Skip to main content

Biomimetic model of the cell cytoskeleton: polymer networks cross-linked with DNA strands

Final Report Summary - BIOLINK (Biomimetic model of the cell cytoskeleton: polymer networks cross-linked with DNA strands)

Summary description of project objective and expected final results:

The central aim of this project is to understand the physics behind the remarkable mechanical adaptability/tunability of cytoskeletal networks. To achieve this goal, the plan is to adopt a unique experimental approach based on an model system consisting of actin filaments grafted with deoxyribonucleic acid (DNA) strands that form cross-links by specific base-pairing (hybridization). The complexity of the system will be gradually increased.

Schematic outline of the four key objectives of the proposal.
(1) Establishment of an experimental assay consisting of actin networks grafted with DNA strands that form cross-links by specific base pairing (hybridization).
(2) Study of passive actin network mechanics as a function of cross-linker compliance (by varying DNA linker length).
(3) Addition of molecular motors (myosin II) to the network, to study active network formation and mechanics as a function of cross-linker compliance.
(4) Creation of biomimetic model of a mechanosensing cell consisting of actin/DNA-encapsulating liposome to explore the formation of adhesions as a function of the cytoskeletal network mechanical properties.

The use of DNA strands as cross-linkers will provide unprecedented control over both the strength and compliance of cross-links between actin filaments. Since DNA strands are 3 orders of magnitude more flexible than actin filaments, networks can be conveniently tuned between the limits of a rigidly cross-linked gel and a flexibly cross-linked gel by varying the DNA linker length and base-pairing.

The specific objectives for this project are:

1.Development of cross-linked biopolymer networks using DNA strands as cross-linkers:
It is well accepted that the mechanical properties of the cell cytoskeleton are determined to a great extent by the nature and concentration of protein cross-linkers that connect the cytoskeletal filaments. However, we still lack a model system allowing us to precisely control cross-linker compliance. To fill this gap, this project aims to develop an assay to produce actin networks with DNA strands serving as cross-linkers. DNA strands are polymers of nucleotides which each has one of four types of bases attached. Single-stranded DNA (ssDNA) will bind only with a strand that carries the complementary sequence of bases, to form stable double-stranded DNA (dsDNA). Over the past two decades it has been realized that DNA hybridization offers the opportunity to construct designer materials (e.g. arrays of nanoparticles, colloids, polymers, and complex DNA-only structures). However, the potential of DNA cross-links for tailoring mechanical properties has remained largely unexplored. In this project, the extraordinary selectivity and programmability of DNA will be exploited to create biomaterials whose viscoelastic properties can be precisely predetermined on the molecular scale.

2.Establish a quantitative multi-scale relation between molecular cross-linker compliance and macroscopic network mechanics.

Cross-linked networks present an extremely complex mechanical behavior, showing deviations from continuum elasticity and simple affine strain fields. Some studies have also addressed the cooperative effect of pairs of different cross-linkers, showing that their combination can have either a simple linear additive effect on network mechanics, or an enhanced synergistic cooperation. The biophysical basis determining which of these two behaviors occurs remains obscure.

Using the experimental assay developed in Key Objective #1, I will systematically study the relation between molecular cross-linker compliance and macroscopic network mechanics. I will first study the dynamic rheological properties, which are expected to be rich due to the disparity of flexibilities of DNA and actin.

I will then focus on the nonlinear viscoelastic response, which is expected to exhibit strain-stiffening to an extent that will strongly depend on DNA-linker length and concentration. I will combine rheology with direct imaging of the networks, using a home-built shear cell. I will spatially map non-affinity in these networks as a function of shear by measuring displacements of fiducial markers in the network by confocal microscopy. I expect to observe marked differences in non-affinity depending on cross-linker compliance and concentration. Using series of confocal z-plane images of fluorescently labeled networks, I will obtain a 3D reconstruction of the network structure that will serve as input for building realistic theoretical models. Additionally, I will study actin networks cross-linked with pairs of DNA cross-linkers of different compliance, aiming to elucidate the physics governing the different cooperative effects reported in the literature. Comparison of actin networks cross-linked with physiological cross-linkers and with DNA linkers of matching compliance will provide insights on the effect of cross-linker geometry and binding affinity on network mechanical properties.

3.Obtain a systematic understanding of the interplay between motor activity and cross-linker structure in controlling self-organization and mechanics of active biopolymer networks:

Intriguingly, molecular motors are able to actively stiffen cross-linked actin networks by generating internal pre-stress. These networks are inherently out of thermodynamic equilibrium due to the continuous consumption of ATP by the myosin II motors. It was found that flexibly cross-linked biopolymer networks are much more strongly stiffened than rigidly cross-linked networks (at high ATP). The reason for this difference is poorly understood. Moreover, it remains unknown how the complex mechanical properties of actin networks such as non-affinity influence motor-driven stiffening, and how the actin-myosin machinery adapts to the application of external loads.

To resolve these questions, I will use the shear cell to visualize directly the network structure under shear. I will spatially map non-affinities in the strain field as a function of external shear and internal stress applied by the motors. This innovative approach will allow me to relate for the first time active network reorganization by the motors with the ensuing rheological properties. Moreover, to resolve the influence of motor activity on network rheology, I will use laser tweezers to perform active and passive microrheology. By active microrheology (AMR), I can measure the dynamic rheological properties. By passive microrheology (PMR). I can characterize the activity of the molecular motors as the deviation of probe particle fluctuations from thermal equilibrium. In addition, combining the shear cell with active/passive microrheology I will test adaptation of the actin-myosin machinery to load.

This will be the first study of the interplay between motor activity and cross-linker compliance in the determination of the self-organization and mechanical properties of cytoskeletal networks. This project will increase our understanding of the role that physical interactions may play in the regulation of processes such as tissue morphogenesis, as suggested by recent work in from the host lab.

4.Examine adhesion formation in a biomimetic model of a mechanosensing cell as a function of cross-linker compliance and motor activity:
The mechanical properties of the cytoskeleton determine how mechanical forces are propagated through and sensed by the cell. These forces are transmitted to the extracellular matrix through protein complexes localized in the cell membrane, focal adhesions, which involve integrin receptors, which tend to cluster. Integrin clustering has been associated with myosin II forces, pointing towards an inside-out signaling mechanism whose basic principles remain unknown. Focal adhesions are connected to the cytoskeletal network, and the formation and coarsening of that structure are determined through the interplay of myosin II motor activity and cross-linker molecular structure. Therefore, I expect the organization of focal adhesions to present an important dependence on motor activity and cross-linkers, and particularly to observe an enhancement of integrin clustering due to network coarsening.

To answer these questions, I will embed active and passive cross-linked actin/DNA networks inside a biomimetic model of a mechanosensing cell developed by the group of Prof. Dr. M. Dogterom (AMOLF), consisting of a giant liposome containing focal adhesion proteins (integrins). I will combine fluorescence microscopy (to visualize and relate network architecture and integrin clustering), with Reflected Interference Contrast Microscopy (RICM, a technique specially suited to study cell adhesions, established at AMOLF by Dr. M. Leunissen). The proposed experimental assay will also allow probing the liposomes using external forces, to explore the outside-in signaling mechanisms controlling adhesion formation.

Description of the work performed since the beginning of the project and the results achieved so far:

During the first six months of the project, we have worked on the development of the assays to bind actin and DNA in order to form cross-linked biopolymer networks. So far, we have explored two approaches:
-Covalent binding between actin and DNA
-Non-covalent binding (biotin-streptavidin) between actin and DNA,
Covalent binding between actin and DNA.

We tried to covalently bind actin and DNA strands by using an assay consisting in:
-Single stranded DNA (ssDNA), functionalized with an amino group (NH2) at either the 3' or the 5' end,
-Polyethylene glycol spacer (PEG spacer) with a maleimide and a succinimidyl group at each end,
-Actin (preferable in monomeric or G- form).

The assay consists in:
1) bind the DNA-(NH2) and the PEG-(succinimidyl), and
2) bind the PEG(-maleimide) and the Actin (at the active actin cysteine, SH, groups).

Sketch of the approach adopted for the covalent binding between DNA and actin as a two step process requiring an intermediate PEG spacer. First step (top panel): binding ssDNA functionalized with a NH2 group and a PEG spacer functionalized at one of its ends with a succinimidyl group. Second step (bottom panel): binding the PEG spacer, functionalized at its other end with a maleimide group, with actin at its accessible cysteine groups. The green starred symbol at the end of the ssDNA strand represents a fluorescent group, to be used for visualization with confocal microscopy.

Results indicate that the protocol devised is able to label G-actin with the ssDNA-PEG, but they also show that actin is then rendered polymerization incompetent. A possible alternative to achieve an actin network with bound DNA would be to bind the ssDNA-PEG to actin in filamentous form (F-actin). In this way, we would get rid of the polymerization competency problems, although it is clear that this assay would not allow for flexibility in the preparation of the actin-DNA networks.

Non-covalent binding between actin and DNA using biotin-streptavidin (STV).

We have explored two approaches to bind actin and DNA non-covalently:
1.Actin Biotin + STV + ssDNA Biotin
2.Actin Biotin + dsDNA-STV

1.Actin Biotin + STV + ssDNA Biotin
To proceed with this protocol, we use 2 biotinylated DNA strands which are complementary at their ends. We then hybridize these two strands of biotin-DNA, and mix them G-actin in polymerization buffer, resulting in a cross-linked network.

The obvious drawback of this assay is that we cannot avoid the formation of rigid cross-links between two actin-biotin units through STV. However, we can try to minimize the formation of those by tuning the stoichiometry of actin-biotin and biotin-DNA.

Sketch of the non-covalent binding assay between biotin-actin (green) and biotin-DNA (black strands with red circles) using a biotin-streptavidin (STV) union. Gel electrophoresis confirms that both DNA strands hybridize.

Despite the inconvenient of the formation of rigid cross-links, because of its simplicity, we used this assay to verify the presence of DNA cross-links in our networks.

We used a bulk-rheometer with a plate-plate geometry to compare the elastic and viscous component of the mechanical properties of 24μM actin network under 3 conditions:
a)No DNA cross-links,
b)DNA cross-links (with rigid cross-links contributing),
c)No DNA cross-links (and only rigid cross-links contributing).

This network mixed actin with DNA-biotin of only 1 type of ssDNA-biotin, therefore not able to hybridize with the other strand and form DNA cross-links.

The results show that, as a first approximation, the DNA cross-links increase the elastic modulus of the actin network in approximately 3Pa, which is the difference in G' between networks a) and b).

Resulting Elastic Modulus, G', and Viscous Modulus, G'', for an actin-DNA network created using the first non-covalent binding assay.

2.Actin Biotin + dsDNA-STV
Right before the preparation of this report, we started exploring the possibility of preparing double stranded DNA to be biotinylated and bound to STV, to later mix it with actin-biotin. This assay will avoid the problem of the formation of rigid cross-links described previously.

POTENTIAL IMPACT AND USE OF THE RESULTS FROM THIS PROJECT (including the socio-economic impact and the wider societal implications of the project so far):

This project will have an impact on biotechnology and materials science, given the almost direct applicability of the experimental assay developed for this project for the creation of cell-inspired materials (for materials science purposes) and tissue-like matrices of accurately defined properties (for tissue engineering/repair purposes). The European Union will then be placed in an extraordinary position for the development of these technologies. The valorization of this project, that is, its translation from basic research into practical applications, will be enhanced by the active concern that the host institution, AMOLF, has in cooperating with industry, as expressed in its mission statement and proved by the active collaborations established with industrial companies (i.e. Philips, Unilever, Shell). The technological and biomedical applications derived from this project can potentially give rise to long-term synergies between the host institute and biomedical/engineering companies, likely to have a relevant structuring effect in the European economy.

Related documents