3D-MINSTED nanoscopy and protein tracking in densely labelled living cells

This paper introduces a 3D MINSTED nanoscopy setup capable of tracking individual proteins with sub-nanometer precision in living cells, successfully resolving the 16 nm steps of kinesin-1 along microtubules even amidst high background and dense protein labeling.

The Gist

The Big Picture: Finding a Needle in a Haystack (While the Haystack is on Fire)

Imagine you are trying to watch a single, tiny firefly (a protein) fly through a crowded, dark forest at night. The problem? The forest is full of thousands of other fireflies, and they are all blinking at once. If you try to watch just one, the light from the others drowns it out.

For a long time, scientists could only watch these "fireflies" if they were in a quiet, empty room (fixed cells) or if they were very far apart. But life is messy and crowded. To understand how our bodies work, we need to watch these proteins move inside living, bustling cells.

This paper introduces a new, super-powerful flashlight called 3D-MINSTED. It's like a magical spotlight that can ignore the thousands of other blinking lights and focus only on the one firefly you want to watch, even in the most crowded part of the forest.

The Problem: The "Donut" Dilemma

To understand the new solution, we first need to understand the old one. Scientists previously used a method called MINFLUX. Imagine trying to find a lost coin in a dark room by shining a flashlight with a hole in the middle (a "donut" shape of light). You move the donut around until the coin is in the dark hole, where it stops reflecting light. By knowing exactly where the hole is, you know where the coin is.

The Flaw: This "donut" light is huge compared to the coin. It covers a massive area. If there are other coins (proteins) nearby, their light spills into your view, confusing the system. In a crowded living cell, this method fails because the background noise is too loud.

The Solution: The "Laser Eraser" (MINSTED)

The authors created a new tool called MINSTED. Instead of just using a donut to find the coin, they use a two-step magic trick:

1. The Flash: They shine a normal, bright laser to make the protein glow.
2. The Eraser: Immediately after, they shine a second, donut-shaped laser (called a STED beam) that acts like an "eraser." This eraser laser is tuned to cancel out the glow of any protein that isn't in the very center of the donut.

The Analogy: Imagine you are in a room full of people shouting. You want to hear one specific person.

• Old Method: You ask everyone to be quiet, but you can only ask a few people at a time. The noise is still too loud.
• New Method (MINSTED): You have a magic noise-canceling headset. You shout "Speak!" to the whole room, but your headset instantly silences everyone except the one person standing in the exact center of a tiny, invisible circle. Everyone else is silenced by the "eraser" laser. Now, you can hear your target perfectly, even if the room is packed with people.

What They Did: The Experiments

The team built a microscope that uses this "eraser" trick in 3D (up, down, left, right, forward, backward). They tested it in two ways:

1. The "DNA Lego" Test (The Ruler Check)
First, they built a tiny grid out of DNA (like a microscopic Lego structure) with known distances between the pieces. They used their new microscope to measure it.

• Result: They measured the distances with incredible precision—better than 1 nanometer (that's about 1/100,000th the width of a human hair). It proved their "eraser" was sharp enough to see tiny details.

2. The "Walking Ant" Test (The Real Deal)
Next, they looked at Kinesin, a motor protein that acts like a tiny ant walking along a rope (a microtubule) inside a cell. These ants carry cargo around the cell.

• The Challenge: They tried to watch these ants walking in living cells that were densely packed with other proteins. Usually, this is impossible because the background is too bright.
• The Result: The 3D-MINSTED microscope successfully tracked the ants. It saw them take steps of exactly 16 nanometers (the size of the ant's stride). It did this even though the "forest" was full of other glowing proteins.

Why This Matters

Before this, watching a single protein move inside a living cell was like trying to watch a single drop of water fall in a hurricane. You couldn't see it clearly.

This new method allows scientists to:

• See clearly in the crowd: They can track proteins even when the cell is "crowded" with other labeled proteins.
• Go 3D: They can see the protein moving up and down, not just side-to-side.
• Be precise: They can measure movements smaller than the width of a virus.

The Bottom Line:
This paper gives us a "super-vision" tool. It allows us to watch the tiny molecular machines that keep us alive as they actually work inside our living bodies, without needing to clean up the mess first. It turns a chaotic, blurry movie into a crisp, high-definition documentary of life at the smallest scale.

Technical Summary

1. Problem Statement

While fluorescence nanoscopy methods like MINFLUX have revolutionized single-molecule tracking with nanometer/millisecond spatiotemporal resolution, they face significant limitations in crowded, live-cell environments:

• Background Noise: MINFLUX relies on a "donut-shaped" excitation beam with a central minimum. Due to diffraction, these beams cover a large sample volume and concentrate power in their periphery, generating substantial unspecific background fluorescence.
• Labeling Density Constraints: To avoid overlapping signals from neighboring fluorophores, MINFLUX requires extremely sparse labeling (typically 3–5 times sparser than standard localization). This makes it difficult to track proteins in living cells where high labeling densities are often necessary or unavoidable.
• 3D Limitations: Previous MINSTED implementations were limited to 2D or fixed cells. There was no established method for 3D tracking of individual proteins in living cells with high precision amidst high background noise.

2. Methodology

The authors developed a 3D-MINSTED (MINimal photon fluxes for STimulated emission depletion) setup that combines the localization precision of MINFLUX with the background suppression capabilities of STED (Stimulated Emission Depletion).

Key Technical Components:

• Beam Configuration:
  • Excitation: A pulsed 532 nm laser (200 ps pulses) excites the fluorophore.
  • STED Depletion: A pulsed 636 nm laser (1.5 ns pulses) is overlapped with the excitation beam.
  • Lateral (x,y) Localization: A STED beam shaped into a 2D donut (via a vortex phase plate) creates a zero-intensity center. The beam is steered in a circle around the estimated fluorophore position using Electro-Optic Deflectors (EODs).
  • Axial (z) Localization: Two additional STED beams (STED Z1 and STED Z2) are used. Their wavefronts are modulated by a Spatial Light Modulator (SLM) with top-hat phase profiles and defocus to create 3D donuts (or "bottle beams") positioned above and below the focal plane ($\pm R_z$).
• Interleaved Pulsing: The system operates at 30 MHz for excitation and 10 MHz for STED. The three STED beams (XY, Z1, Z2) are temporally interlaced to probe lateral and axial positions independently within a single pulse cycle.
• Feedback Loop: Upon detecting a photon, the system updates the focal point toward the position that defined the effective Point Spread Function (E-PSF) causing the detection. This allows the focal point to "circle" or "wiggle" around the fluorophore, refining its position in real-time.
• Polarization Management: A complex optical path involving Wollaston prisms, quarter-wave plates, and resonant electro-optical modulators (Res-EOM) combines the orthogonally polarized Z-beams into a single stream to minimize losses.

3. Key Contributions

• First 3D-MINSTED Live-Cell Tracking: The paper introduces the first setup capable of 3D single-molecule tracking in living cells with sub-nanometer precision.
• Superior Background Suppression: By using a STED donut to suppress fluorescence everywhere except near the intensity minimum, the effective sampling volume is reduced by 30 times compared to 3D-MINFLUX. This allows tracking in environments with **20 times higher protein densities** than possible with commercial MINFLUX systems.
• High Precision in 3D: The system achieves a localization precision ($\sigma$) of < 1 nm in all three dimensions (x, y, z) when sufficient photons are collected.
• Automated Tracking in Crowded Environments: The method successfully isolates single-molecule signals from dense backgrounds, enabling the automatic selection of valid tracks in living cells where traditional methods would fail due to signal overlap.

4. Results

The authors validated the system using two distinct models:

A. 3D DNA Origami Grids (Static Calibration)

• Sample: 3×3 DNA origami grids with known spacing (10.9 nm and 12 nm).
• Performance: The system resolved the grid structure in 3D with high fidelity.
• Precision: Achieved localization precisions of < 1 nm (median) in both lateral and axial directions.
• Measurement: Measured grid constants of $d_x = 10.6 \pm 1.1$ nm and $d_y = 9.6 \pm 0.8$ nm, confirming the ability to resolve structures smaller than 10 nm.

B. Kinesin-1 Motor Protein Tracking (Dynamic Application)

• Fixed Cells: Tracked kinesin-1 moving along microtubules in fixed U-2 OS cells.
  • Observed the characteristic 16 nm step size of kinesin.
  • Measured a mean step length of 16.7 ± 5.8 nm.
• Living Cells: Tracked kinesin-1 in living U-2 OS cells with dense labeling (multiple fluorophores per diffraction-limited volume).
  • Successfully tracked > 10,000 raw tracks, filtering down to 337 valid single-fluorophore tracks.
  • Extracted 1,207 steps with a mean length of 16.7 ± 6.6 nm.
  • Signal-to-Background Ratio (SBR): Achieved a median SBR of 10.8 ± 3.6, demonstrating robust signal isolation despite high background noise.
  • Temporal Resolution: Achieved a temporal resolution of 2.4 ms.

5. Significance

• Overcoming the "Crowding" Barrier: This work demonstrates that 3D-MINSTED can overcome the fundamental limitation of MINFLUX regarding labeling density. It enables single-molecule studies in physiologically relevant, crowded cellular environments without the need for extreme sparse labeling.
• Unprecedented Resolution: The combination of < 1 nm spatial precision and millisecond temporal resolution in 3D provides a new tool for studying conformational changes and motor protein mechanics in their native context.
• Scalability: The ability to acquire hundreds of valid tracks per measurement day suggests that 3D-MINSTED is a viable, routine tool for high-throughput single-protein tracking in live cells, opening doors to studying complex biological processes previously inaccessible to super-resolution microscopy.

In summary, this paper presents a major advancement in optical nanoscopy by extending MINSTED to 3D live-cell imaging, effectively solving the background noise problem that has hindered single-molecule tracking in dense biological samples.