse protocol, generated by the AFM, triggering the ILM and controlling basic system tasks of joint optical experiments (see Figure 3.1).
Fig. 3.1: Schematic representation of device interfaces.
Basing on these hardware connections between the devices, a software framework was designed and implemented. The project partners preferred to extend and integrate the existing special software systems over re-implementing everything in one software in order to keep all the expert software features for the integrated system.
Thus, in the SMW context there are 4 different software systems for hardware control and image analysis: the TILL Photonics Live Acquisition software (TILL LA) for ILM control, the Agilent PicoView (PicoView) software for AFM control, the HOLOEYE software for controlling the spatial light modulator (SLM) and the arivis Browser software for image overlay, visualization and analysis. To integrate these applications they were connected as shown in Figure 3.2.
Fig. 3.2: High level scheme of software scenario.
For the scenarios specified in this project, the integrated system must be able to perform three different tasks:
• Calibrating the two devices to get a transformation between their coordinate systems,
• Positioning the AFM tip interactively and
• Acquiring ILM images and correlated AFM Regions of Interest (ROI).
All of these tasks have to be performed with two different AFM types – the Agilent 5500 AFM and the Agilent 6000 AFM. Both devices differ completely, in mechanical construction as well as in electronic interfaces. Thus, different calibration and operation modules had to be implemented.
Fig. 3.3: Experiment control UI in TILL LA software.
As a result of the software development the application partners could use a powerful AFM-ILM software for their experiments (refer to Figure 3 and 4). Each of the four integrated software packages has its clearly defined tasks in this setup:
• TILL LA controls the whole experiment. It performs the calibration procedure and is used as the control center to set ROIs and pass parameters and results. Furthermore, it controls all ILM parameters.
• Agilent PicoView is used to control more advanced parameters of the AFM. Moreover, AFM images can be pre-processed in this software.
• HOLOEYE SLM software controls the instruments Spatial Light Modulator
• Arivis Browser is the visualization and analysis tool. Combined image sets can be visualised in any combination, colour properties can be set and analysis procedures as segmentation can be applied.
Fig. 3.4: Combined software for controlling the integrated setup.
Conclusion
In this work package a stable, usable toolkit for everyday work with the combined setup was developed and implemented. The existing software systems were highly integrated and the typical workflows (calibration, point scan, ROI scan) were implemented for ease of use. As the evaluation work packages showed, electronic and software of the combined setup is highly usable, even for complex and sophisticated experiments.
Work package 4: Mechanical and optical integration
Work package leader: Sven Krüger (HEPAG)
Partners involved: TID, Agilent, HEPAG
Summary
Within the project time HOLOEYE has grown into an independent fabless SLM/microdisplay producer for photonic and industrial applications. That is, the application-oriented customization of the SLMs was enabled: cell gap, LC-Material, packaging, flex, interpixel gap, AR-coatings etc.
The implemented SLM was integrated into the MORE™ microscope platform of partner TILL Photonics. Furthermore real-time hardware-accelerated CGH calculations were developed (CUDA, Clearspeed), demonstrated and tested with biophotonic applications.
Optical tweezers with multiple traps and aberration correction, SI and FRAP were implemented and tested in the microscope platform.
Task 4.1-4.3: SLM mode for dynamic phase shaping, for shaping intensity patterns and for using dynamically altered structured illumination
Task 4.4: Optical tweezers functionality for force measurements
Introduction
The general goal in the project for HOLOEYE was the implementation of the liquid crystal on silicon (LCOS) spatial light modulator (SLM) technology for biophotonic applications. This comprised the development and production of the dedicated SLM and integration of the SLM into the commercial microscope platform.
The new SLM
The existing commercial LCOS phase SLMs are working mostly in a relatively limited spectral band. This does not fit commercial requirements for biophotonic applications, because the novel microscope products implement numerous of techniques and functionalities at different wavelength from blue to IR part of spectrum. The blue-green part of spectrum is well suitable for structural illumination, FRAP and TIRF techniques, which were implemented with the developed SLM within the project. The IR part was intended mostly for optical tweezers (OT) application. The red part of the spectrum was used for SLM-based phase contrast imaging.
Started with the single wavelength version of our phase-only SLM PLUTO NIR with HD resolution, which was compatible with the wavelength of 1064 nm only, we developed in the first reporting period and brought to the market a new phase-only broadband SLM PLUTO NIR-II, which was integrated into the MORE™ microscope platform of TILL Photonics. The SLM is specified for working in the 400-1064 nm spectral band with the full phase modulation over 2pi.
Further Customisation
Within the project time HOLOEYE has grown into an independent fabless SLM producer. This allowed further improvement of the SLM technology in the second reporting period, using application-oriented customisation, which comprises various options in addressing of the SLM and in design of the panel itself.
In the driving options a new technique was implemented, which decorrelates time-dependent flicker-noise, hence providing a significant reduction of this kind of noise. The fig. 4.1 show flicker noise with standard (left) and with improved (right) addressing (the amplitude mode was used for simplicity). As one can see, the flicker noise in this case is much lower (about 6 times) and the base frequency of the flicker is two times higher. The last fact is also advantageous, because in most cases higher-frequency flicker has lower affect to the performance of the system. The reduction of flicker noise, which is typical for all digitally addressed microdisplays, is important for certain applications, because it affects the performance of OTs, as well as the performance of SI/FRAP, especially when pulsed light source is needed.
The other approach for reduction of the flicker noise is the optimisation of the design of the panel itself. The flicker noise is dependent mostly on the addressing frequency, viscosity of the LC material used and the cell gap. That is why HOLOEYE produced additional versions of PLUTO NIR-II with different cell gaps in the range between 4 and 12 microns. The versions with different LC materials were also implemented.
To analyse phase temporal response in more general way, HOLOEYE implemented a high-speed interferometer, based on CCD-line camera, which allows measuring phase response of the panel at high frequencies (up to 32kHz).
The problem of the panel curvature also was addressed here. The curvature of the panel introduces aberrations into the system, which can be compensated with the panel itself. This software solution is based on the measurement of the wavefront from non-addressed panel (fig. 4.2 left). The addressing of the certain phase map compensates the curvature. The procedure can be made at low cost using a simple interferometric setup and the software, where phase map can be represented e.g. with Zernike polynoms. The interferogram of the compensated panel is shown at fig. 4.2 (right). One can see, that only the addressable area of the SLM is compensated. The real-time compensation can be achieved using developed Framework PhaseFrame (see below), where the static phase map will be subtracted from the holograms, addressed to the display.
Nevertheless a hardware solution is preferable here, so other approach was to improve design of the panel itself as well as the packaging process. This allows in general improving flatness of the panel (fig. 4.3). The performed analysis showed, that the thickness of the display coverglass affects the flatness of the panel. The thicker coverglass helps to keep panel flat, that is why the versions of NIR-II with the coverglass thickness of 1.1 mm were produced. Additional option was the usage of the ceramic stiffener instead of metall.
Further developments include improvements in light and diffraction efficiency of the SLM. Currently wafers with reduced inter-pixel gap (200 nm vs. 500 nm) are ordered and processed. The SLMs based on this wafers will be available in the near future.
Fig. 4.1: Flicker for standard (left) and for optimised (right) addressing.
CGH calculations towards biophotonic applications
During the second reporting period, HOLOEYE used part of resources to implement an advanced software package in order to help developers to realize own real-time algorithms for SI, FRAP and OT using GPU calculations.
Fig. 4.4: Screenshot of the sample program, written with the Framework: Real-time hologram calculation of the checker board pattern.
For advanced researchers HOLOEYE developed PhaseFrame Framework (Fig. 4.4) which enables calculation of holograms on the nVidia graphic card (GPU), using CUDA programming language. This Framework allows developers to implement quickly real-time FFT (for up to 3 SLMs simultaneously) using as input a still image, image sequence, video data or video stream from extern camera as well as data from frame buffer.
Integration of the SLM into the microscope platform
With the integration of SLM into the MORE™ microscope platform, the following techniques were enabled with the SLM: SI, SI TIRF, FRAP, OT.
OTs were demonstrated with multi-traps with help of the optical tweezers software developed by Institut für Technische Optik (ITO), Stuttgart. The software works in Labview environment and allows managing multiple optical traps real-time using GPU computations.
Fig. 4.5 shows two beads that are trapped in the MORE™ platform with the IR laser (825 nm). The size of each bead is 16,3 µm. During experiments also beads of other sizes were trapped (6µm, 10µm). The number of traps is limited to a large extent only with the power of the laser source.
Fig. 4.5: Two beads 16,3um are independently trapped with OT Software, image courtesy J. Madl JLU Linz.
At first, the quality of the trapping laser beam was suboptimal and we demonstrated the flexible aberration correction with SLM, which could improve the spot quality (fig. 4.6). The original non-corrected spot has a significant amount of astigmatism and coma (fig. 4.6 left). This was corrected with the optical tweezers software (fig. 4.6 right).
The realisation of the SI/FRAP/uncaging techniques is very similar with OTs, except of the laser used (480 nm). The calculation of the periodic pattern for SI is made usually with iterative Fourier transform algorithm (IFTA). At the fig. 4.7 one can see a reconstruction of the biological object, imaged with SI (colour represents the depth). For FRAP, the basic algorithms remain the same. Fig. 4.8 shows a FRAP image of the pattern with letters “BIZ”, that was generated at the sample plane onto a SIP chart. The SLM can be easily synchronized with the camera in order to make quantitative FRAP measurements. Structured illumination for TIRF, where SLM is located in the object-conjugated plane, requires block of the zero order. This can also be done with the SLM.
Task 4.5: CCD-based wide-field imaging with TIRF and focus hold system
Test experiments have been performed with the novel algorithm for structured illumination to further optimise image resolution and the signal-to-noise ratio. This new approach was favoured instead of the CCD-based imaging due to a better sectioning and image resolution.
The observed temperature variations in the ILM causing tiny changes in the output polarisation of the lasers and signal changes indistinguishable from changes of the focus position could be solved with an optimised TIRF and focus hold system. The focus sensor configuration has been also implemented into the acquisition software.
Work package 5: Hardware integration and proof of concept building
Work package leader: Rainer Daum (TILL)
Partners involved: TID, TILL, Agilent, HEPAG
Summary
In the course of the SMW project, two combi microscope setups have been mechanically built. The first workstation was made up of the TILL light microscope platform iMIC and the Agilent AFM 5500, which was fully functional and achieved satisfactory results in performance and stability tests. With this platform, simultaneous topography and recognition imaging was performed on biological samples. As soon as the novel light microscope platform MORE™ was available, the HOLOEYE SLM was implemented into this platform for multiple microscopic applications. This ILM/SLM platform was mechanically combined with the new Agilent AFM 6000 to a fully integrated, highly stable and versatile microscope platform.
Task 5.1: Implementation of SLM functionality into ILM-AFM setup
In a first step, TID and TILL have been working with HEPAG on the implementation of SLM functionalities into the light microscope platform. The SLM hardware provided by HEPAG was designed into the new TILL microscope platform and a first microscope prototype was built, in which the SLM as well as optical components for trans-illumination have been optically and mechanically implemented (Fig. 5.1). The SLM was equipped with the new “D”-type of phase modulator for a broad waveband (400-1100nm). This device has been characterized for use in OT (1064 nm) as well as in optical filter applications (630 nm).
Furthermore, the control electronics, a new DSP control-board, for the ILM were developed and the appropriate firmware for real-time control of all required ILM functions and their synchronisation with the SLM and AFM was created by TILL. HEPAG provided driver electronics for the SLM.
In order to use the same SLM both in epi-illumination mode (together with an IR-laser for optical tweezers, or with a visible laser as pattern generator for structured light or FRAP) and in transmitted light mode as filter element for contrast-enhanced viewing, a special trans-illumination mode had to be established, which strikes the sample at small angles only.
Fig. 5.1: 3D optical concept (left) and first mechanical prototype (right, turned by 90° to the left) of microscope platform with implemented SLM (upper left corner on the right image), two diode lasers and CCD camera
Agilent mechanically integrated full AFM functionalities on top of the ILM/SLM setup, including the possibility to run TREC measurements, force experiments and imaging (in contact and MAC mode). For the mechanical integration of AFM and ILM, the “normal” 4-objective-slider (movement in one direction) of the MORE™ platform had to be mechanically modified by TILL. The filter slide was equipped with the various filter sets (e.g. for Cy3- and Cy5-like dyes).
The combined ILM/AFM/SLM setup built of MORE™ light microscope and the Agilent AFM 6000 is shown in Fig. 5.2. The system was equipped with a 828 nm IR laser (for OT) and a TILL Oligochrome light source for fluorescence (with excitation filters for GFP, Cy3 and Cy5) and set up in an acoustic isolation chamber mounted on a massive stone plate.
Fig. 5.2: The SMW setup. The Agilent AFM 6000 is mounted on a TILL MORE™ stage. The HOLOEYE SLM and the 828nm laser are integrated underneath the stage (not visible here, see fig. 5.1). The setup is placed on an active dampening system. Everything is isolated acoustically.
With the SMW setup the HOLOEYE SLM application software can be used to Fourier-transform images and transmit this information to the SLM. As shown in the figure 5.3 that alignment is good enough to project images into the sample plane, at least with a 20x air objective.
Furthermore, the functionality of the AFM in the combi setup was intensively tested on fixed T24 cells and it worked properly using all required AFM modes (dynamic force imaging, force spectroscopy, TREC…. Combined dynamic imaging with both brightfield and fluorescence imaging was successfully performed using a 20x air objective (see fig. 5.4).
Fig. 5.4: Combined optical microscopy and dynamic AFM imaging on the SMW setup. The AFM scans were carried out at the marked section in the brightfield image. The AFM scans were performed in liquid.
Task 5.2: Noise level characterisation
For a fully functional ILM-AFM setup the noise level for any AFM measurement induced by the ILM is a critical point, which had to be characterised in detail. The ILM itself contains moving parts, such as filter cubes, objectives. It was clear from the beginning that there will be certain limitations for an AFM measurement in terms of what kind of movements or vibrations inside the ILM can be tolerated. For instance, moving the filter cube or changing the objective while the AFM tip is attached and performing a force-distance cycle in certainly impossible, but luckily this is not required by any real-world ILM-AFM applications.
Therefore the involved partners mainly focused on improving the mechanical interface of the ILM-AFM and the full integration of the voice-coil z-drive. Finally the damping characteristics of the ILM body (made of mineral cast) in combination whit the low-noise z-drive induced a noise level into the AFM measurements, which is sufficient to carry out force-distance cycle experiments while running a live z-stack (oil objective moves up & down while in contact with the sample cover slide). For most application the ILM objective can be set to a distinct focal position (no movement) enabling even AFM experiments requiring exceptional low noise levels. While the first test were still performed on the iMIC platform, more comprising data were recorded on the much more stable MORETM platform, which is of course the natural choice for a combined ILM-AFM setup.
Fig. 5.5: Simultaneous topography and recognition imaging; (a) signals coming from the maxima and minima of each sinusoidal cantilever oscillation are recorded as topography and recognition images in TREC. (b) A specific ligand, attached to an AFM tip via a flexible PEG tether, binds to a target molecule. (c) Simultaneous topography (left) and recognition (middle) images of CD1d-?-GalCer complexes using TCR modified AFM tips. Recognition sites are indicated by dark spots in the recognition image. The right panel shows the 3D topography superimposed with the distribution of receptor spots. The scan size was 1.1× 1.1 ?m2 with the height scale ranging from 0 to 0.28 ?m on the upper panel and the scan size was 3.2 × 3.2 ?m2 with the height scale ranging from 0 to 0.5 ?m on the bottom panel (from Duman et al., 2010).
A high signal-to-noise ratio is particularly important for simultaneous topography and recognition imaging (TREC) to explore cell shapes, membrane domains and an accurate localisation of receptor binding sites with a resolution of 5 nm. With the SMW setup, theses TREC measurements can be combined with fluorescence imaging to determine the overall expression level and the distribution of receptor sites on the cell surface. The signal-to-noise ratio achievable with the first prototype of the AFM-ILM-combi setup (AFM 5500/iMIC) was high with an apparent spot diameter of 25 nm. Results of TREC imaging experiments are shown in Fig. 5.5.
The noise levels were also measured for force-distance cycles in order to analyse the impact on force spectroscopy experiments and we achieved ~12pN peak-to-peak fluctuations in the measured force. At this range, biologically relevant receptor-ligand unbinding forces can be easily studied.
Work package 6: SMW test, validation and optimisation
Work package leader: David Grandy (Anasys)
Partners involved: TID, TILL, Anasys, Arivis, Agilent, IFJ PAN, HEPAG, JKU
Summary
This work package was the most important part of the SMW project, because it aimed at the testing and validation of the integrated SMW platform with all its functionalities (ILM, AFM, SLM/OT and PTNS) and the optimisation of noise levels and vibration, light efficiency, optical tweezers functionality and many more. Only after the technology partners have successfully finished the integration and test of the hardware and software of the SMW microscope, the application partners could start with the validation and the specification of additional requirements for their biological experiments.
Task 6.1: Verification and optimisation of ILM imaging performance
Before the integration of ILM and AFM, the optical imaging functionalities of the new ILM microscope from TILL have been tested with respect to high speed, low vibration and long-term performance with satisfactory results. In parallel the controlling and analysis software was also tested intensively. Due to some problems in the calibration process and imaging, the controlling software had to be partly redesigned (see also WP3). After re-implementation of the software, further tests were started, which showed a general improvement of the acquisition results. With the optimised ILM-AFM platform, a number of optical performance tests were performed and for instance, the measured point spread function is symmetrical and its dimensions fit to the expected value.
Task 6.2: Verification and optimisation of AFM performance
The combi AFM was tested using standard AFM imaging and force-distance cycle acquisition. High resolution topographical images of protein layers showed resolution of few nanometres while molecular recognition imaging showed single molecule detection sensitivity. After optimisation of the magnetic coil of the AFM, magnetic AC imaging MAC-mode showed proper performance in amplitude-frequency plots as well as phase-frequency plots.
As a biological test, JKU used the AFM 6000 system to study the distribution of GPI-GFP molecules on the T24 cell surface and obtained good correlation between the GFP fluorescence and the anti-GFP recognition signals. The accuracy of the overlay was proved by optical and topographical imaging of FITC beads on the same position. The specificity of the recognised sites was investigated by using amplitude block and by injecting free anti-GFP antibody molecules, both of which showing convincing results (Fig. 6.1).
Fig. 6.1: Overlay of optical and topographical images of FITC beads.
Further tests were performed (details in 2nd periodic report) and in conclusion, the combined AFM-optical setup has shown high accuracy for overlay of optical and AFM images. The good correlation between the fluorescence and the recognition images enhanced the reliability of both techniques and provided additional information of molecules on cell surface and within the cell.
Task 6.3: Test of PTNS functionality
Anasys decided to start with the characterisation of pharmaceutical drug morphology using PTNS as this is a sample system that we were very comfortable with. We hope to switch to looking at cells in the next phase. Shown below are some of the results obtained regarding characterisation of drug morphology using photothermal FTIR spectroscopy
Determining drug morphology is an important issue in the pharmaceutical industry. In this work the photothermal technique was used to investigate a model drug supplied by GSK. The PhotoThermal technique initially characterised tablets made purely of the drug in its amorphous and crystalline forms. Figure 6.2a shows the topographical image of each tablet surface whilst 6.2b shows the corresponding photothermal spectra. Whilst there are clear differences in the topography, it is not possible to state which surface represents each material. The spectrum of the amorphous material shows much broader peaks that that of the crystalline. This is a common observation with IR spectra of amorphous materials due to non-specific intramolecular interactions in this state. An investigation was then carried out into a mixed system of amorphous and crystalline materials. Figure 6.3a shows a topographical image with locations at which photothermal spectra (Figure 6.3b) were obtained. Point 1 shows a spectrum characteristic of a crystalline material whilst points 5 and 6 show that of an amorphous material. The remaining points, however, appear to show intermediate spectra. It is likely that these points are close to the interface between the two materials and that the thermal waves travelling through the materials will cause spectral mixing. Subsequent thermal analyses were carried out at these locations, using the same thermal probe, to distinguish between the materials, thus demonstrating the power of the technique to obtain information on the physical, chemical and thermal nature of the sample.
(a) Crystalline Amorphous (b)
Fig. 6.2: (a) AFM topography images (b) photothermal spectra of a crystalline and amorphous model drug.
Fig. 6.3: (a) AFM topography image showing locations at which photothermal spectra (b) were obtained.
Experimental results on cells
The procedure is described in detail in the 2nd periodic report. Four cell lines were prepared on glass slides before immobilisation on the probe. Once the particle was immobilised the tip was located at the focal point of the IR beam so that a spectrum could be measured (fig. 6.4). Spectra were acquired at a resolution of 8 cm-1 for a total of 1000 scans.
Fig. 6.4: Schematic digram of the process of immobilisation of cellular material on the tip of the probe.
The aim of this work was to compare cancerous cell lines with noncancerous lines. As mentioned previously there are four lines in total, two pairs each of homologous cell lines. Within each pair there was a cancerous and non cancerous cell line. To compare the spectra within each pair, the three individual spectra were averaged and overlaid (see detailed figures in 2nd periodic report and Fig. 6.5 below).
Concluding remarks
The results show that it is possible to obtain IR spectra of small amounts of cellular material with the PTMS technqiue. In accordance with previosuly reported work using conventional techniques, spectral signatures were observed for cancerous lines by comparing ratios of specific peaks. In these studies several measurements were made (at least ten) of each cell line to establish a true spectral signature. It was therefore agreed that many more repeat measurements would be carried out on fresh samples.
Task 6.4: Establishment of test system for cancer cell recognition
Two human bladder cell lines (HCV29 – non-malignant cancer of cell urether and T24 transitional cell cancer of bladder) were chosen as a model system used to test capability of AFM to identify cancerous cells. The studied bladder cells are different in many features - they have different morphology, actin filaments and microtubules organization, and N-cadherin expression.
The structure of cancerous T24 cells revealed quite different morphology as for example cell nucleus with irregular shape and cytoplasm less rich with organelles resulting in homogenous image of cell interior by TEM. The cell membrane is smoother, and there are fewer protrusions present on its sided. The observed differences in cell morphology of single cells were the criteria for choosing these cells as model system.
Task 6.5: Characterisation of chemical composition of cellular structures
The aim of this work was to compare cancerous cell lines with reference cell lines (non-malignant in case of bladder and less-differentiated cancer in case of melanoma). As mentioned previously there are four lines in total, two pairs each of homologous cell lines. Within each pair there was a cancerous and reference cell line.
a) b)
Fig. 6.5: a) The average IR spectra for studied cell lines: HCV29 – non-malignant transitional epithelium of urether (blue); T24 – transitional cell cancer of bladder (red); WM115 – melanoma cells from primary tumor site on skin (green), and WM266-4 – melanoma cells from metastasis to skin (magenta). b) Raw data with the fits. The ratio between peaks centred at 1035 cm-1 and 1087 cm-1 corresponds to the level of glycogen consumption.
Test measurements with Raman spectroscopy have been carried out in order to define the differences between the studied cells – human bladder (HCV29 and T24) and melanoma (WM115 & WM266-4) cell lines. The obtained results showed major difference within the range 1000-2000 cm-1 and 500-1400 cm-1 for bladder and melanoma cells. Afterwards, the IR spectra for the four studied cell lines were measured and the corresponding pairs of cells, i.e. HCV29 with T24 and WM115 with WM266-4, were compared.
To compare the spectra within each pair, the three individual spectra were averaged and overlaid (Fig. 6.5). Based on this data, the spectral signature of cancer progression was calculated. The R-value for non-malignant cells varies from 0.6 to 0.8. The R = 753 obtained for non-malignant HCV29 cells lies within the limits of this range. For malignant cells, the R-value should be equal to or smaller than 0.5 which corresponds well the obtained values for T24, WM115 and WM266-4 cells.
The results show that it is possible to obtain IR spectra of small amounts of cellular material with the PTMS technqiue. In accordance with previosuly reported work using conventional techniques, spectral signatures were observed for cancerous lines by comparing ratios of specific peaks.
Task 6.6: Test of trapping capabilities and force calibration system
After optimisation of the poor light efficiency of the OT system (cf. WP4 and WP5) and substantial increase of laser power, the trapping capabilities of the setup were tested with beads of different sizes and the results showed that the number of traps is limited only by the power of laser source.
Task 6.7: Test of AFM functionalities on living cells
The goal of this task was to established AFM functionalities by (1) characterisation of the cellular stiffness accompanied by imaging the cytoskeleton using optical microscope; (2) establishing the model for single molecule interactions; (3) elaboration of sample preparation; and (4) establishing test system for cancer recognition.
(1) Characterisation of the cellular stiffness (AFM) and actin filaments imaging (fluorescence microscope)
It is known that differences between non-malignant and malignant cells are larger than between normal and non-malignant cells. Therefore, the two human bladder cell lines were chosen as a model, reference system. They were non-malignant transitional epithelial cell of urether (HCV29) and transitional cell carcinoma of urine bladder (T24). These cells are characterised by distinct shape and distribution of actin filaments, enabling the correlation between the fluorescent images and surface topography measured by AFM (Fig. 6.6).
Fig. 6.6: (left) Morphology of human bladder (HCV29 & T24) cells visualised by AFM topography imaging and fluorescent staining of actin filaments by phalloidin labelled with Alexa Fluor 488. These changes were accompanied by distinct elastic properties (right).
(2) Model for single molecule interactions in cancerous cells
The measurement of N-cadherin expression in T24 and HCV29 cells using Western blot (left) and fluorescence microscopy (upper right corner) showed its various levels depending on the type of the studied cells (Fig. 6.7).
Fig. 6.7: Measurement of N-cadherin expression in T24 and HCV29 cells using Western blot (left) and fluorescence microscopy (upper right) and measurement of unbinding force using AFM (down right).
The larger N-cadherin expression was detected in cancerous T24 cells as compared to HCV29 ones. Both the abundance and unbinding force was determined delivering information on larger amount of N-cadherins and on larger unbinding force (above 2 fold) pointing out the formation of more stable complexes in cancerous cells. With this data, a model system basing on elasticity measurements and N-cadherin expression in human bladder cells has been developed.
(3) Sample preparation for AFM- based recognition measurements
Within this task, the detailed growth conditions and steps for sample preparation were intensively tested and defined both for the separate growth of cells and for the simultaneous growth of two cell lines. They are described in detail in the 2nd periodic report.
(4) Test system for cancer recognition
The observed distinct elastic properties of bladder cells together with morphological differences were the basis for cancer cell recognition studies. Each cell line was cultured both separately and as a mixture of two corresponding cell types i.e. HCV29 together with T24 cells.
Once a mixture of cells was prepared, the AFM-based elasticity measurements have been performed on unstained cells. For each single cell, the Young’s modulus was calculated showing the dependence of the Young’s modulus versus cell number ordered as a function of time. If there is no deviation from the mean, the obtained Young’s modulus is not affected by other external factors influencing the cell elasticity. The average elasticity of non-malignant HCV29 cells was 4-fold larger than that of cells from malignant cell cancer of bladder. The mixed cells showed two modulus values that overlapped with those obtained for each cell line separately. The obtained values of the elastic modulus were in good agreement with those measured earlier, so with this data a reliable test system for cancer recognition could be established.
Work package 7: Developing the SMW towards an accurate tool for cancer detection
Work package leader: Malgorzata Lekka (IFJ PAN)
Partners involved: IFJ PAN
Task 7.1: Cancer recognition using the SMW platform
The goal of this task was the characterisation of morphological properties of cancerous cells alone using both optical and fluorescence imaging. The cell morphology of studied cell lines (melanoma and bladder cells) has been characterized using optical and fluorescence microscopy. Two types of samples were studied. Those containing only one cell type (either non-malignant bladder HCV29 or malignant bladder T24 or not-metastatic melanoma WM115 or metastatic melanoma WM266-4 cells) or two corresponding cell types mixed (i.e. pairs of two bladder or two melanoma cell lines). Imaging was carried out as a function of time of co-existence of two cell types. Cells were visualised through staining of actin filaments using phalloidin labelled fluorescently with Alexa-Fluor 488. To verify the growth conditions the doubling time was determined for cells grown separately and in co-cultures.
A) B)
Fig. 7. 1: Morphology of human (A) bladder (HCV29 & T24) and melanoma (WM115 & WM266-4) cells visualized by AFM topography imaging and fluorescent staining of actin filaments by phalloidin labelled with Alexa Fluor 488.
Task 7.2: Quantitative description of cancer cell parameters
Quantitative description of the expression of cell surface molecules (distribution, interaction forces, binding/unbinding kinetics)
The expression of N-cadherin has been studied in bladder and melanoma cells in search of differences both in an unbinding force and distribution on a cell surface. The measurements of the unbinding force as a function of loading rate, performed for melanoma cells, showed distinct unbinding kinetics. The obtained TREC images showed the N-cadherin distribution directly on a surface of melanoma cells.
These results were partially presented in February 2011 at the XIII Linz Winter Workshop as a poster presentation “Unbinding properties of N-cadherin in Human Bladder Cancer”.
Determination of stiffness of cancerous cells via AFM und 2nd harmonics imaging
The AFM measurements of cancer cell elasticity using AFM are completed. The analysis of the stiffness dependence of as a function of the indentation depth showed the decrease the of the Young’s modulus with the increase of the indentation depth. However, the relation between melanoma cells (WM115 & WM266-4) remained the similar. Independently of the indentation depth, the WM266-4 are always more deformable as compared to cells originating from primary tumour site.
The 2nd harmonics imaging was performed using AFM-ILM-OT platform. The results showed that distribution of storage and loss modulus is more homogenous in melanoma cells coming from metastasis to skin (WM266-4), what was conformed by the fluorescent microscopy showing less differentiated actin cytoskeleton in these cells.
Quantification of cytoskeleton organisation (relative Young’s modulus value)
To quantify the observed alterations in cell elastic properties, the 2D-distribution maps of the Young’s modulus are created based on the AFM elasticity measurements. Both in histograms and in 2D-maps two fractions of the Young’s modulus were observed i.e. softer one corresponding to places where AFM probe indents only cell membrane and more rigid fraction related to probing the actin filaments underlying beneath cell membrane. The fraction of stiffer filaments was larger for non-malignant cells (HCV29) as compared to malignant counterpartner (T24), and also for melanoma cells from primary tumour site (WM115). The results were verified by 2nd harmonics imaging performed using SMW platform.
Quantification of correlation between the Young’s modulus and cancer malignancy
The metastatic cells independently on the cell studied are characterized by higher deformability (lower Young’s modulus) independently on the indentation depth dependence.
Visualisation of cytoskeleton by fluorescence imaging
Visualization of cell cytoskeleton (actin filaments and microtubules) were performed using both AFM and fluorescence microscopy in order to find the correlation with the state of cells (non-metastatic or metastatic one). Four different cell lines, human bladder (non-malignant HCV29 & malignant T24) and melanoma (non-metastatic WM115 & metastatic WM266-4), have been measured in physiological conditions. The differences between actin filaments organization in their cytoskeletons have been observed using AFM, and verified through fluorescence microscopy by staining them with fluorescently labelled phalloidin.
Work package 8: Application of SMW to the immunological process of T cell activation
Work package leader: Peter Hinterdorfer (JKU)
Partners involved: JKU
Task 8.1: Establishment of test system for T cell activation using AFM-mediated stimulation
For the experiments of T cell activation, both the biological systems and the measurement instruments have been established. Jurkat T cell (human cell line, TIB-152) can be maintained using normal cell culture procedure. Immobilization of T cells have been tested with bare glass slides, epoxy functionalised glass slides, and poly-L-lysine coated glass slides. T cells can be gently adsorbed on bare glass surface. It is possible to approach the AFM cantilever onto the living Jurkat cell and measure the force-distance curves for 1000 times, which supports the feasibility of T cell activation experiment. The above experiments were tested with two AFM systems: Agilent AFM5500 and AFM6000. The X and Y translation systems worked very well. The activation of the T cells is monitored by the fluorescence measurement of the calcium response in the cell. The optical microscope more from TILL is equipped with Fura-2 filter set, which can continuously switch between 340 nm and 380 nm for excitation. The software Live Acquisition from TILL provides the option of the online calculation of the ratio of the fluorescence signals from the two channels which allows the quick and sensitive monitoring of the change of calcium concentration in the cells.
Task 8.2: Characterisation of the interaction of CD3 with Jurkat T cells
For AFM imaging and force-distance curve measurements, fixed Jurkat cells were prepared. The fixed cell sample was at first measured with contact-mode AFM imaging in PBS by a soft cantilever with a spring constant of 10 pN/nm. From the topography (Figure 8.1) we can see that the Jurkat cells have a diameter of about 10 µm and the cell surface is not too rough. Therefore, cantilever tip functionalised with OKT3 was used to measure the simultaneous topography and recognition (TREC) images (Figure 8.2) on the top of the fixed Jurkat cell. The upper panel shows the topography (left) and recognition (right) images measured before block. From the recognition image, we can see that there are some nano-domains of CD3 molecules on the Jurkat cell surface. To examine the specificity of the recognition events, block experiment was performed by injection of free anti-CD3 antibody into the measurement solution. The images after the block are shown in the lower panel of Figure 8.2. The topography is still similar to that before block. However, most of the recognition signal is reduced.
We also used the OKT3 functionalised cantilever tip to measure the force-distance curves on the Jurkat cell. Two typical curves with binding events are shown in Figure 8.3(A and B). From measurements of more than 100 curves, the statistic distribution of the unbinding force between the OKT3 tip and the Jurkat cell can be obtained (Figure 8.3C).
Figure 8.3: (A, B) Typical force-distance curves measured on Jurkat cell with the tip functionalised with OKT3. (C) Distribution of unbinding force between OKT3 tip and the Jurkat cell measured with scanning range of 1 µm and scanning time of 8 s/cycle.
The force-distance curves were also measured with different force loading rate by using different scanning time and scanning range. It was found that the unbinding force is proportional to the logarithm of the force loading rate.
Task 8.3: Quantification of T cell triggering and response
The experiment configuration for the activation of living Jurkat cells is shown in Figure 8.4. The cantilever tip was treated with APTES, conjugated with NHS-PEG-aldehyde or NHS-PEG-acetal crosslinker, and covalently linked with anti-CD3 antibody (OKT3). Living Jurkat cells were loaded with Fura-2 before the activation experiment. During the experiment, the functionalised cantilever tip was used to continuously measure the force distance curves on a randomly selected cell. Meanwhile, the fluorescence images were continuously recorded at two excitation wavelengths: 340 nm and 380 nm. The ratio of the fluorescence signal from the two channels was calculated, which can sensitively monitor the change of the calcium concentration in the cell. One example of the measurement was shown in Figure 8.5. In panel A, the functionalised cantilever tip was localised above a Jurkat cell. Before force-distance curve measurement, the calcium concentration in this cell is as low as most of other cells, and after the force-distance curve measurement was started for about 3.5 min, the calcium concentration in this cell became higher. The results in Fig. 8.5 clearly demonstrated the increase of calcium concentration in this cell.
As a control experiment, we performed the measurement with a cantilever tip functionalised with non-specific goat IgG and none of the cells could be activated.
In conclusion, the single molecule workstation has been successfully used for the experiment of T cell activation by functionalised cantilever tip. The Jurkat cells are much more active in HBSS with Ca++ & Mg++ and with 10% FCS at 37°C than in PBS at room temperature. The cantilever tip functionalised with non-specific goat IgG cannot activate the Jurkat cell, but the tips with anti-CD3 antibody can. The recognition imaging revealed that there are nano-domains of CD3 molecules on the Jurkat cell. The force-distance curve measurements at different force loading rate allowed the calculation of energy landscape of the interaction between the anti-CD3 antibody tip and the Jurkat cell.
Figure 8.5: (A) A cantilever tip functionalised with OKT3 was approached to a Jurkat cell loaded with Fura-2. (B) Before force-distance curve measurement, the calcium concentration in the cell under the cantilever tip is low as indicated by the ratio of fluorescence signals excited by 340 nm and 380 nm. (C) After the force-distance curve measurements were started for 228 s, the calcium concentration in the cell under the tip increased significantly, which showing the activation of the T cell. (D) The Fura ratio signal of the cell measured by the tip indicates that the cell was activated about 3.5 min after the start of the force-distance curve measurement, and that the increase of calcium concentration lasted for about 70 s.
Additional Task: Study of High-Density-Lipoprotein particles (HDL) and the transfer of lipids to cells by the optical tweezers (OT)
Details of this task have been reported in the 2nd periodic report and deliverable D6.3. In conclusion, the optical tweezing capability of the SMW setup was applied for studying lipid transfer from HDL particles to living cells. We developed three different functionalisation strategies. All of them are straight forward and easy to implement without the need for sophisticated surface chemistry. The protocols described here should be applicable to a large number of biomolecules for various OT experiments. We found specific binding of HDL-coated beads to living cells with SRB1 whereas cells without SRB1 yielded no binding events. The interaction force even surpassed the maximal trapping force of the OT setup. Despite the strong binding we found no specific transfer of DiI from the HDL-coated beads to the cell. With respect to the AFM transfer experiments, this observation indicates that force plays an important role in the transfer mechanism. OT has a lower force range than AFM. Furthermore, in AFM the force is concentrated in a sharp tip, which increases the effective force applied between HDL particles and the plasma membrane. Our observations here indicate the importance of further studying this process in more detail.
Potential Impact:
1.4.1. Potential impact
Understanding how cells function requires characterising not only the molecular constituents of the cell, as afforded by the family of “-omics” disciplines, but also their spatio-temporal interplay. The two microscopes combined in the results of the project approach this goal from different sides, from the outside with maximal spatial resolution (AFM), and from the inside of the cell, with maximal temporal resolution. The possibility to gain overlapping information combined with two different approaches to initiate and modulate cellular responses, mechanically from outside and by light from the inside, opens a whole new dimension for characterizing and eventually understanding cell function.
1.4.2. Dissemination activities and exploitation of results
Beginning at the end of the first period, but mainly in the second period, the results of the SMW project were made available to the public in a variety of dissemination demonstration activities, such as the participation in scientific and industrial conferences and the publication of project results in scientific journals. The dissemination activities of the project period are listed in detail in Template A1, but the main activities were the following:
Conferences
- 3rd International Workshop HoloMet 2010 (June 13-16) on Perspectives of Optical Imaging and Metrology in Balatonfüred, Hungary
- XIII Linz Winter Workshop 2011 (February 4-7) in Linz, Austria
- AFM Biomed 2011 (August 23-27) in Paris, France
Publications
- Duman M. et al., “Improved localization of cellular membrane receptors using combined fluorescence microscopy and simultaneous topography and recognition imaging”, Nanotechnology, 2010, No. 21(11):115504
- Lekka M. et al., “Characterization of N-cadherin unbinding properties in non-malignant (HCV29) and malignant (T24) bladder cells”, Journal of Molecular Recognition, 2011, 24(5), 833-42
- Pogoda K. et al., “AFM depth-sensing analysis of cytoskeleton organization in fibroblasts based on AFM data”, Eur. Biophys. J. (Oct 2011)
- Woerdemann M. et al., “Controlling ghost traps in holographic optical tweezers”, Optics Letters, 36, 3657-9 (2011)
2. Use and dissemination of foreground
The research work carried out in SMW project aimed at developing an integrated ILM-AFM-OT platform with possible PTNS extension. Such tool can have a major impact for the biology community and hence it was extremely important to plan the dissemination activities for propagating the project results to the largest number of potential users.
Since the topics of the SMW project were not especially suited for non-specialised general audience, we also made an effort to reach a wider public than the scientific community, through press releases, disclosing the main achievements in an understandable way for European citizens and through presentations given on conferences and workshops for a more industrial audience.
Synergy with education levels was also foreseen towards the participating universities. The universities have integrated some of their SMW activities and result in the curricula, and several PhD students have been involved in SMW research activities and became very familiar with the SMW setup when performing the biological experiments. TID will also provide a microscope setup to the BioImaging Zentrum of the Ludwig-Maximilians-Universität München, with which it is closely collaborating for many years now. This ensures the technology transfer from industry to research and will help to foster academic research and education in biological and biophysical subjects.
Goal of the project was to develop new features for both the AFM- and the light microscope part of the instrumentation. One should be able to use them not only together, but also separately. Thus a number of features, developed in the course of the project, have meanwhile been turned into product features and are available as individual instrument modules from Agilent (AFM related) and TILL Photonics (light microscopy). There are more to come after the conclusion of the project, and the commercial success will thus not only rely on the success of a combined instrumentation, but also on the success of the two individual system modules.
List of Websites:
www.till-id.com/smw.html
Contact:
Dr. Rainer Uhl
rainer_uhl@me.com
