Nano fountain pen patterning of microtubule tracks for guiding molecular motors

Abstract

This study explores a new method for transporting biological cargo using motor proteins along their natural tracks, microtubules. Utilizing the nano fountain pen (NFP), we aim to fabricate microtubule pathways connecting designated origin and destination points, offering tracks for kinesin molecules to transport biological cargo. To achieve this, we patterned avidin lines, to which biotinylated microtubules are adhered. The functionality of these immobilized microtubules is confirmed by observing the progressive movement of single kinesin molecules. Furthermore, we demonstrate microtubule alignment through hydrodynamic drag. This method serves as a foundational step toward developing a conveyor network for miniaturized ‘lab-on-a-chip’ systems, with continued potential for advancements in MT alignment and polarization techniques.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-54307-4.

Introduction

The increasing demand for monitoring environmental and biological threats, such as biohazards, food pathogens, and clinical diagnostics, has highlighted the limitations of current centralized laboratory methods. These methods, involving skilled personnel and expensive, time-consuming equipment, are not ideal for quick and portable detection. To address these challenges, the concept of “lab-on-a-chip” has surfaced, with the goal of condensing complex laboratory processes to fit onto a single, miniature chip. The fundamental concept driving lab-on-a-chip technology is the integration of diverse processing and assaying stations onto a compact chip. Consequently, samples must be transported from one point to another with precision and efficiency. This requirement has led to significant advances in engineered shuttle systems within the field of nanotechnology and Microelectromechanical Systems (MEMS)^1,2.

A distinct approach in addressing the challenge of cargo transport has drawn inspiration from biology, specifically the intracellular transport of molecular cargo^3–5. This natural process involves nano-sized motors belonging to the kinesin and dynein families, which move along microtubule (MT) tracks to transport vesicles and organelles between well-defined sites within the cell. Harnessing and mimicking this intracellular cargo transport process in artificial platforms can lead to significant advances in biofluidics, lab-on-a-chip and personalized medicine applications and has thus led to considerable scientific interest⁶. Nonetheless, this approach poses substantial technological challenges, demanding meticulous manipulation and precise control of motors between predefined stations. Over the years, diverse strategies, including chemical modifications, topographical techniques, and magnetic and electrical manipulations, have been explored to achieve MT organization and guided movement^7–10. Most of these efforts have focused on transporting MTs on top of surface-immobilized motors, a method referred to as the “gliding assay”. While this approach facilitates motion, controlling the directionality and velocity of transport has proven to be exceedingly challenging^11–13.

In this study we explore the alternative approach where MT paths are controlled, and the motors are guided along these paths, analogous to the situation in the natural cellular environment (or to guiding a train along fixed tracks rather than moving tracks upon fixed trains). Our method involves the use of a nano fountain pen (NFP)^14,15, for nanoscale precision writing of avidin line patterns. Biotinylated MTs are hydrodynamically aligned in the flow direction within a liquid chamber and subsequently attach specifically to avidin lines prewritten (with NFP) on the bottom of the liquid chamber.

To demonstrate MT functionality, Green Fluorescent Protein (GFP)-tagged kinesin-5 motors (Cin8) were introduced into the chamber under high ionic strength conditions. Previous work from our laboratory had demonstrated that under such ionic strength conditions, the majority of full-length Cin8 motors move processively towards minus-end of the MTs. The average running length of Cin8 molecules was ~ 2 μm on Taxol-stabilized MTs^16,17 and GMPCPP-stabilized MTs¹⁸. The velocity of minus-end directed motility of Cin8 exhibits a wide range of 20–600 nm/s on Taxol-stabilized MTs^16,17 and 20–1000 nm/s on GMPPCP-stabilized MTs¹⁸. Thus, examination of Cin8 motility under high ionic strength conditions should allow for precise control of cargo direction, contingent on ensuring MT’s consistent polarization, to be addressed in a future step. Achieving the same polarity of the aligned MTs would facilitate the shuttling of cargo over long distances.

In the foundational step reported herewith, we concentrated on parallel alignment of MTs, disregarding their polarity, adhered specifically to pre-written avidin tracks and demonstrated normal transport of molecular motors on them. In the next step, we will concentrate on polarization alignment using a combination of magnetic attributes^19–22 and flow fields⁸.

Materials and methods

Nano fountain pen (NFP) system

The nano fountain pen is a customized glass or a quartz nano pipette (Nanonics Imaging Ltd.) with an aperture of diameters 500–1000  nm at its end^14,15. This nano pipette is mounted as a probe of a scanning probe microscope (SPM) system (Nsom Topaz – Nanonics Imaging Ltd.) allowing control of its movement in three dimensions with nanometer precision. The pipette is filled with a solution that migrates to the tapered tip due to capillary forces but is held at the edge of the tip by surface tension. When contact is made with the surface, the solution flows to the surface, similar to a fountain pen. The tip can then be dragged along the surface, under the control of a feed-back loop that maintains constant pressure, resulting in writing of lines. The line thickness is controlled by a combination of aperture diameter and tip velocity.

Surface preparation for NFP writing

Clean and freshly silanized coverslips were prepared as previously described²³. Briefly, the coverslips were silanized by dipping in a solution of dimerthyldichlorosilane diluted in trichloroethylene (Sigman-Aldrich). These silanized coverslips were stored under vacuum for up to two weeks.

The surfaces were then coated with biotinylated bovine serum albumin (bBSA) (Sigma-Aldrich). This coating was achieved by immersing the coverslips in a bBSA solution diluted in triple distilled water (TDW) for 15 min, followed by a brief 1-minute rinse in TDW and drying with a dry spinner. This procedure created a uniform, smooth coating of bBSA on the coverslip surface.

Fluorescent avidin solution for patterning

A solution containing a mixture of unlabeled and fluorescently labeled avidin was used to pattern macro-drops (see details below) and NFP-written lines. It was prepared by combining unlabeled NeutrAvidin (Thermo Scientific) with fluorescent tetramethylacetilrhodamine (TAMRA)-conjugated streptavidin (Invitrogen by Thermo Fisher Scientific) in 10 mM Hepes, pH 7.0. The optimization of this mixture served two purposes: (1) to ensure a high density of avidin binding sites for the later attached biotinylated MTs, and (2) to achieve optimal contrast for imaging MTs (red fluorescence) against the avidin lines (also red fluorescence). The optimal mixture was determined to be 50 µg/ml of unlabeled avidin (neutravidin) and 0.1 µg/ml of labeled avidin (streptavidin). This mixture is hereinafter referred to as “avidin.”

Microtubule preparation

MTs were polymerized at 37 °C in the presence of 1mM guanosine- 5’-[(a, b)-methyleno] triphosphate (GMPCPP) by mixing biotinylated tubulin, rhodamine-labeled tubulin, and unlabeled tubulin (Cytoskeleton Inc.) in a 3:1:10 molar ratio in PEM buffer (PEM; 80 mM PIPES pH 6.9, 2 mM MgCl₂ and 0.5 mM EGTA). Following centrifugation, MTs were resuspended in PEM buffer supplemented with 10 µM Taxol (PEM-Taxol)^18,24 and stored at room temperature for further use.

Microtubule attachment on avidin patterns

Fluorescent avidin mixture (as previously described) was applied onto bBSA coated coverslips (prepared as described above) and dried under vacuum. The solution was applied either as 0.3 drops, resulting in ~3 mm diameter disks (referred to as “macro-drops”), or written with NFP, producing lines ~ 30 μm wide. A clean coverslip was later affixed atop the avidin-bound coverslip using double-sided sticky tape, thus creating a flow chamber that encloses the avidin structures (either macro-drops or NFP written lines) within.

For flow-alignment experiments, the assembled chamber was first washed with PEM buffer to dispose of unbound avidin. It was then perfused with MTs and incubated at room temperature for 15 min. Subsequently, the chamber was washed with PEM-Taxol to remove the unbound MTs.

For experiments with un-aligned MTs, the MTs were attached to the avidin structures without using the flow chamber. This was achieved by directly applying the MT solution onto the macro-drops or NFP lines and subsequent washing by dipping in PEM-Taxol.

Single molecule motility assay

Single kinesin molecule motility assays were performed with Saccharomyces cerevisiae kinesin-5 Cin8-GFP, as described previously²⁴. Briefly, purified Cin8-GFP, in motility buffer (30 mM Pipes, 50 mM Tris, 1mM EGTA, 140 mM KCl, 1 mM ATP, 2 mM MgCl₂, 0.1 mg/ml casein, 10 mM Glucose, 1 mM DTT, 0.1 mg/ml glucose oxidase, 0.8 mg/ml catalase, ATP-regeneration system containing 10 mM phosphocreatine, 0.05 mg/ml creatine phosphokinase, pH 7.2,) was added to the flow chamber containing MTs immobilized to the avidin patterns and imaged. For Cin8 motility on unoriented MTs, motors solution in motility buffer was applied directly onto the surface-bound MTs, incubated for 5 min, washed with motility buffer and imaged (see below).

Imaging

Two different microscope systems were used in these experiments. Fluorescent NFP lines were imaged by Axioplan 2 Zeiss fluorescence upright microscope with a 60X oil immersion Olympus objective (N.A. 1.45) or a 10X Zeiss objective (N.A. 0.3), 100 W mercury lamp and an iXon3 EMCCD camera (Andor Technology Ltd.). Alternatively, rhodamine-labeled MTs and time lapse images of single molecule motility with GFP-labeled Cin8 kinesin motors were acquired using an Axiovert 200 M Zeiss fluorescence inverted microscope equipped with a 100X oil immersion Zeiss objective (N.A. 1.4) and an sCMOS Neo camera (Andor technology Ltd.).

The images were taken with the use of two sets of filters. For imaging of the GFP labeled kinesin motors, excitation maximum of with emission maximum ofFor the imaging of the labeled MTs and avidin NFP lines, excitation of nm] with the emission of . Images were processed and analyzed using ImageJ (Image Processing and Analysis in Java) software. All the reported values represent average ± SEM.

Results and discussion

In the initial optimization stages, we studied the attachments of fluorescently labeled biotinylated MTs onto avidin macro-drops with and without hydrodynamic alignment. In the next steps, we studied the MTs onto NFP created lines.

While avidin-biotin chemistry is generally considered benign in bulk applications, its application via localized NFP deposition may introduce unique micro-scale variables, such as altered surface energies or local concentration gradients at pattern boundaries. It was therefore scientifically rigorous to verify - rather than assume - that the structural and functional integrity of the microtubules remains uncompromised under these specific patterning conditions.

Unaligned MTs on macro-drops

Avidin macro-drops were deposited on silanized cover slips as described in Materials and Methods. MTs were then attached to the avidin macro-drops by applying the MT solution directly on the macro-drops and subsequent washing by dipping in PEM-Taxol.

MTs are attached specifically to the avidin patterns macro-drop, as can be seen in the fluorescence image (Fig. 1 A). The dark (right) side of the image represents the glass surface, whereas the rest of the image is populated by a part of the avidin macro-drop. Attached MTs are visible as thin lines on the macro-drop, but not on the glass (dark background), demonstrating the specificity of attachment. The MTs are unaligned, as demonstrated in the angle distribution of MTs (Fig. 1B).

Fig. 1.

Unaligned microtubules on avidin macro-drops. (A) Specific attachment of biotinylated MTs (thin lines) to an avidin macro-drop. The right edge of the drop is imaged, no MTs are observed on the surface outside the avidin drop (dark region in the far right of the image). (B) Angle distribution of the MTs orientation. (C) Zoomed images inside the avidin drop: left- MTs (red channel), middle- Cin8-GFP motors bound to the MTs (green channel), right- merged image of red and green channels. Average MT length was µm (n = 40).

In the next step, Cin8-GFP motors in motility buffer were applied to the macro-drop, without flow chamber assembly. Figure 1 C (left panel) shows the avidin drop and the MTs visible in the red fluorescence channel. Figure 1 C (middle panel) shows the Cin8-GFP motors visible in the green fluorescence channel. Figure 1 C (right panel) shows the merged image of the red and green channels. To examine the motility of the motors, time-series frames were acquired (supplementary movie S1), from which kymographs of Cin8-GFP motility were generated (Fig. 4).

Fig. 4.

Motility of GFP-tagged Cin8 motors on rhodamine-labeled MTs, attached to avidin macro- drops (A, B, supplementary movies S1, S2) or NFP-written lines (C, D, supplementary movies S3, S4). (A–D) kymographs of motor runs on MTs attached to the surface. Directionality of the MTs is indicated at the top of each kymograph, based on motor accumulation at the minus-end of the MTs, by an asterisk. Fast minus-end directed runs are indicated by yellow arrows; slow plus- or minus-end directed runs of Cin8 dusters are indicated by green arrows. Time and length scales are indicated. E. Box-plots representing velocity distribution of the four categories (A–D). Box height represents IQR, horizontal line is the median, small square indicates the average, whiskers are at the data point adjacent to 1.5 IQR.

We used here the GFP-labeled kinesin-5 Cin8 motors as a proof-of-concept, to examine if the adsorbed MTs can function as tracks for molecule motor motility, since it has been previously demonstrated that the same MT can serve as a track for single molecule motility for both kinesin-1 and kinesin-5 motors, applied simultaneously^25,26. Although in cells, kinesin-5 motors crosslink and slide apart MTs, single molecules of these motors have been demonstrated to move processively (with run length of ~ 2 µm^16–18) along single MTs in vitro under high ionic strength conditions^18,24, as those used in the current experiments (see materials and methods).

Aligned MTs on macro-drops

Avidin macro-drops were deposited on silanized cover slips, as described previously. In the next step, a flow chamber with a height of 0.15 [mm] was assembled by securing a coverslip above the deposited avidin macro-drop using two strips of double-sided adhesive film. Subsequently, a solution of MTs was introduced into the chamber by bringing the tip of a pipettor into contact with the gap between the coverslip and the slide. Capillary forces between the coverslip and glass rapidly drew liquid into the chamber, creating a strong flow that aligned MTs parallel to the flow direction. The fluorescence image in Fig. 2 A shows MTs specifically attached to the avidin drop, displaying a near-parallel alignment determined by the flow direction. The angular distribution in Fig. 2B clearly indicates flow-induced alignment of the MTs. The three panels of Fig. 2 C show the red (left panel), green (middle panel) and merged image (right panel) of the sample after the introduction of Cin8-GFP motors into the chamber. Time-series frames were acquired (supplementary movie S2), from which kymographs of Cin8-GFP motility were generated (Fig. 4). We conclude that this mode of specific attachment of MTs renders them parallel, along the direction of flow, and this arrangement proves functional for the motility of motor proteins.

Fig. 2.

Aligned microtubules on avidin macro-drops. (A) Specific attachment of biotinylated MTs (thin lines) to an avidin macro-drop. The right edge of the drop is imaged, no MTs are observed on the surface outside the avidin drop (dark region in the far right of the image). (B) Angle distribution of the MTs orientation showing a pronounced peak at ~ 10°. (C) Zoomed images inside the avidin drop: left- MTs (red channel), middle- Cin8-GFP motors bound to the MTs (green channel), right- merged image of red and green channels. Average MT length was µm (n = 40).

MTs on NFP-generated avidin lines

The nano-pipettes were filled with red-labeled avidin solution and used to inscribe lines ~ 30 μm wide and ~ 400 μm long (referred to as “micro-lines”), on surfaces functionalized as described in Materials and Methods, as shown in Fig. 3A.

Fig. 3.

Micro scale avidin lines written with the NFP. (A) fluorescent image of the avidin lines. (B) Specific attachment of MTs to the avidin micro-line with a random alignment and (C) with a parallel alignment (D) Merged image of the red unaligned MTs and the green Cin8-GFP motors and (E) Red aligned MTs and the green Cin8-GFP motors. (F) Angle distribution of the unaligned MTs (n = 144) and (G) parallel aligned (n = 211). Average MT length was µm (n = 40), and µm (n = 40), for unaligned and aligned MTs, respectively.

MTs were attached in an unaligned manner, by first dropping the MT containing solution onto the avidin lines and subsequently assembling the chamber. This resulted in specific attachment of the MTs to the avidin micro-lines with random orientation (Fig. 3B). When MT solution was introduced into the pre-assembled chamber by bringing the tip of a pipettor into contact with the gap between the coverslip and the slide, MTs were specifically attached to the avidin micro-lines, aligned in parallel with the direction of the flow as shown in Fig. 3C. The angle distributions of both of the above cases are shown in Fig. 3D.

We found that microtubule lengths were highly consistent across all substrates:

NFP Lines: (Aligned) vs. (Unaligned).

Drops: 11.61 (Aligned) vs. 9.74 (Unaligned).

Since the lengths are uniform across conditions and the standard error is low, we concluded that microtubule length is not a confounding variable in our determination of orientation angles. Introducing Cin8-GFP to the chambers, resulted in Cin8 binding to the MTs (Fig. 3C and supplementary movies S3 and S4 for unaligned and aligned MTs, respectively). Processive motility of the motors was observed in both types of samples. Single molecule motility was demonstrated by kymographs (Fig. 4).

Characteristics of Cin8 motility were in accordance with previously reported results, including accumulation of Cin8 at the minus-end of the MTs, fast minus-end directed motility of the majority of moving motors and slow plus-end, and minus-end directed motility of bright clusters of Cin8 (Fig. 4 and ref¹⁸). Kymographs of Cin8-GFP motility were generated (Fig. 4A-D) from the time-series frames that were acquired. The average speeds of the Cin8 motors (Fig. 4E) align with the previously reported values^18,24. While minor variations in mean velocity were observed between conditions, these fall within the broad stochastic range characteristic of Cin8 motility (20–1000 nm/s) and do not indicate a systematic effect of track alignment on motor kinetics. Moreover, the average running length of Cin8 motors obtained in this study is 1.7 ± 0.1 μm (n = 34), which is within previously published range^16–18. These results indicate that the adsorbed MTs onto NFP-generated patterns, either aligned by hydrodynamic flow or not, preserve their ability to function as tracks to movements of molecular motors (N = 8 to 10 microtubules per condition, with specific velocities detailed in Fig. 4E).

Conclusions

We have demonstrated the targeted binding of MTs to both macro-drops and micro-lines of avidin with high specificity. The micro patterning of MT tracks was made possible due to the employment of the Scanning Probe Microscopy based NFP. We achieved hydrodynamic, flow-induced alignment of the MTs which ensured their functional integrity, as demonstrated by motility assays of kinesin motors (Cin8). This constitutes a method for the specific attachment of MTs to predefined micro patterns while preserving their functionality, thereby allowing the normal motion of molecular motors.

The current results suggest that increasing the concentration of MTs and decreasing the width of the avidin micro-lines (by reducing the NFP diameter) holds the potential to create continuous tracks of MTs aligned with pre-designed patterns. While the present study focuses on low-density distributions to facilitate single-molecule observation, the transition to high-density, continuous tracks is a logical progression for future optimization. While hydrodynamic alignment is a straightforward approach, it is important to note that the polarity of the MTs remains uncontrolled. This work thus reports on the first important step towards the stated goal. To establish a robust platform for long-range motor-driven movement, it is crucial for the MTs to attach to the avidin tracks with consistent polarity. Furthermore, investigating steering fidelity relative to pattern curvature and the influence of microtubule length will be essential for full system optimization. These aspects require further investigation and exploration in our future research endeavors.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

O.F. Performed all the experiments and analyzed the data. H.P. Performed the initial imaging of Cin8 prepared the MTs and helped developing the motility assays. M.P. Extended technical aid with MTs polymerization and attachment of magnetic beads. L.G. and L.A.G. designed, supervised and oversaw the research and obtained funding.

Funding

This work was supported in part by the Israel Science Foundation (ISF) grant no. 629/23, awarded to Larisa Gheber.

Data availability

The raw data supporting the conclusions of this article will be made available by the authors upon request.

Declarations

Competing interests

The authors declare no competing interests.