My PhD



Project management
 Each step brings you closer to your goal, but each step is already a victory.
"Protein and cell adsorption: topographical dependency and adlayer viscoelastic properties determined with oscillation amplitude of quartz resonator"
Remark: for lisibility, all references have been omitted.


Presentation of the work

The understanding of protein and cell adsorption is of high importance for the production of biocompatible implants as well as for the biosensor field, where a protein layer has to be designed as a functional coating. It is well-known that a living body brought into contact with a surface will induce protein adsorption, on which further proteins or cells will adsorb. The foreign body reaction is composed of a succession of events. The first protein adsorption occurs within minutes. During the first day cells (neutrophils and macrophages) will attack the foreign body. When the macrophages find they cannot digest the implant, they fuse into giant cells to engulf the object in the following next 4 days. However the implant is too large to be completely ingest, therefore inducing cytokines (messengers) secretion to alert other cells. In response fibroblasts arrive and begin synthesizing collagen during the next 10 days. After 3 weeks the implant is completely encased in an acellular and avascular collagen bag. The properties of the first protein adlayer can therefore play a crucial role in the integration of implant in the body.

Depending on the application, varying adhesion qualities are required. For example bones or tooth implants demand good protein adhesion to allow the growth of the bone and the formation of a stable and strong interphase between the implant and the bone. In contrast the protein adhesion has to be poor for catheters or contact lenses, where a low contact with the tissue is required. The interactions between proteins and a solid surface can be influenced by modifying the surface properties, which are the chemistry, the surface charge and the topography. A number of studies have shown that cells "sense" the topography of the surface at the micrometer scale, growing along grooves of determined width and depth.

Because proteins are always present between the cells and the substratum, the first part of my work (see Paper I and II) has been devoted to study the role of topography on protein adsorption, which had never been addressed as only parameter. Created structures had to be scaled down to reach sizes accordingly to proteins, i.e. in the nm range, which is smaller than the lithography resolution. The Scanning Probe Microscopy (SPM) permits manipulation of molecules, measuring surfaces at the atomic scale or modifying the surface at the nm scale. An Atomic Force Microscope (AFM) has been used to create nanostructures applying the Local Anodic Oxidation (LAO), which allows us to change the topography (figure 2) without modifying the surface chemistry.
Figure 2: Nanostructures created by LAO.
A: The first two letters of my university FR.
B: Wire netting structure

On the created nanostructures proteins have been adsorbed, and the new surface can be measured by AFM. These experiments were static giving no information on the kinetics of protein adsorption.

The second part of my work (see Paper III to V) has therefore been focussed on the building of a Quartz Crystal Microbalance (QCM), which allows us firstly to follow the adsorption kinetics, and secondly to get information about the viscoelastic properties of the adlayer. In comparison to commercial QCM a new parameter has been introduced, which is the maximal oscillation amplitude of the quartz crystal (see Paper III). The adsorption kinetics of different systems has bee measured. Protein A-BSA-IgG is a system well-known in immunology and fibronectin is viscoelastic in solution. For both systems the mechanical properties of the adsorbed layers have been investigated on Ti and Au surfaces (see Paper IV). Cell adsorption and spreading has also been performed with different cell types, where the cytoskeleton has been disturbed with drugs (see Paper V). This system range has allowed us to improve the interpretation of QCM measurements, opening new applications in biotechnology research.


The Quartz Crystal Microbalance (QCM) is a powerful technique to follow the adsorption kinetics of quite different systems, such as atoms, molecules, proteins or even living cells. The physical phenomenon of the sensor is based on the piezoelectric effect, whose discovery is attributed to Pierre and Jacques Curie. In 1880 they showed that crystals of Rochelle salt could produce electricity when pressure is applied in certain crystallographic directions. One year later they showed the reverse effect, which consists of the production of strain by the application of electricity. Nevertheless it is only at the end of the 1950s that this phenomenon has been investigated. AT-cut quartz crystal is the most used cut for frequency control applications, because it has nearly zero frequency drift with temperature.

Two electrodes are evaporated on each side of the thin quartz plate for the application of an alternative voltage, which induces oscillation of the crystal. In 1959 Sauerbrey published an article showing that the frequency shift of quartz crystals is proportional to the adsorbed mass under vacuum. The QCM was born.

In 1981 the application field of QCM is enlarged to liquid environments, opening new research insights in biotechnologies and electrochemistry.

The general setup of our QCM is presented in figure 8.
Figure 8: Setup of our home-made Quartz Crystal Microbalance.

It is composed of three different main components, which are the sensor environment, the QCM-computer interface and the computer with the program allowing to acquire, treat and present the data.

  The sensor environment top

10MHz quartz crystals have been used, whose electrodes show two different diameters. Only the larger one enters in contact with liquids, in order to avoid perturbating effects due to the electrical properties of the liquid. The quartz crystal is entrapped between two plexiglas pieces, which are sealed with VITON O-rings. The liquid exchanges are performed manually with micropipettes and a tubing system.

The sensor and the solutions are placed in a temperature control box in order to maintain a constant temperature of T ±0.1oC by water circulating inside its walls. No artefacts due to thermal modification of liquid properties can therefore occur.

It is possible to combine the QCM with microscopy, by mounting the crystal under an optical or fluorescence microscope.

  The QCM-computer interface top

The electronic interface controls the relay and digitizes, stores and sends the measured data to the computer.

A frequency generator excites the crystal near its series resonance frequency f and a second frequency generator determines the sampling frequency fsampling. The relation between the measured alias frequency and the quartz resonance frequency is given by the following relation:
falias=fsampling ·frac (  f


When fsampling » f, Delta fsampling = Delta f, so that measured shifts of the alias frequency are proportional to resonance frequency shifts.

The relay is closed about 3ms to excite the crystal, and is thereafter opened. Due to energy losses occurring in the system the crystal amplitude A begins to decrease according to
A(t) = A0 e-t/tau sin(2 pi f t + phi)

where tau is the decay time constant, f the measured frequency and phi the phase angle. One measurement under liquid lasts about 1.5ms. During the measurement the amplitude signal is digitized and stored in the RAM. It is thereafter transmitted to the computer with the GPIB bus, and a new measurement can be performed by closing the relay.

  The program top

A program has been developed in LabView in order to acquire, treat and show the data over hours. The measured amplitude is transmitted to the computer via the GPIB bus, and the program fits the data in order to determine f, tau and A0 each 300ms. The Dissipation factor D=2/(w tau) is also calculated.

Figure 10 (see pdf file) shows the developed Graphical User Interface (GUI). It is composed of three different parts.

In part A the acquired data are monitored. First of all the amplitude decrease of each measurement is shown in the left part. It allows us to adjust precisely the frequency generator at the resonance frequency of the crystal and to control continuously the crystal behavior. The decay time constant is determined and the exponential decrease is also shown in this first graph. The amplitude decrease is transformed in a pure sinusoidal curve in order to determine the resonance frequency. This signal is presented under the amplitude decrease. For each measurement f, D and A0 are plotted in two different ways. The general overview of the adsorption kinetics is visualized in the three small graphs (middle of part A). The parameters of the last 60 measurements are shown in the three larger graphs in the right part.

The input parameters such as the user profile, the measurement time, the acquisition rate and the created file name have to be given in part B.

Finally the part C allows us to control the Start/Stop of the measurements. The user can choose to start/stop the data acquisition directly or at a scheduled date and time.

Similarly the measurement can be immediately stopped, or its end can be determined.

  The resolution top

Under vacuum we obtain an excellent resolution of f ±0.03Hz, tau±1.5E-8sec, A ±3a.u. and D ±0.5E-7. The measured mass sensitivity corresponds therefore to m ±0.135ng/Hz cm2 using the Sauerbrey relation.

Under liquid loading the resolution decreases due to damping of the liquid. When nanodeionized water is in contact with the quartz at 25oC, we reach a resolution of f ±2Hz, tau±0.5E-6sec, A ±50 a.u. and D ±1E-6. The mass sensitivity is m ±9ng/Hz cm2, corresponding to the better resolution which is possible to obtain with 10MHz quartz crystals.

Protein adsorption on topographically nanostructured titanium    
C. Galli, M. Collaud Coen, R. Hauert, V.L. Katanaev, M.P. Wymann, P. Gröning and L. Schlapbach
Surface Science 474 (2001) L180-184

We study the protein adsorption on surfaces in order to investigate their predominant role in biocompatibility. Nanostructures are created by local anodic oxidation on titanium using the atomic force microscope. A remarkable specificity of the actin filament adsorption on the nanostructure height is noticed. F-actin is observed to have a low adsorption on nanostructures of a height of 4 nm and the adsorbed proteins appear to be randomly oriented. In contrast high protein adsorption is observed for structure height between 1 and 2 nm, moreover the filaments adsorb preferentially parallel to the nanostructured pattern.

Keywords: Atomic force microscopy; Surface structure, morphology, roughness, and topography; Chemisorption; Titanium oxide; Biological molecules-proteins

Creation of nanostructures to study the topographical dependency of protein adsorption    
C. Galli, M. Collaud Coen, R. Hauert, V.L. Katanaev, P. Groening, L. Schlapbach
Colloids and Surfaces B: Biointerfaces 26 (2002) 255–267

Nanostructures of sizes comparable to protein dimensions are created on Si and Ti surfaces by local anodic oxidation (LAO) using the atomic force microscope (AFM). The characterization of the surface by X-ray photoelectron spectroscopy (XPS) reveals that this method assures a modification of the topography of the surface without a change of its chemical composition. Surfaces structured by LAO therefore represent ideal systems to study the dependence of protein adsorption on topography. We are able to visualize the created nanostructures with an AFM and successively adsorb the proteins in situ, rinse and image the new surface. The densities of adsorbed proteins on the nanostructured and neat surfaces are compared and we find that the protein arrangement depends on the underlying nanostructures, showing that proteins can ‘‘sense’’ the topography of surfaces at the nanometer scale. This result can be considered as the nanoscale analogous of the adsorption found for cell systems on micrometer structures.

Keywords : Nanostructures; Topography; Protein; Adsorption; Surface

The simultaneous measurements of the maximal oscillation amplitude and the transient decay time constant of the QCM reveal stiffness changes of the adlayer    
C. Galli Marxer, M. Collaud Coen, H. Bissig, U.F. Greber, L. Schlapbacha
Accepted for publication in Analytical and Bioanalytical Journal, May 2003

The interpretation of adsorption kinetics measured with Quartz Crystal Microbalance (QCM) can be difficult for adlayers undergoing modifications of mechanical properties. We have studied the behavior of the oscillation amplitude A0 and the decay time constant τ of quartz during adsorption of proteins and cells using a home-made QCM. We are able to measure simultaneously the frequency f, the dissipation factor D, the maximal amplitude A0 and the transient decay time constant τ each 300ms under liquid, gaseous or vacuum environments. This enables distinguished analyses between adsorption and modifications of liquid/mass properties. Moreover the surface coverage and the stiffness of the adlayer can be estimated. These improvements promise to increase the appeal of the QCM methodology for any applications, measuring intimate contacts of dynamic material with a solid surface.

Keywords: QCM, amplitude, viscoelasticity, decay time constant, protein, cell

Study of adsorption and viscoelastic properties of proteins with a quartz crystal microbalance by measuring the oscillation amplitude    
C. Galli Marxer, M. Collaud Coen, and L. Schlapbach
Journal of Colloid and Interface Science 261 (2003) 291–298

The adsorption kinetics of protein A, BSA, IgG, and fibronectin has been investigated using a homemade quartz crystal microbalance. Information about the energy losses appearing in the system is measured by the maximal oscillation amplitude and the dissipation factor. Only the maximal oscillation amplitude allows us to distinguish the different contributions of liquid and mass to the total frequency shift. The adsorption of proteins has been performed on Ti and Au surfaces at different concentrations. The amount of irreversible adsorbed protein A and IgG increases with increasing bulk concentrations. On Au more proteins adsorb, but their biological activity is reduced in comparison to Ti. Protein A forms a first monolayer in a few seconds, which shows practically no energy losses, and following this a second monolayer is formed. The adsorption rate for the second monolayer is much smaller and energy losses are present. Fibronectin is forming a very viscoelastic system, whose mechanical properties are affected by immersion in different buffer solutions.

Keywords: Protein A; BSA; IgG; Fibronectin; Adsorption; QCM; Amplitude; Dissipation factor; Viscoelasticity

Cell spreading on Quartz Crystal Microbalance elicits positive frequency shifts indicative of viscosity changes    
C. Galli Marxer, M. Collaud Coen, T, Greber, U. F. Greber, L. Schlapbach
Accepted for publication in Analytical and Bioanalytical Journal, May 2003

Cell attachment and spreading on solid surfaces was investigated with a home-made Quartz Crystal Microbalance (QCM), which measures the frequency, the transient decay time constant and the maximal oscillation amplitude. Initial interactions of the adsorbing cells with the QCM mainly induced a decrease of the frequency, coincident with mass adsorption. After about 80min, the frequency increased continuously and after several hours exceeded the initial frequency measured before cell adsorption. Phase contrast and fluorescence microscopy indicated that the cells were firmly attached to the quartz surface during the frequency increase. The measurements of the maximal oscillation amplitude and the transient decay time constant revealed changes of viscoelastic properties at the QCM surface. An important fraction of these changes was likely due to alterations of cytosolic viscosity, as suggested by treatments of the attached cells with agents affecting the actin and microtubule cytoskeleton. Our results show that viscosity variations of cells can affect the resonance frequency of QCM in the absence of apparent cell desorption. The simultaneous measurements of the maximal oscillation amplitude, the transient decay time constant and the resonance frequency allow an analysis of cell adsorption to solid substratum in real time and complement cell biological methods.

Keywords: cell, QCM, adsorption, cytoskeleton, amplitude, decay time constant

Conclusion and outlook top

The thesis work presents two different studies. In the first one the fundamental question addressed was to know, whether proteins can "sense" the topography, similarly to what had been shown for cells. It required to find a method, which allowed us to modify the topography at the nanometer scale without any change of the surface chemistry. Moreover it was necessary to visualize the nanostructured surface before and after protein adsorption. The AFM is the only technique, which allows us to modify the surface at a nanometric scale and to measure in situ adsorbed biological molecules such as proteins, which are insulators. Nanostructured surfaces by LAO have been characterized by XPS, demonstrating that LAO treatment effectively induces no change of the surface chemistry. Therefore ideal systems can be created by LAO to study the only influence of the topography on protein adsorption. Globular protein A shows no adsorption difference between parallel 1nm high nanostructures and neat Si. The bounded IgG also shows no preferential adsorption on or outer the nanostructures covered with protein A. In contrast, preferential adsorption of filamentous F-actin has been measured on 1nm high created lines on Si, and the density of adsorbed protein on neat Si is higher than on nanostructures. Ti nanostructures of different heights have been created. The F-actin adsorption is the greatest on 1-2nm high lines with a preferential orientation along them. In contrast the adsorption density is low on 4nm high nanostructures, and no defined orientation is noted. On neat Ti adsorbed F-actin shows a smaller density than on the 1-2nm high nanostructures. Therefore proteins are able to react to modifications of the topography at the nanometer scale, what has to be taken into account in the design of new biomaterials.

In order to approach in vivo conditions, further studies have to be performed under aqueous conditions and with the presence of protein mixtures. Moreover the next experimental step consists on the adsorption of cells on the proteins oriented by such nanostructured surface. In this view new methods would be required, because the work area of most actual commercial AFM is restricted to 104 micro m2 and the method is too slow to create nanostructures over larger areas. The lithography technique will be perhaps the solution to produce nanostructures at large scale, because its resolution is becoming improved and its work area is in the cm2 range. Nevertheless further developments are required before the lithography reaches the LAO resolution.

The second part of this work was concentrated on the development of the QCM technique, and especially on the improvement of the interpretation of QCM data. At the moment the QCM is the only method, which allows us to follow the adsorption kinetics and to get simultaneously viscoelastic information of the adlayer. Nevertheless the interpretation of measurements performed with commercial QCM is still difficult, because modifications of liquid and mass properties can not be separated from the variation of adsorbed mass. We have demonstrated that the introduction of the maximal oscillation amplitude and the decay time constant improves the determination of the phenomena arising at the quartz surface. Contribution of adsorbed mass can effectively be distinguished from the one of modification of viscosity or density of the liquid or of the mass. Using the protein A-BSA-IgG system, the biological activity of adsorbed protein A has been investigated on Au and on Ti. At concentration of 160 mM no difference is noted, but divergences have been measured at higher concentrations. More proteins always adsorb on Au surfaces and the adlayers dissipate less energy than on Ti. However the biological activity is better on Ti. Further experiments with other protein systems are required in order to determine if the biological activity can be directly related to the viscoelastic properties of the protein layer. The effect of buffer properties on adsorbed fibronectin has also been investigated. Immersion in Citrate-Phosphate buffer induces only delamination of the adlayer. In contrast the mechanical properties of the fibronectin are modified in Hepes solution.

The QCM open new applications in cell biology and biotechnologies, because the interactions occurring between a surface and a cell can be directly investigated. We have demonstrated that different phenomena influence the frequency, the amplitude and the decay time constant during cell adsorption. During the first 80min the frequency decreases due to an increase of adsorbed cells. Thereafter the frequency begins to increase, what would correspond to mass desorption. Nevertheless living cells are still present at the surface. We demonstrate that the development of the cytoskeleton induces stiffening of the cell and increases the total cell viscosity. This phenomenon influences the QCM sensitivity in frequency, although no modification of the amount of adsorbed cells occurs. The cytoskeleton state has also been modified with drugs, which increase or decrease the polymerisation state of actin and microtubules. It allowed us to reproduce rapid increasing and decreasing of frequency in less than 30min due to fast changes of the cell viscosity and cell adsorption.

The understanding of complex systems such as cells requires nevertheless the combination of different technique. We tried to combine a QCM with a fluorescence microscopy. The quartz electrode has to be modified in order to avoid fluorescence bleaching on Au. Moreover it is essential to work with oil objectives, which have very short working distances. The building of the whole setup is therefore complicated in order to fulfill all requirements, such as temperature stabilization and exchange of fluid. To date no reproducible measurements could be performed during several hours. Nevertheless this process is the only one, which would allows us to correlate the number of developed focal points with the measured frequency, amplitude and decay time constant. Moreover this method combination would allow us to track modification of cell properties during for example virus infection. In order to quantify the variation of the frequency, the amplitude and the decay time constant during viscoelastic modifications of adlayer, colloidal systems can be very interesting. They are firstly much less complex than living cells and secondly they allow us to study phase transitions. Our first tests are very promising.

PhD table of content

Presentation of the work
    The sensor environment
    The QCM-computer interface
    The program
    The resolution

Paper I
Protein adsorption on topographically nanostructured titanium
C. Galli et al, Surface Science (2001)

Paper II
Creation of nanostructures to study the topographical dependency of protein adsorption 
C. Galli et al, Colloids and Surfaces B: Biointerfaces (2002)

Paper III
The simultaneous measurements of the maximal oscillation amplitude and the transient decay time constant of the QCM reveal stiffness changes of the adlayer
C. Galli Marxer et al, Analytical and Bioanalytical Chemistry (2003)

Paper IV
Study of adsorption and viscoelastic properties of proteins with a quartz crystal microbalance by measuring the oscillation amplitude
C. Galli Marxer et al, Journal of Colloid and Interface Science (2003)

Paper V
Cell spreading on Quartz Crystal Microbalance elicits positive frequency shifts indicative of viscosity changes
C. Galli Marxer et al, Analytical and Bioanalytical Chemistry (2003)

Conclusion and outlook

 © 1998-2018 by Carine Galli Marxer