Main text

Need to image sample from a specific well-defined angle
Inhomogeneous samples have an optimal orientation when studied by light microscopy. For example in developmental biology highly scattering tissues such as pigments, the eyes or the yolk can obscure the organ or tissue of interest. Typically, the sample is carefully embedded in its ideal orientation prior to the experiment. However, after embedding the sample’s orientation is fixed and cannot be adapted to the sample’s development during in vivo experiments or to another region of interest anymore.

Define explicitly what the goal is (with respect to your sample, zebrafish).
To study organisms with optimal resolution during development, a technique to dynamically orient the sample in the microscope in any arbitrary orientation is needed. (While different organisms may require tailored solutions), the popular model organism zebrafish is transparent in the early stages making it the ideal sample for in vivo imaging.

Embedding the work in the literature
To align many zebrafish larvae for high-throughput applications in the preferred orientation microfluidic systems have been developed Lin 2015. For ... neuro  zebrafish larvae have been embedded between two coverslips and imaged from two sides by manually turning the sample Ronneberger 2012. While these techniques allow a precise, static sample orientation they still did not offer an adaptive reorientation of the sample. In light-sheet microscopy (or SPIM Huisken 2004) the sample can be rotated about one axis with a rotation stage.  ...sample orientation...Single-axis insufficient for total control over three-dim. orientation. This may pose a problem in developing organisms when imaging tissues deeply buried in the embryo.


[In addition, multi-view: The entire sample is then reconstructed by fusing views from several angles. In each view the area facing the objective lens is well resolved. Consequently, in such a single rotational–axis system uniform resolution is achieved only along the equator, which faces the detection lens. Therefore the lateral resolution at the polar caps is no better than the axial resolution of the microscope (Figure 2b). ] LATER

{Here, we present a remote positioning technique that after the injection of magnetic beads can freely position the sample by applying a magnetic field and show its excellent performance for a zebrafish larva on a commercial stereoscope and a zebrafish embryo on a custom-built SPIM system.} see above

Requirements of an orientation technique
Typically microscopes are equipped with a xyz stage to position the sample in the focal plane of the objective but there is no control of the sample orientation. A positioning method for biological samples needs to fulfill the following criteria: it needs to add a degree of freedom to visualize the sample from the optimal side. The positioning technique must not interfere with the imaging. The method should be easy to integrate in any automated microscopy routine. Ultimately, the technique has to ensure the sample integrity. While optical methods have successfully been used to position and to orient single cells for light microscopy Torres-Mapa 2012 Kreysing 2014 Kreysing 2008, forces are not high enough to position a millimeter-sized zebrafish embryo.

Introduce the idea and show how you go about implementing it
We tried whether we could use magnetic forces to orient a zebrafish embryo within its chorion. We injected magnetic beads into the yolk and we discovered that we could orient the injected embryo within its chorion by approaching a permanent magnet close to the embryo (Supplementary video S1).
Injection of magnetic beads

The super paramagnetic beads were injected with glass needles into the yolk of the zebrafish embryo to ensure the integrity of the zebrafish embryo. We performed the injections at 4 hpf and the injection needle was inserted from either the vegetal pole or the lateral side and beads were deposited close to the yolk membrane (Figure 1b). We found that performing the injections at extremely low pressure (5-10 psi) and long injection duration (100-150 ms) avoided the dispersion of beads. By applying a strong constant magnetic field after the injection of the beads with a permanent magnet the beads were attracted and clump. This aggregation of beads preserved the single beads from translating through the yolk and eased the rotation.

Find the right magnetic beads to orient the zebrafish embryo

To ensure the sample integrity no residual force should affect the sample after it has been oriented. Compared to ferromagnetic beads it turned out that super-paramagnetic beads show no residual force once the magnets are removed since they exhibit magnetic properties in the presence of a magnetic field with no residual magnetism Neuman 2007. The size of these beads needed to be large enough that the beads are stationary and do not translate within the yolk when a magnetic field is applied but they should be small enough that they can be injected without damaging the zebrafish. 


Injection of beads does not alter the zebrafish development
Injected embryos were monitored for 4 days and we found no visible delay or defect in development when compared to non-injected wildtype embryos (Supplementary Figure S1). As the embryo developed, the yolk was being used with the beads staying in the remaining yolk, close to the yolk extension.






The injected beads were attracted by the permanent magnet but the zebrafish couldn’t translate towards the electromagnet since it was encapsulated in its chorion. ROTATION We found that by moving the permanent magnet we can orient the zebrafish embryo contactless within its chorion. (Moving magnet too slow) varying magnetic fields

Magnets
The external magnetic field can be generated by permanent or electromagnets. Although permanent magnets have a higher magnetic field, electromagnets can be controlled by the applied current and hence allow a dynamic adaptation of the magnetic field without any movement of the magnet. The applied force on a super-paramagnetic bead is proportional to the magnetic field gradient Neuman 2007. Therefore, we sharpened the core of the electromagnets to create a sufficiently strong magnetic field gradient, even with a moderate current (300 mA). Heating of the electromagnets was also negligible. We designed the electromagnets to be small enough to be oriented around the sample in any geometry giving the freedom of full 3D sample orientation.

4 Magnets: FIGURE 1

To study developing zebrafish from the optimal orientation we adapted our magnetic orientation technique to a SPIM.

We found that a bead size of 2.8 µm allows a rotation of the sample with moderate magnetic fields without a translation of the beads within the zebrafish chorion. The applied force on the fish is proportional to the number of injected beads hence a higher volume of beads is favorable but the injected volume may damage the zebrafish. We found that a volume as low as 1 nl bead solution corresponding to about 15ng of beads or about 1000 beads is sufficient to rotate the embryo at low currents of 300mA.

Why do we need more than one axis of rotation
SPIM has become the technique of choice for in vivo imaging in developmental biology Huisken 2004 Weber 2014. In contrast to most light microscopy techniques, SPIM provides an axis of rotation about which the sample is rotated to take several views. However, a single rotational-axis system cannot give the full three-dimensional freedom of rotation (Figure 2b). Therefore, not all parts of the sample can be turned in front of the detection objective and are accessible, e.g. for photo-manipulation.

Introduce the magnetic manipulation on a SPIM
To orient the sample in three dimensions at least four magnets were needed since every magnet acts as an attractor fixed in space. Here, we included a tetrahedral electromagnet configuration in a SPIM setup (Figure 2a). We designed a sample chamber that held the four magnets in a tetrahedral geometry and the sample tube.

Performance of the system
We measured the performance of the magnetic rotation and found that a rotation of the zebrafish embryo within its chorion from one magnet to another one (109.5 degrees) took less than 30s at a current as low as 300 mA (Figure 1d). This transition time from one magnet to the next one was anti-proportional to the applied current and could therefore be tuned according to the application. A low current led to slow and homogenous rotation whereas a high current led to a rapid reorientation. After releasing the embryo from the magnetic force the zebrafish retracted less than 10s and settled at its final position. We found that the embryo remained stable in its settled position for over a minute longer than the time it needs to take a three-dimensional stack of the whole embryo on a SPIM. Applying a low magnetic force could stabilize the orientation of anisotropic samples in a preferred orientation. This method is especially suitable to orient samples in an out-of equilibrium orientation

Intermediate angles
To study the sample in the optimal orientation intermediate orientations between two/three magnets need to be accessible. We found that these intermediate orientations could be accessed in two different ways. First, spherical samples with no preferred orientation (as the zebrafish embryo till the bud stage) could be stopped while rotating from one magnet to the next one and settled at this intermediate orientation. Second, by applying a current onto two or more magnets at the same time the zebrafish oriented to the resulting magnetic field. By changing the ratio of the applied currents between the different magnets the sample could be positioned in any orientation in three dimensions.

Tetrahedral geometry gives better coverage
In single sided detection SPIM microscopes the sample is commonly rotated in 90deg steps and four different orientations are acquired to obtain a more homogeneous resolution and better coverage of the sample. These orientations are equally distributed in the plane of rotation. However, due to the 3D nature of the sample this geometry does not cover the whole sample. The four magnets and thereby the four views by switching on only one magnet are homogeneously distributed in 3D (Figure 2b/c). This tetrahedral geometry of the four magnets enables a better coverage of the sample than the single rotational axis SPIM with the same number of stacks. 

4 views of the same embryo
We imaged a zebrafish embryo (??) from four different angles in the single rotational axis SPIM mode by manually rotating the tube in 90deg steps and in the multi-view, multi-axes mode by sequentially orienting the sample with one magnet and taking a stack for each magnet. For the registration of the we aligned the data we manually aligned the different angles in the 3d Viewer (????). The single-rotational axis data could be simply aligned
the data the four different views were successfully fused by






The  FUSION is still missing!

10 different orientations from the same embryo
To obtain an almost isotropic resolution across the whole zebrafish embryo we imaged the embryo from ten different orientations. We took a stack for the four sample orientations given by the tetrahedral magnet geometry and the six orientations for all the possible combinations of two magnets. The resulting data showed no deformations and the embryo could be imaged from ten different orientations (Figure 2)



Limitations of light microscope as spinning disk
In contrast to SPIM, common light microscopes as spinning disk are designed to image the sample from only one side. To orient the sample in the preferred orientation and to study the sample from multiples sides we developed an inset consisting of a plate and an arc holding the two electromagnets.  This inset can easily be adapted to any commercial light microscope.  Here, we show the in situ orientation of a zebrafish larva (5dpf, kdrl?) on a stereoscope. The magnet orientation could be adapted continuously by sliding the magnets along the arc. The sample was positioned on the axis of rotation. The zebrafish larvae was embedded in a glass capillary to prevent it from translating towards the magnets and to ensure the ball bearing mechanism.  

2 different orientations from the same zebrafish larvae
We used two electromagnets to orient the zebrafish larva about its anterior-posterior axis making use of the elongated morphology at the larvae stage. We found that the rotation of the zebrafish larva worked best with an angle of about 35 degrees between the electromagnet and the plate. By switching on one magnet the zebrafish was slightly translated towards the wall of the glass capillary and rotated about 180 degrees towards the magnet. By switching off the magnet the larvae was released from the force and settled in its resting position (Figure 3). A 180 degrees rotation from the left lateral to the right lateral resting position about the anterior-posterior axis took not longer than 10 s at a current of 1 A. By rotating the zebrafish larvae we can image the larvae from the right and the left lateral side within a single experiment.