Laser Micromachining of Bone as a Tool for Studying Bone Marrow Biology

Abstract

The bone marrow (BM) has traditionally been a difficult tissue to access because it is embedded deep within the bone matrix. It is home to the hematopoietic stem cells (HSCs) that give rise to all blood cells in the body. It is also the site of origin for malignant blood cells such as leukemia and multiple myeloma, as well as a frequent site of metastasis for many solid tumors including prostate and breast cancer. The following chapter describes how laser micromachining of bone can be used to improve both optical and physical access to the BM. For example, laser thinning of the overlying bone can improve optical access, enabling deeper imaging into the BM as well as enhancing optical resolution by reducing scattering and aberration. Laser micromachining can also be used to provide physical access into the BM by creating access ports for micropipette insertion and delivery of cells to precise locations in the BM, as well as for the extraction of BM cells and interstitial fluid, all under image guidance. This chapter provides a detailed protocol for installing a laser-micromachining capability for users with an existing multiphoton microscope. Additionally, we briefly outline how such a system improves the optical resolution during imaging as well as its potential use to study injury response.

1. Introduction

The bone marrow (BM) is home to the hematopoietic stem cells (HSCs) that sit at the apex of all blood and immune cell production. It is also home to hematologic malignancies like leukemia, myelodysplastic syndrome, and multiple myeloma, as well as a frequent site of metastasis for many solid tumors, including prostate and breast cancer. Intravital microscopy, in combination with transgenic mice and fluorescent labeling strategies, has provided important insights into the basic biology of normal and malignant hematopoiesis [1–11]. It has also been used to study the homing and engraftment of HSCs in the BM after transplantation [12, 13]. The imaging depth, however, remains limited due to the highly scattering bone matrix which restricts the ability to reconstruct the 3D spatial organization of BM tissue in vivo.

In this chapter, we describe how laser micromachining of bone can be used to improve imaging depth and resolution. Moreover, we describe how to generate small defects in the bone as a methodology for studying injury response. These defects can also be used as access ports for insertion of a micropipette, allowing for the delivery of cells to the BM under image guidance, as well as the isolation of cells from precise anatomical locations [14, 15]. As a central part of this experimental strategy, plasma-mediated laser ablation is used to etch bone tissue away with a precisely defined geometry. Plasma, a highly ionized state of matter, is generated when the energy of femtosecond laser pulses (typically kept below 1 nJ for multiphoton microscopy) exceeds a few nJ, generating electric fields at the focal point that are sufficiently strong (~100 MV/m in our system) to initiate tunneling and multiphoton ionization [16] which leads to a dense region of ions and electrons. Because of the short laser pulses and the tight-focusing geometry, the plasma is tightly confined in space and time, with a dimension of ~1 μm³ that collapses after termination of the laser pulse. This removes a ~1-μm³ volume of bone with each laser pulse while imparting minimal thermal damage to the surrounding tissue. If the ablation is performed under fluid immersion, as in most highresolution multiphoton imaging applications, each pulse is accompanied by a cavitation bubble that expands and collapses on the timescale of ~1 μs. In essence, this is a highly precise microsurgery procedure that limits tissue damage within <20 μm of the surgery (ablation) site. It is sometimes referred to as a “laser scalpel” and is used clinically in ophthalmic surgery [17]. In the following, we describe how to integrate a plasma-mediated laser ablation setup into a conventional multiphoton microscope (Subheading 3.1) and provide a detailed protocol for bone thinning (Subheading 3.2) and the generation of bone defects (Subheading 3.3). Laser bone thinning provides greater imaging depth by reducing the bone thickness and concomitant scattering [18]. On top of that, this has the added benefit of improving optical resolution. With some modifications, the technology can be used for cell delivery to the BM [19, 20] as well as the extraction of BM cells [18, 19] and interstitial fluid.

2. Materials

2.1. Materials and Supplies for Installation of the Ablation Capability

We organize the components necessary for installing the ablation capability on the multiphoton microscope into two categories: optics and mechanical components. Whenever necessary, we provide information on the quantity of the required items. We assume the ablation capability is being integrated into a functioning multiphoton microscope (including an imaging laser, scanning optics, and objective lens) and therefore do not provide information on these components.

2.1.1. Optical Components

1. A femtosecond laser source with a repetition rate of ~1 MHz (e.g., Satsuma; central wavelength, 1030 +/−5 nm; see Note 1).

2. Mirrors M1–M4 (see Fig. 2): Round, protected aluminum mirrors (Thorlabs, PF10–03-g01, qt. 4) with corresponding mirror mounts (Thorlabs, KM100, qt. 4).

3. Lens L1 and L2 (see Fig. 2): Mounted achromatic doublet lens (f = 75 mm) (Thorlabs, AC254–075-AB-ML, qt. 2) with corresponding lens mount (Thorlabs, CP35, qt. 4), attached to a post and holder (see Note 2). Lens L2 is mounted on a translation stage (Thorlabs, MS1S) to facilitate its positioning along the laser propagation axis (needed to optimize collimation; see Note 3).

4. For combining imaging and ablation beam (see Note 4): Dichroic mirror (DC) 950-nm cutoff (Thorlabs, DMSP950R, short pass) or 980-nm BrightLine^® single-edge laser-flat dichroic beamsplitter (Semrock, BLP01–980R-25). Corresponding mount (Thorlabs, FFM1 (Filterholder) and B4C (Platform), along with C6WR (Cage Cube)). The cage cube can be fixed to a regular post and holder but helps ensure a 45° angle with respect to both the imaging and ablation beam.

5. Apertures for the intermediate image plane (see Note 5).

6. For characterizing beam diameter: CMOS camera, Zelux monochrome (Thorlabs, CS165MU). Only required for troubleshooting (see Note 7).

7. For measuring laser power (if not already purchased for the multiphoton microscope): Slim Photodiode Power Sensor (Thorlabs, S130C) and Compact Power and Energy Meter Console, Digital 4^″ LCD (Thorlabs, PM100D).

8. Alignment disk for beam alignment (Thorlabs, VRC4CPT, qt. 2).

Fig. 2.

Detailed optical diagram for installation of the ablation beam. M1–M4 mirrors 1–4, L1,2 lens 1,2, DC dichroic needed for overlapping the ablation beam and imaging beam, M mirror, L lens, OL objective lens, IP image plane, im. IP intermediate IP

General Comments

• We recommend using either imperial or metric-sized components. Do not mix and match.

• All mounts should be connected to a post (Optical Post, SS, 8–32 Setscrew, ¼^″−20 Tap, L=4”, Thorlabs, TR4, qt. ~20) inserted into a post holder (Post Holder, Spring-Loaded Hex-Locking Thumbscrew, L=3”, Thorlabs, PH3, qt ~20) and fixed to the optical table by clamping forks (Thorlabs, CF175, qt. ~20) with M6 screws (Thorlabs, SH6MS16V, qt. ~30). If the space on your optical table is limited, we recommend using short clamping forks (e.g., Thorlabs, CF038 or CF125C).

• As a general recommendation, it is advised to install optical components/mounts in a way that leaves room for translation and optimizing their positioning.

• To maintain a consistent experimental performance level, it is strongly recommended to mark the position of the laser beam at multiple locations throughout the optical system by installing irises (Thorlabs, SM1D25) along with their mounts (post, post holder, clamping fork). As part of the daily alignment, you can check the laser position with respect to the irises and make any necessary adjustments.

2.1.2. Mechanical Components

1. Motorized Micromanipulator (Sutter Instrument, MPC-385). The MPC-385 System consists of the MPC-200 controller and one to four MP-285 micromanipulators.

2. A post and holder, used to mount the sample on the micromanipulator. A right-angle clamp (Thorlabs, RA90) or rotation stage (Thorlabs, PR005) can be used to obtain greater flexibility in sample positioning.

2.2. Materials and Supplies for Laser Ablation

To obtain high-quality ablation geometries with a water immersion objective lens (used on most multiphoton microscopes), it is necessary to generate a continuous stream of water or saline across the sample (flushing), removing gas and debris that is generated during the ablation procedure (see Fig. 3). In the following we have listed all relevant components that are required for setting up a flushing system:

Fig. 3.

Flushing system. A peristaltic pump is used to flow a continuous stream of PBS, saline, or water across the sample. Flow directions marked by red arrows

1. Peristaltic pump, MPM REGLO ICC 2CHNL 6RLR (Cole-Parmer, EW-78001–58), and tubing (L = 16^″, D = 1.85 mm, ISMATEC, 95714–24) for controlling flow speed.

2. 2 Male Luer Lock to Hose Barb Adapters, 1/16^″ (Cole Parmer, UX-45513–00).

3. Blunt-tip needles (22G, 2^″, Grainger, 5FVC4, 2x) that are positioned at the tip of the objective lens and attached to the tubing of the peristaltic pump with the adapters.

4. Falcon^® 50-mL High Clarity PP Centrifuge Tube (CORNING, 352098, qt. 10), that is filled with PBS (Thermo Fisher Scientific, #10010002).

5. Rubber-Cal Neoprene Sheet (0.5 in thick × 12 in wide × 24 in length, Lowes, #2696930). This is used to install the flushing system needles onto the objective lens (see Fig. 4).

6. A Hamilton needle (Luer Lock, 22G, Thermo Fisher Scientific, 14–815–412).

7. An excised piece of bone that can be mounted to a glass slide (BoliOptics, SL39101003) with Krazy Glue (#457099, Staples) or a mouse (see below).

Fig. 4.

Preparation of rubber slab to become a needle holder. Flushing requires a holder for the needles that can be prepared as shown here. (a) Remove a circular piece with dimensions of the objective lens. (b) Etch small holes into the rubber fitting, so that needles can be positioned at the tip of the objective lens

General Comment

While we describe using a piece of rubber to fix the needles on the objective lens, we typically use a custom-designed, 3D-printed holder (drawing available upon request).

2.3. Materials for In Vivo Imaging

1. General supplies: Deionized water, molecular biology grade (Boston BioProducts, WT-035), isopropanol (CVS, 152074), lens tissue (Thorlabs, MC-5) for cleaning optics, Ethilon 6–0 suture, 18^″ (Zogo Medical, ETH-1956G), Curad triple antibiotic ointment (Curad, CUR001231H), PBS (Thermo Fisher Scientific, #10010002), and Qtracker 655 vascular labels (Thermo Fisher Scientific, Q21021MP).

2. Mice: C57BL6/J #000664 (Jackson Laboratories) or a fluorescent reporter mouse of your choice.

3. Mouse surgical kit (Kent Scientific, INSMOUSEKIT). When autoclaving your instruments, please familiarize yourself with the manufacturer’s specifications.

4. Sterile, individually, wrapped transfer pipettes (Thermo Fisher Scientific, 13–711–20).

5. Q-tips (Thorlabs, CTA10, qt. 10). Sterilize by autoclave.

6. Anesthesia: Isoflurane (Piramal Critical Care) and vaporizer (WPI, EZ-155). We recommend the SomnoFlo starter kit that includes an induction chamber and tubing (Kent Scientific, SF-MSEKIT) or ketamine/xylazine (Sigma-Aldrich, K113–10ML) as an alternate anesthetic.

7. Analgesics such as bupivacaine, buprenorphine and Tylenol. Administer analgesics in accordance with your institution’s IACUC policy, and consult an OAR/IACUC veterinarian.

3. Methods

3.1. Installing a Laser Ablation Capability on Your Multiphoton Microscope

All laser-scanning multiphoton microscopes have a built-in device that translates the laser beam along the x and y dimensions of the microscope field of view (FOV) (termed xy-scan), thereby generating a 2D image (see Fig. 1). This device can be a combination of two galvo mirrors, a resonant galvo and galvo mirror, or a scanning polygon and a galvo mirror (we use the latter on our home-built microscope). However, most commercial 2P imaging systems have a galvo–galvo scanner implemented. The same scanning device can be used to translate the ablation beam across the field of view (FOV), achieved by co-propagating the imaging beam and the ablation beam with the help of a dichroic mirror (DC; see Fig. 1).

Fig. 1.

Co-propagation of ablation and imaging beams in the multiphoton microscope. The ablation beam is overlapped with the imaging beam using a dichroic mirror (DC). Both beams are steered across the microscope field of view using scanning optics. M mirror, L lens, IP image plane, im. IP intermediate IP. Blue arrows mark the entrance of the scanning optics and back aperture of the objective lens, respectively

3.1.1. Installation of the Ablation Laser Unit

1. Fix the ablation laser to the breadboard of the microscope table using screws of the appropriate dimensions (typically M6 screws).

2. Turn on the ablation laser, and measure its output power to ensure that it is functioning as advertised. Reduce its power to ~10 mW, and proceed with the alignment. As a safety precaution, laser alignment should “never” be performed at full power. Additionally, we advise the operator to wear appropriate laser safety glasses as the operating wavelength of the ablation laser (1030 nm) is beyond the visible spectrum and therefore cannot be seen by the human eye. It can cause severe damage to the retina if operated inappropriately.

3. Using the alignment disk, visualize the laser propagation axis. Mark its height with respect to the table surface using an iris, adjusting the iris height so that the beam passes through the center of the iris. Install mirror M1 and M2 (see Fig. 2) by placing them at the correct height and at a 45° angle with respect to the propagation direction of the laser beam. Ensure that the height of the laser beam remains constant as it propagates from M1 to M2 and beyond using the iris. Never place an optical element such as a lens or prism at the position of M1. These elements have a reflective surface that is perpendicular to the propagation direction of the laser and can reflect part of the beam back into the laser head, damaging its function.

4. Install lens L1 (focal length 75 mm) into the beam path, ensuring that it is perpendicular to the propagation direction and that the beam passes through its center. For this, install short cage rods (Thorlabs, ER1) into the lens mount, and use the alignment disk for positioning.

5. Place L2 (focal length 75 mm) onto a mount, and fasten it to a translation stage, placing it roughly at a 150-mm distance from L1. Use the translation stage to fine-tune the distance between L1 and L2. Proper positioning is ascertained by testing the collimation of the beam that exits L2 (see Note 3). If the beam is divergent, the distance between the lenses needs to be increased. If it is convergent, the distance needs to be decreased.

6. Install the 950-nm short-pass dichroic (DC) while ensuring that both laser beams have a 45° angle with respect to it.

7. Overlap the imaging with the ablation beam by adjusting M3 and M4 as follows: use M3 to overlap the two beams at the dichroic. Use M4 to overlap the beams before entry to the scanning optics. Repeat iteratively until the two beams overlap completely at all positions between these two points.

8. Unscrew the objective lens, and ensure that the two beams overlap at its back aperture, tweaking the alignment of M3 and M4 if necessary. Measure the transmission of the imaging beam through the microscope by measuring its power at the scanning optics and at the back aperture (blue arrows in Fig. 1). Compare its transmission to that of the ablation beam to further confirm proper alignment (see Note 7).

3.1.2. Installation of Flushing System

To perform plasma-mediated laser ablation of bone using a water immersion objective lens, it is necessary to flow a steady stream of PBS, saline, or water across the sample, thereby removing any gas and debris that is created during the procedure. Install the flushing system by proceeding as follows:

1. Install tubing on your peristaltic pump, and attach a Luer Lock to Hose Adapter at one end of each hose, screwing the blunt-tip needles into place (see Fig. 3). Place the other end of the hose into a Falcon tube filled with PBS (see Fig. 3). Turn on the peristaltic pump, and verify flow from the needle on channel 1, as well as suction into the needle on channel 2 (see Fig. 3).

2. Take the neoprene rubber slab, and cut out a circle with the correct dimensions, fitting it around the objective lens (OL; see Fig. 4). Advance a sharp-edged needle (same gauge as blunt) along the cone of the OL, and slowly stencil a hole into the rubber slab (see Fig. 4). Stencil a second hole directly opposite, and thread the two blunt-tip needles through the respective openings, positioning them as shown (see Fig. 4). Turn on the peristaltic pump, setting channel 1 (efflux) to 62 rpm and channel 2 (influx) to 75 rpm.

3. Position an excised piece of bone in the image plane of the OL, adjusting the flow rate of influx/efflux to achieve a liquid cone with the correct dimensions (see Fig. 5).

4. Use your imaging software to perform laser ablation by taking a z-stack and adjusting the pulse energy of the ablation laser to ~12 nJ in the image plane. Set the step size along z to 0.25 μm, frame average to 10, and the number of z-slices according to the ablation depth that you would like to achieve, i.e., for an ablation depth of 20 μm, you would need 20 μm/0.25 μm = 80 z-slices. Your ablation geometry should be smooth (see Note 6).

5. To obtain an arbitrary ablation shape, adjust the scan geometry of your galvo mirrors in the microscope imaging software. Alternatively, if your system does not provide this function (see Note 5), insert a mask in the intermediate image plane (im. IP in Figs. 1 and 2). Optimal dimensions will vary depending on the magnification of your microscope, and it could be necessary to custom-design your mask (see Note 9).

Fig. 5.

Diagram showing correct flow rate. When influx and efflux are matched, a PBS cone with the correct dimensions is observed (left). If influx is too high (middle), PBS overflows the sample and dripping is observed. If influx is too low, the cone is too small, and it is impossible to generate a plasma due to the air interface (right)

3.2. Laser Bone Thinning to Improve Image Quality

Intravital microscopy of the skull BM is typically limited to depths of ~100–150 μm due to strong scattering and absorption of UV/Vis and NIR wavelengths by bone and BM tissue. Bone thinning presents one method for improving imaging depth. However, mechanical thinning produces rough surface quality along with low precision and proclivity for thermal damage. Laser thinning of bone has been shown to improve imaging depth, while only minimally damaging the surrounding tissue, and results in flat bone surfaces that maintain high image quality [18, 21].

3.2.1. Perform Skin Flap Surgery

Place the mouse into an induction chamber, start the oxygen flow, and set the isoflurane to 3%. Alternatively, you can use ketamine/xylazine at a dose of 100 mg/kg (IP) to anesthetize the mouse. Administer analgesics in accordance with your institution’s animal care and use committee’s (IACUC) policy. Sterile equipment should be used throughout the procedure:

1. Set the isoflurane to 2%, and place the mouse into a mouse holder, ensuring that the skull is securely fixed (see Note 10).

2. Shave the mouse’s head hair using a blade or scissors, then disinfect it first with an alcohol and then an iodine wipe. Using a pair of tweezers, pull the mouse skin away from the skull, and cut a half-elliptic shape with dimensions of ~3 × 5 mm (see Fig. 6a), making sure to point the scissor tips upward to avoid damaging the skull or eyes.

3. Fold the skin flap toward the back of the skull, and gently secure it in place with a drop of antibiotic ointment. Using a sterile transfer pipette, place a drop of sterile saline onto the mouse skull, and gently remove the periosteum using two Q-tips, drawing the periosteum back from the skull center in a unified motion (see Fig. 6b). Use an additional Q-tip, and rub it across the mouse skull while applying a good amount of pressure, thereby removing any further periosteal tissue (see Note 11).

4. Add a fresh drop of sterile saline onto the skull, and set the isoflurane to 1.5%. Ensure that the mouse is breathing at a rate of õnce every 1–2 s, adjusting the isoflurane if necessary. If using ketamine/xylazine, re-inject the mouse every ~40 min with 100 mg/kg (IP) to maintain anesthesia. Transfer the mouse to the multiphoton microscope.

Fig. 6.

Skin flap surgery for in vivo imaging. (a) Incision site for skin flap surgery. After folding the skin flap backward (b), moving two Q-tips from the skull center outward in a concerted motion (arrows) leads to removal of the periosteum

3.2.2. Perform In Vivo Multiphoton Microscopy

1. Turn on the microscope’s photomultiplier tubes, set your imaging laser to the correct wavelength for two-photon excitation, and adjust its power to ~40 mW (see Note 12). You should also ensure that the correct set of dichroic mirrors and filters is in place in order to visualize your fluorophores (see Note 13).

2. Bring the mouse skull into focus by adjusting either the mouse or the microscope along the z dimension (see Note 14). Record the x, y, and z position of lambda, bregma, and bifurcation on the mouse skull (see Fig. 7). These anatomical landmarks are most readily visualized by the second harmonic generation of bone (generated by your imaging laser or ablation laser at half the emission wavelength). Recording the position of your in vivo images with regard to these landmarks will allow you to compare results from different mice, as well as to identify sites from previous experiments when doing longitudinal imaging experiments.

3. Translate the mouse/scope along the x/y/z dimension, and identify sites of interest for intravital imaging. Record a z-stack by setting the step size to between 1 and 5 μm and averaging at least ten imaging frames. You can record z-stacks in multiple, adjacent locations that can later be stitched together to generate a larger, tiled image.

Fig. 7.

Mouse skull with anatomical landmarks. Position of bifurcation, bregma, and lambda shown by arrows

3.2.3. Perform Bone Thinning

To perform bone thinning during intravital imaging, turn on the flushing system, and set the ablation pulse energy to ~12 nJ, the step size along z to 0.25 μm and the frame average to 10. You can thin multiple, adjacent regions and stitch individual microscope fields of view to create tiled, high-depth images of larger bone marrow regions.

3.3. Generation of Bone Defects

Laser ablation is not only a precise method for thinning bone tissue and improving imaging depth and resolution; it can also be used to generate small defects in the bone and track the injury response by in vivo imaging [21]. In combination with fluorescent reporter mice for specific cell populations, this can provide important insights into wound healing and immune function.

3.3.1. Perform Skin Flap Surgery, Multiphoton Microscopy, and Laser Injury

Perform skin flap surgery and in vivo imaging as detailed in Subheadings 3.2.1 and 3.2.2, respectively:

1. Inject a vascular contrast agent such as Qtracker prior to imaging. This can be done either retro-orbitally or by tail vein injection.

2. If applicable, insert an iris/mask into the intermediate image plane of the microscope. Set the ablation pulse energy to ~12 nJ, the step size along z to 0.25 μm, and the frame average to 10.

3. Ablate a cone-shaped crater (see Note 8), and gently perforate the BM at its tip, taking care to avoid rupturing underlying blood vessels. Slowly close the iris as the sample/scope moves along the z dimension to further minimize collateral damage (see Ref. [21] for details).

3.3.2. Survival Surgery

1. Once you have completed the imaging and injury generation procedure, remove the mouse from the microscope, and gently rinse the calvarium with ~10 mL of sterile saline.

2. Fold the skin flap back onto the skull, and suture it in place using an Ethilon suture. The spacing between individual sutures should be no more than 2 mm. Apply triple antibiotic ointment onto the freshly sutured skin and administer analgesics in accordance with your institution’s IACUC policy. Allow the mouse to recover from anesthesia.

3.3.3. Follow-Up Imaging

The procedure for follow-up imaging is the same as in Subheadings 3.2.1 and 3.2.2, although it can become necessary to rub the skull vigorously after the skin flap surgery to remove small amounts of scar tissue that has begun forming ~1 week after the first imaging session.

4. Notes

1. Using a low repetition rate laser. We recommend operating the laser at a repetition rate of 1 MHz for two reasons:

   1. At relatively low repetition rates such as 1 MHz (compared to the maximum possible rep. rate of 40 MHz for Satsuma laser), the average power of the laser will leave less thermal damage behind as opposed to higher laser rep. rates.
   2. The interval between two laser pulses is long enough for a plasma-induced cavitation bubble to collapse before the next laser pulse hits the sample, thereby maintaining a high ablation efficiency.

2. Changing the beam diameter. Depending on the objective lens that is used, the diameter of the ablation beam will need to be adjusted to fill out its back aperture. This can be done by installing two lenses in a 4f arrangement. This leads to a beam magnification of M = f₂/f₁, where f₁ is the focal length of L1 and f₂ is the focal length of L2, respectively.

3. Measuring collimation. One method for testing collimation is by using a commercially available laser beam profiler. These can be quite expensive, however, and we have found that a simpler method can also be used and delivers high-quality ablation geometries. Insert the alignment disk into your laser beam, and use a pencil to mark its circumference and beam center. Install a mirror at this position, and shoot the laser beam along the length of your optical table. If possible, extend the distance to the nearest wall. Move the alignment disk along the newly generated beam path (which is, ideally, several meters in length), and check to make sure that the beam doesn’t focus at any point between the mirror and the wall. Compare the geometry of the laser beam at the wall to the markings on the alignment disk. If it is smaller, the beam is convergent, and the distance between the two lenses needs to be increased. If it is larger, the beam is divergent, and the distance needs to be reduced.

4. Choice of dichroic mirror. Depending on the existing imaging beam path set by the 2P microscope, the implementation of an ablation arm can be realized in two ways:

   1. The ablation beam is merged with the imaging beam by a short-pass dichroic mirror (see Fig. 2).
   2. The ablation beam is merged with the imaging beam by a long-pass dichroic mirror (DM). In this case, the imaging beam is deflected into the 2P microscope, while the ablation beam passes through the DM and enters the microscope scanning optics. In essence, this would correspond to switching ablation laser and imaging laser in Fig. 2.

5. Aperture in intermediate plane. Most commercial 2P microscopes have a programmable scanning area. Therefore, these types of systems do not need an aperture in the intermediate image plane. However, for the sake of completeness, an aperture has been included in this protocol in case your system does not have a programmable galvo–galvo scanner.

6. Troubleshooting: The ablation geometry is highly irregular. If the ablation beam is well aligned, the ablation quality should be good. However, here are some common reasons why the ablation has poor quality:

   1. Alignment needs to be tweaked. Move the sample below the image plane, and turn on the flushing system and ablation beam. The emission of the ablation plasma should be visible (in our images this looks like tiny bubbles or speckles) and should fill the entire FOV (see Fig. 8). If not, adjust the tilt angles of M4 to optimize the ablation plasma.
   2. The flushing needle is too far from the ablation plasma, and therefore the flow rate does not suffice to remove gas/debris. This becomes noticeable because large bubbles form during the ablation procedure, and they scatter the ablation beam, thereby preventing plasma generation. Move the needle closer to the image plane, advancing it along the cone of the OL, while paying attention not to block the imaging/ablation beams with the needle.
   3. An ablation step of 0.25 μm with ten passes per plane (ten-frame average) yields optimal results on our system. You may need to adjust these values if the scanning optics on your system or the NA of your objective lens are different (our NA is 1.0).

7. Troubleshooting the transmission efficiency of the ablation beam. If the ablation and imaging beams are well overlapped, their transmission efficiency should be similar, unless the beam diameters differ significantly. You can check the specs of the imaging laser to determine its diameter or measure it using an inexpensive CMOS camera (e.g., Zelux monochrome (#CS165MU, available from Thorlabs). If necessary, adjust the diameter of the ablation beam by installing two lenses with proper magnification.

8. Ablation crater shape. The shape of the ablation crater is determined by the focusing geometry of the objective lens. Ideally, the angle of the crater wall takes the value of the focusing angle of the objective lens (given by its NA).

9. Determining the magnification. The fastest and most reliable way of doing this is by inserting an aperture of known geometry into the intermediate image plane. The same geometry should be visible in the multiphoton images, and calculation of the ratio between the two sizes will yield the magnification (see Fig. 9).

10. Mouse holder. We use a 3D-printed mouse holder for our imaging and ablation experiments. The SolidWorks file is available upon request. Discussing its design and fabrication is beyond the scope of this chapter. Please consult the in vivo imaging chapter in this book for information on alternate mouse holders.

11. Bleeding. Slight bleeding is often observed when removing the periosteum and can reduce image and ablation quality. The bleeding will generally stop if you put a fresh drop of saline onto the mouse skull and wait for a few minutes. Once bleeding has stopped, use fresh Q-tips to gently remove coagulated blood, and add a fresh drop of saline to the skull.

12. Two-photon excitation frequency and laser power. We recommend using one of the following resources to determine the two-photon excitation frequency of your fluorophore:

    1. https://www.bioimaging.bmc.med.uni-muenchen.de/manuals-protocols/multiphoton/index.html
    2. https://www.drbio.cornell.edu/cross_sections.html
    3. https://www.thermofisher.com/us/en/home/references/molecular-probes-the-handbook/technical-notes-and-product-highlights/fluorescent-probes-for-two-photon-microscopy.html

    The optimal power for in vivo multiphoton microscopy depends on the repetition rate of your laser. For a repetition rate of 80 MHz, we recommend using ~40 mW for bone marrow imaging. This value is obtained by measuring the laser power in the image plane. Note that at the bone surface lower powers are recommended.

13. Dichroic mirrors and filters. The optimal set of dichroics and filters will depend on the different fluorophores you are planning to excite. We recommend using Thermo Fisher’s Fluorescence SpectraViewer [22] to help guide your choices since the emission spectra for one- and two-photon excitation are typically similar.

14. X, Y, and Z positioning. Depending on your system, this will entail either moving the microscope or the sample (mouse) along the x, y, and z dimensions.

Fig. 8.

Ablation plasma. Image of the ablation plasma, which fills the entire imaging FOV, if correctly aligned. The plasma emission is visible as tiny specks or bubbles

Fig. 9.

Insertion of an aperture into the intermediate image. Comparing to the aperture dimensions in the image plane of your microscope makes it possible to calculate your microscope’s magnification