In this study, second harmonic generation (SHG) and third harmonic generation (THG) spectroscopic imaging were performed on biological samples using a femtosecond laser source in the third near-infrared (NIR) optical window (NIR-III). Using a visible-NIR spectrometer, the SHG and THG signals were simultaneously detected and were extracted using spectral analysis. Visualization of biological samples such as cultured cells (HEK293 T), mouse brain slices, and the nematode Caenorhabditis elegans was performed in a label-free manner. In particular, in an SHG image of an entire coronal brain section (8 × 6 mm2), we observed mesh-like and filamentous structures in the arachnoid mater and wall of the cerebral ventricle, probably corresponding to the collagen fibers, cilia, and rootlet. Moreover, the THG images clearly depicted the densely packed axons in the white matter and cell nuclei at the cortex of the mouse brain slice sample and lipid-rich granules such as lipid droplets inside the nematode. The observations and conclusions drawn from this technique confirm that it can be utilized for various biological applications, including in vivo label-free imaging of living animals.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
Second harmonic generation (SHG) is a second-order nonlinear optical process that was discovered by Franken et al.  approximately 60 years ago. Following the development of short-pulse laser sources, SHG has been applied widely as an effective method of wavelength conversion in ultrafast science. Shortly after its application in wavelength conversion, SHG emerged in biological research fields such as SHG microscopy [2–4]. SHG activities due to biological harmonophores have been found in biomolecules and their assemblies in living cells and tissues and have been compiled in textbooks . These well-known harmonophores include biomolecules such as collagen (e.g., that found in mouse tail tendons) [3–5], muscle myosin , special types of tubulins such as spindle fibers and nerve axons , and spider silk . In terms of protein filaments, collagen, myosin, tubulin, and fibroin have been studied widely and intensively . However, despite these widespread studies, SHG imaging has been applied to limited harmonophores thus far [8–10]. Recently, we previously reported that ciliary rootlets are SHG-active owing to their non-centrosymmetric structures . One of the characteristic features of these rootlets is that the SHG signal intensity is approximately 1/1000 times weaker than that of collagen at the sclera. Inspired by this observation, we believe that other biologically intrinsic asymmetric structures with weak SHG activity exist as well. In order to explore these weak SHG signals in non-transparent tissues, however, it is desired to use laser pulses with high peak power and large penetration depths into the samples.
As optical windows are suitable for deep tissue imaging, the NIR-I (650–950 nm), NIR-II (1000–1350 nm), and NIR-III (1550–1850nm) optical windows [11–15] and various types of laser sources have been developed in these zones. To date, the direct outputs of mode-locked femtosecond laser sources such as Ti:sapphire , Cr:forsterite [16,17], and Yb3+-doped fiber laser sources [18–21] have played important roles in developing an excitation light source for SHG microscopy. Although these laser sources have exhibited satisfactory performance and proven themselves to be useful, these typical laser systems are bulky, and their emission wavelengths are restricted to the range of 700–1350 nm, corresponding to the NIR-I and NIR-II optical windows. In contrast, the longer attenuation length in the NIR-III window  indicates the great potential of this optical window. In addition, a laser wavelength of 800 or 1064 nm should be avoided while performing simultaneous detection of third harmonic generation (THG) signals and SHG signals as the wavelength of the THG signal corresponds to 266 or 355 nm. Therefore, it may be absorbed by tissue samples or microscope objectives. Furthermore, two- or three-photon excitation processes can cause unwanted auto-fluorescence.
Recently, laser sources in the NIR-III optical window were utilized to develop SHG, THG, and multiphoton excitation fluorescence microscopy [11–15]. As its excitation wavelength (1550 nm) is longer compared with those of the NIR-I and NIR-II windows, the microscopic system offers great advantages in increasing the penetration depth, owing to reduced scattering loss [22–24] and photodamage suppression. However, despite these advantages, the biological applications of NIR-III are focused primarily on well-known targets such as collagen and myosin [13,25,26].
In this study, we developed an SHG and THG microscope using a compact Er3+-doped fiber laser oscillator. As demonstrated in the pioneering works [13–15], a laser wavelength of 1550 nm provided SHG and THG signals at 775 and 517 nm, respectively, which are perfectly suitable for standard optics and detectors with near-maximum efficiency. We built the spectroscopic SHG and THG microscopy systems by combining a microscope with a spectrometer. This modality was reported in our previous reports on sub-nanosecond  and 100 ps  laser sources for CARS microspectroscopic setups as well. We simultaneously detected SHG and THG signals in the visible and NIR regions using an ultraviolet–visible–NIR spectrometer with a coverage range of 365–930 nm. Based on the spectroscopic analysis, we recognized clear SHG and THG signals and distinguished them from the background or baseline artifacts. Therefore, the proposed method exhibited great potential to reveal weak and buried SHG signals that were not clearly observed in living cells and tissues.
2.1 Spectroscopic SHG and THG imaging system
The experimental setup is illustrated in Fig. 1(a). The laser source was an Er3+-doped fiber laser source (CNT-1550-TK-A, KPHOTONICS, Az, USA). The output wavelength, repetition rate, typical pulse duration, and average power were 1550 nm, 40 MHz, 150 fs, and 100 mW, respectively. The laser pulses were guided by an optical fiber and collimated using a fiber collimator (F280APC-1550, Thorlabs Inc., NJ, USA). As the THG of the fundamental was observed at the output of the laser source, only the NIR-III spectral components were selected by an optical filter (IR80, Kenko-optics, Tokyo, Japan). The laser pulses were introduced into a custom-made inverted microscope (TE2000, Nikon Corporation, Tokyo, Japan) and were focused on the sample through a microscope objective (CFI Plan Apo 60x NA 1.27, water immersion, Nikon, Tokyo, Japan). The sample was placed on a motorized stage (Micro Stage (2-axis), Mad City Labs, WI, USA) with a maximum scanning range of 25 × 25 mm2 combined with a piezo stage (MCLS03542, Mad City Labs) in the axial direction. The visible and NIR signals were collected using a second objective lens (Plan S Fluor 40× NA 0.6, dry, Nikon, Tokyo, Japan). The respective SHG and THG signals at 775 and 517 nm were detected using a spectrometer (SpectraPro-300i, Princeton Instruments, NJ, USA) equipped with a charge-coupled device (CCD) camera (PIXIS 100 BR, Princeton Instruments, NJ, USA). The SHG and THG signals were extracted by fitting each spectral profile using a Gaussian function, and their respective amplitudes for each spatial position were mapped out.
2.2 Sample preparation
A starch granule is considered to be a model biological sample for SHG microscopy due to its helical amylose structure [28,29]. Therefore, we utilized a potato starch granule to evaluate the performance of the microscope. The potato starch was purchased from a local supermarket. The evaluation was conducted by placing dry starch granules without further purification between a cover glass and glass slide and then sealed with nail lacquer.
HEK293 T cells (RCB2202) were obtained from the RIKEN BRC Cell Bank. The cells were cultured in Dulbecco's modified Eagle’s medium (Wako, Japan) supplemented with 10% fetal bovine serum and penicillin/streptomycin at 37 °C in a humidified atmosphere of 5% CO2. The HEK293 T cells were grown to ∼80% confluency in a 35 mm glass bottom dish (Matsunami Glass, Japan). The culture dish was removed from the incubator immediately before measurement and recording was performed without changing the medium.
Coronal brain sections of mouse were used as samples for wide-area (8 × 6 mm2) imaging. A wild type (C57BL/6 N) male mouse at p118 was deeply anesthetized with 5% isoflurane and an i.p. administration of somnopentyl (32.4 mg/kg). Thereafter, the mouse was transcardially perfused with 0.1 M phosphate buffered saline (PBS) with heparin for 1 min, followed by being transcardially perfused with 4% paraformaldehyde for 10 min. The brain was removed and immersed in the same post-fix buffer at 4 °C for 24 h. The removed brain was washed three times with PBS (5 min each, shaking in a shaker) and subsequently sectioned into 50 µm thick sections using a vibratome (Leica). The resulting brain sections were used for label-free SHG and THG imaging.
Concerning depth-resolved measurement and immunostaining, we prepared the sample in the following manner. The protocol was the same as that described before removing the brain and also used a wild type (C57BL/6 N) male mouse at p84.The brains was then postfixed in 4% PFA at 4 °C overnight, cryoprotected in 30% sucrose in PBS for 2 days, embedded in OCT compound (Sakura Finetech), and stored at −80°C until use. The brain was cryosectioned coronally at thicknesses of 40 and 300 µm by Cryostat (Leica CM3050S).
These sliced sections were subsequently mounted on glass slides with Mowiol. The samples were stored at 4 °C for 24 h. Optical images were obtained using a stereoscopic microscope (Zeiss AXIO Zoom.V16, Carl Zeiss GmbH, Jena, Germany).
We used anti-neuronal nuclei (NeuN) antibody to visualize neuron cell nuclei and anti- glial fibrillary acidic protein (GFAP) antibody to visualize astrocyte. After sectioning the brain as described above, the 40-µm sections were rinsed with PBS and incubated in 0.4% Block Ace (Snow Brand Milk Products) in PBS with 0.1% Tween20 (0.1% PBST) for 2 h at room temperature. This procedure was followed by overnight incubation with mouse-NeuN antibodies (1:1000; Millipore MAB377) and rabbit anti-GFAP (1:1000; Dako # Z0334) antibodies in 0.2% Block Ace in 0.1% PBST overnight at 4 °C. The sections were rinsed with PBS and then incubated with AlexaFluor488-conjugated donkey anti-mouse IgG (1:1000; Thermo Fisher # A-10667) and AlexaFluor594-conjugated goat anti-rabbit IgG (1:1000; Thermo Fisher # A-11012) antibodies in 0.2% Block Ace in 0.1% PBST for 2 h at room temperature. The sections were rinsed with PBS and incubated with trihydrochloride (Hoechst 33342) (10 µg/ml; Thermo Fisher # H1399) for 10 min at room temperature, then rinsed again and mounted with Vectashield mounting medium (Vector Lab, #H-1700). Fluorescence imaging was conducted using an LSM 800 confocal microscope system (Carl Zeiss, Germany) and NanoZoomer XR (Hamamatsu Photonics, Japan).
Wild-type strain N2 of Caenorhabditis elegans was obtained from the Caenorhabditis Genetics Center (CGC) and was cultured as described previously  on agar plates containing a nematode growth medium with OP50 Escherichia coli at 20 °C. A gravid hermaphrodite was mounted on a fresh 2% agar pad with M9 buffer before the measurements.
All animal experiments were approved by the Institutional Animal Care and Use Committee of the University of Tsukuba. All animals were maintained in accordance with the institutional guidelines of the animal facilities of the Laboratory of Animal Resource Center.
3. Results and discussion
3.1 SHG and THG imaging of a starch granule
To evaluate the performance of the proposed system, we measured a starch granule. Figure 2 depicts the (a) conventional wide field, (b) SHG, and (c) THG images of the starch granule. The image and step sizes were 50 × 100 µm2 (51 × 101 pixels) and 1 µm/pixel, respectively. The exposure time at each spatial position was 50 ms. The laser power at the sample position was approximately 20 mW. The spectral profiles of the SHG and THG signals are displayed in Fig. 2(d). As described in Section 2.1, we fitted the spectral profiles using a Gaussian function to extract their amplitudes as the signal intensity. As demonstrated in Fig. 2(b), the SHG image displays high image contrast in some granule areas. The spectral profile of the SHG signal at the cross of position (i) in Fig. 2(b) is shown in Fig. 2(d), indicating multiple peaks. The spectral feature of the SHG signal is reflected by that of the fundamental laser spectrum (inset in Fig. 1). On the other hand, a strong THG signal is observed outside the granule. The spectral profiles of the THG signal at the crosses of positions (ii) and (iii) in Fig. 2(b) are shown in Fig. 2(d). The THG signal originates from the interface between the air and underlying cover glass, providing a uniform contrast except in the granule area . The THG image visualized the granule, whose peripheral area gave relatively high image contrast as well, which could be due to the refractive index difference between the granule and surrounding air. Similar results were also obtained by using a multimodal nonlinear optical microscope with a 1064-nm laser source , which is shown in Fig. S1 in the Supplement 1.
We evaluated the limit of detection (LOD) using the incident laser power dependence of the SHG signal of a starch granule. The details are described in the Supporting Materials (Fig. S2). With decreasing incident laser power, the SHG signal firstly showed quadratic dependence on the incident laser power, followed by linear dependence, as shown in Fig. S2. Based on this result, the LOD was estimated to be roughly in the order of 10 CCD counts in the present study.
3.2 SHG and THG imaging of cultured cells
For biological applications, we firstly visualized the cultured cells (HEK293 T). Figures 3(a) and (b) display the SHG and THG images, respectively. The image and step sizes were 100 × 100 µm2 (201 × 201 pixels) and 0.5 µm/pixel, respectively. The exposure time at each pixel was 100 ms. We found clear bright spots in the SHG image, whereas the THG image displayed several large dark and round areas. These dark and round areas were predicted by cell nuclei , which are encircled by solid lines (Fig. 3(c)). The merged image of the SHG and THG signals (Fig. 3(c)) indicates that the SHG spots were likely to coexist with the cell nuclei. In our previous study , we observed similar SHG spots for cultured cells (COS7) and assigned them as Rootletin filaments that are the linker proteins of centrioles [10,34,35]. Therefore, it was determined that the SHG spots corresponded to Rootletin . Although the SHG spots are weakly observed, the signal is clearly evident in the spectral domain, which is manifested in Fig. 3(d).
We evaluated the upper limit of the lateral spatial resolution using the following procedures. By fitting the spatial profile of the small SHG spots (i) observed in Fig. 3(a) using a Gaussian function, the upper limit of the lateral spatial resolution of SHG imaging was estimated to be 0.8 ± 0.1 µm. Similarly, using the sharp edge indicated by the red line in Fig. 3(b), the upper limit of the lateral spatial resolution of THG imaging was estimated to be 0.8 ± 0.1 µm. Concerning the axial spatial resolution, we evaluated the depth-resolved THG signal at the interface between the glass and air and estimated 1.9 ± 0.1 µm to be the upper limit of the axial spatial resolution for both THG and SHG signals. These values are larger than the numerical simulation results, which are 0.5 (SHG) and 0.4 (THG) µm for the lateral direction and 1.1 (SHG) and 0.9 (THG) µm for the axial direction, probably because the microscope objective was not optimized for 1.55 µm.
3.3 SHG and THG imaging of a mouse brain slice
We performed wide-field SHG and THG imaging to visualize the entire mouse brain slice sample. Figure 4 presents the obtained macroscopic images for the label-free (no staining) (a) and immunostaining (b, c) samples. The macroscopic fluorescent images obtained by Hoechst, GFAP, and NeuN and the correspondences of the regions in (i)–(iv) are provided in Fig. S4 in the Supplement 1. The image and step size used to acquire the SHG and THG images in Fig. 4(a) were 8 × 6 mm2 (801 × 601 pixels) and 10 µm/pixel, respectively. The exposure time for each pixel was 200 ms. The inset in Fig. 4(a) illustrates a bright-field image of the entire region of the slice sample. We state again for emphasis that the SHG and THG images were obtained simultaneously. The SHG image in Fig. 4(a) (left) displays high contrast at the meninges on the brain surface (indicated by (i)) and the meninges between the hippocampus and thalamus (indicated by triangles). These results were reproduced well using the immunostaining sample, as shown in Fig. 4(b), where the image and step sizes for the SHG image were 4.1 × 6.4 mm2 (411 × 641 pixels) and 10 µm/pixel, respectively. The exposure time for each pixel was 200 ms. It should be noted that the SHG image in Fig. 4(a) exhibits medium contrast in the areas indicated by (ii) and (iii), corresponding to the cerebral ventricle and cerebral white matter, which will be discussed later with the microscopic images. On the other hand, the THG image in Fig. 4(a) (right) displays image contrast different from that of the bright-field image (inset in Fig. 4(a)), indicating a characteristic contrast enhancement. One of the prominent features is the high contrast in the cerebral white matter, as indicated by area (iii) and two spots denoted by magenta triangles. It should be noted that the THG spots indicated by the magenta triangles were not observed in the bright-field image (inset in Fig. 4(a)). In the cerebral white matter, axons are tightly bundled. The areas indicated by the two triangles contain bundled axons oriented perpendicular to the THG image as well. As both the areas indicated by the cerebral white matter (iii) and the triangles were composed of the same organelles (bundled axons), we assumed that the THG image contrast originated from the densely packed axons that provide a refractive index different from those of the peripheral areas. Moreover, at the cortex indicated by region (iv), the microscopic THG image showed a large number of spots. It will be discussed later.
In the SHG images, several structures extending vertically from the brain surface are evident. One of them is observed in Fig. 4(c), which corresponds to the box indicated by (v) in the merged image (right) in Fig. 4(b). In Fig. 4(c) (middle), GFAP, a marker for astrocytes, is visualized. Astrocytes are known to form the blood–brain barrier, and the GFAP-positive areas in Fig. 4(c) indicate astrocyte-covered blood vessels. In the merged image in Fig. 4(c), the straightened structure with high contrast in SHG is consistent with GFAP, corresponding to a blood vessel.
Next, we performed microscopic imaging of the brain slice samples. Figure 5 summarizes the images obtained around the brain surface for the label-free (no staining) (a) and immunostaining (b) samples. The SHG images exhibit mesh-like structures. In Fig. 5(b), the SHG imaging results are reproduced by the immunostaining sample. In particular, the merged image (SHG, THG, and GFAP) and zoomed-in view in Fig. 5(b) reveal that the SHG signal extends outside the astrocyte end-feet (indicated by GFAP). In a previous study, SHG was reported to visualize the pia arachnoid complex . However, the position of the pia mater is found just above the astrocyte end-feet, indicating that the mesh-like structures are not pia mater, but rather subarachnoid space and/or arachnoid mater. Since Collagen-1a1-positive inner arachnoid fibroblasts  exists at this area, we consider the collagen fibril structure produced by fibroblasts  are visualized by SHG. The mesh-like structures were partly visualized in the THG image in Fig. 5(a). This tendency is also clearly observed in the merged image in Fig. 5(a); the mesh-like structures in both the SHG and THG images overlap well. The collagen fibers in arachnoid mater additionally exist at the meninges between the hippocampus and thalamus  (indicated by the white triangles in Fig. 4(a) (left)), at which the SHG signal is observed.
Moreover, bright THG spots were sparsely observed in the cortex section that were previously reported and assigned as cell nuclei . The cell nuclei were more clearly observed in area (iv), which corresponded to the deeper position of the cortex where cell bodies densely exist. Therefore, our results are compatible to those obtained using previously reported THG images .
Secondly, Fig. 6 depicts the microscopic imaging results around the cerebral ventricle indicated by (ii) in Fig. 4(a) for the label-free (no staining) (a) and immunostaining (b) samples. The SHG image in Fig. 6(a) reveals filamentous structures (i) and underlying straightened structure (ii), which are indicated by arrows in the merged image. The spectral profiles of the SHG signals at filamentous structures are shown in Fig. S3(b). The THG image indicated by (ii) in Fig. 4(a) displays the image contrast at the ventricle walls. Therefore, as depicted in the merged image in Fig. 6(a), these structures in the SHG image are on the ventricle walls.
The overall tendency is reproduced well by the results obtained using the immunostaining sample, and the SHG image in Fig. 6(b) displays high contrast on the ventricle walls. As clearly shown in the zoomed-in views of the merged (SHG (yellow) and Hoechst (cyan)) images, the straightened structure in area (ii) is observed on the ventricular side of each nucleus of the ventricular cells. The filamentous structures indicated by (i) spread out on the ventricular side from area (ii). According to the literature , harmonophores can be assigned to the cilia lining the inner walls of the ventricles. However, it is known that ependymal cells, underneath the cilia, have SHG-active Rootletin , a component of the rootlet that supports cilia. Moreover, the rootlet appears to be wrapped around the nuclei in ependymal cells . Therefore, we believe that not only the cilia (i), but also the rootlets (ii) of the ependymal cells are observed in the SHG image in Fig. 6.
Thirdly, we investigated the area (iii) in Fig. 4(a) in more detail for the label-free (no staining) (a) and immunostaining (b) samples in Fig. 7. The microscopic THG image in area (iii) depicts the cerebral white matter, where axon bundles exist provides image contrast due to the spatial heterogeneity of the white matter. In addition, the microscopic SHG image in this area reveals numerous distinct SHG spots. The spectral profiles of the SHG signals at several spots are shown in Fig. S3(d). Although the axonal microtubules are SHG-active, their morphologies are not filamentous. Moreover, fixation is known to lead to a significant loss of SHG signal . Therefore, the SHG spots in area (iii) could not be ascribed to the axonal microtubules. This finding is validated by the results obtained using the immunostaining sample, as shown in Fig. 7(b). SHG spots are not observed for the immunostaining sample, probably because the immunostaining procedures makes the organelles SHG-inactive. Based on the immunostaining results, the cell nuclei (Hoechst) are concentrated in area (iii), where dark THG spots are observed in Fig. 7(b) (bottom row left). On the other hand, the SHG image shows bright spots in label-free samples as Fig. 7(a) (combined). Because neuronal cell bodies at this site were confirmed to be absent by NeuN immunostaining, these nuclei can probably be assigned as derived neuroglial support cells such as oligodendroglia and astrocytes . The presence of astrocyte cell bodies in this region is also confirmed by GFAP in Fig. 7(b) (bottom row right). Therefore, we determined that the SHG spots originated from organelles in the neuroglial support cells. Similar spots are also observed at the periventricular (indicated by the arrow) in the SHG image in Fig. 6(a), as confirmed by the spectral profile in Fig. S3(a).
Finally, we will focus on the area of the cortex (iv). The results are summarized in Fig. 8. The macroscopic SHG image indicated by area (iv) in Fig. 4(a) provided negligible visual evidence of the existence of a signal. However, the microscopic SHG image in this area (Fig. 8(a)) exhibits weak but distinct spots. These weak SHG spots are also confirmed by the spectral profiles of the SHG signal in corresponding areas, as shown in Fig. S3(f). At the cortex, numerous cell bodies exist. Area (iv) in the THG image clearly exhibits bright spots, as discussed regarding area (i). These spots originated from the cell nuclei in the cortex . The SHG and THG signals were combined to investigate the origin of the SHG signal in Fig. 8(a), and the immunostaining sample was used to confirm the location of the neuronal nuclei in Fig. 8(b). The label-free sample shows that some of the SHG spots appear close to the cell nuclei. In addition, the immunostaining sample demonstrates that most of the SHG spots are in contact with the NeuN. Therefore, the organelles close to the cell nuclei, such as centrioles, were among the candidates for the SHG source. The centrioles, which are components of the centrosome, are connected to each other by a linker protein called Rootletin [34,35] that is SHG-active . In addition, similar small SHG spots are observed in cultured cells (HEK293) with the same experimental setup (Fig. 3).
To evaluate the deep penetration length obtained with 1.55 µm excitation, we performed depth (z)-resolved SHG and THG imaging. Figure 9 presents the z-stacking results for the SHG and THG images of the cortex with a 10 µm difference in depth position, as well as the spectral profiles. The sample thickness was 300 µm. The results demonstrate that image contrast is not significantly degraded at deeper positions. As shown at the upper images in Fig. 9(a), the SHG signal also displays vascular structures due to collagen. Moreover, the lower images exhibit weak SHG spots, as discussed in regard to Fig. 8. These structures are also confirmed by the spectral profiles shown in Fig. 9(c). On the other hand, the THG images in Fig. 9(a) depict cell nuclei, whose contrast is different at different depth positions.
3.4 In vivo SHG and THG imaging of Caenorhabditis elegans
Next, we performed in vivo SHG and THG imaging of living nematode Caenorhabditis elegans, as they are frequently chosen as model animals. Figure 10(a) provides an optical image of the worm, and Figs. 10(b) and (c) display the corresponding SHG and THG images, respectively. The exposure time at each spatial position was 50 ms. The image and step sizes were 400 × 100 µm2 and 1 µm/pixel, respectively. The SHG image in Fig. 10(b) displayed a high image contrast around the pharyngeal and body-wall muscles. These tissues express high levels of muscle myosin, which is known to be SHG-active. This finding is consistent with the literature [44,45]. In contrast, the THG image in Fig. 10(c) displays numerous bright spots in the intestinal cells, as indicated by arrow A. According to the literature , several types of granules, such as lipid droplets, yolks, endosomes, and other granules, are plausible candidates for the THG spots.
In conclusion, we developed a label-free bio-imaging platform using a femtosecond laser source in NIR-III (1550 nm). We validated the usefulness and efficiency of the system by measuring cultured cells, mouse brain slices, and a living nematode. In particular, the SHG image of the brain slice displayed unique microscopic structures such as mesh-like structures due to collagen in the arachnoid mater and meninges, filamentous structures due to cilia and rootlet at the cerebral ventricle, and spot-like patterns that we assumed were due to neuroglial support cells in the white matter. Some of the features observed in label-free SHG and THG images are validated by comparing them with immunostaining images. Owing to the large penetration depth of the proposed imaging platform due to deep NIR excitation, this technique will pave the way for in vivo label-free imaging of living animals.
JSPS KAKENHI (JP21H04961 Grant-in-Aid for Scientific Research [A]).
The authors thank Mr. S. Ishibashi, Mr. T. Maruyama, Mr. N. O. Grooms, and Ms. A. Imamura for their helpful discussions and measurement of the starch sample. The authors gratefully acknowledge Dr. M. Nuriya, Keio University, for the fruitful discussions. Nematode strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). We would like to thank Editage (www.editage.com) for English language editing.
The authors declare no conflicts of interest.
The data that support the findings of this study can be obtained from the corresponding author upon reasonable request.
See Supplement 1 for supporting content.
1. P. A. Franken, G. Weinreich, C. W. Peters, and A. E. Hil, “Generation of optical harmonics,” Phys. Rev. Lett. 7(4), 118–119 (1961). [CrossRef]
2. C. Sheppard, J. Gannaway, R. Kompfner, and D. Walsh, “The scanning harmonic optical microscope,” IEEE J. Quantum Electron. 13(9), 912 (1977). [CrossRef]
3. I. Freund and M. Deutsch, “Second-harmonic microscopy of biological tissue,” Opt. Lett. 11(2), 94–96 (1986). [CrossRef]
4. F. S. Pavone and P. J. Campagnola, Second Harmonic Generation Imaging (CRC Press, 2013).
5. S. Fine and W. P. Hansen, “Optical second harmonic generation in biological systems,” Appl. Opt. 10(10), 2350–2353 (1971). [CrossRef]
6. V. Nucciotti, C. Stringari, L. Sacconi, F. Vanzi, L. Fusi, M. Linari, G. Piazzesi, V. Lombardi, and F. S. Pavone, “Probing myosin structural conformation in vivo by second-harmonic generation microscopy,” PNAS 107, 7763 (2010). [CrossRef]
7. P. J. Campagnola, A. C. Millard, M. Terasaki, P. E. Hoppe, C. J. Malone, and W. A. Mohler, “Three-dimensional high-resolution second-harmonic generation imaging of endogenous structural proteins in biological tissues,” Biophys. J. 82(1), 493–508 (2002). [CrossRef]
8. Y. Zhao, K. T. T. Hien, G. Mizutani, and H. N. Rutt, “Second-order nonlinear optical microscopy of spider silk,” Appl. Phys. B 123(6), 188 (2017). [CrossRef]
9. W. L. Rice, S. Firdous, S. Gupta, M. Hunter, C. W. P. Foo, Y. Wang, H. J. Kim, D. L. Kaplan, and I. Georgakoudi, “Non-invasive characterization of structure and morphology of silk fibroin biomaterials using non-linear microscopy,” Biomaterials 29(13), 2015–2024 (2008). [CrossRef]
10. T. Akiyama, A. Inoko, Y. Kaji, S. Yonemura, K. Kakiguchi, H. Segawa, K. Ishitsuka, M. Yoshida, O. Numata, P. Leproux, V. Couderc, T. Oshika, and H. Kano, “SHG-specificity of cellular Rootletin filaments enables naive imaging with universal conservation,” Sci Rep 7(1), 39967 (2017). [CrossRef]
11. M. Yamanaka, H. Niioka, T. Furukawa, and N. Nishizawa, “Excitation of erbium-doped nanoparticles in 1550-nm wavelength region for deep tissue imaging with reduced degradation of spatial resolution,” J. Biomed. Opt. 24(07), 1–4 (2019). [CrossRef]
12. J. Xu, D. Murata, J. Ueda, and S. Tanabe, “Near-infrared long persistent luminescence of Er3 + in garnet for the third bio-imaging window,” J. Mater. Chem. C 4(47), 11096–11103 (2016). [CrossRef]
13. K. Kieu, S. Mehravar, R. Gowda, R. A. Norwood, and N. Peyghambarian, “Label-free multi-photon imaging using a compact femtosecond fiber laser mode-locked by carbon nanotube saturable absorber,” Biomed. Opt. Express 4(10), 2187–2195 (2013). [CrossRef]
14. R. Aviles-Espinosa, S. I. Santos, A. Brodschelm, W. G. Kaenders, C. Alonso-Ortega, D. Artigas, and P. Loza-Alvarez, “Third-harmonic generation for the study of Caenorhabditis elegans embryogenesis,” J. Biomed. Opt. 15(4), 046020 (2010). [CrossRef]
15. H. S. Kitsakorn Locharoenrat and G. Mizutani, “Rotational anisotropy in second harmonic intensity from Cu nanowire arrays on the NaCl(110) substrates,” J. Lumin. 128(5-6), 824–827 (2008). [CrossRef]
16. H. Matsuda, Y. Fujimoto, S. Ito, Y. Nagasawa, H. Miyasaka, T. Asahi, and H. Masuhara, “Development of near-infrared 35 fs laser microscope and its application to the detection of three- and four-photon fluorescence of organic microcrystals,” J. Phys. Chem. B 110(3), 1091–1094 (2006). [CrossRef]
17. M. S. Pochechuev, A. A. Lanin, I. V. Kelmanson, A. S. Chebotarev, E. S. Fetisova, D. S. Bilan, E. K. Shevchenko, A. A. Ivanov, A. B. Fedotov, V. V. Belousov, and A. M. Zheltikov, “Multimodal nonlinear-optical imaging of nucleoli,” Opt Lett 46(15), 3608–3611 (2021). [CrossRef]
18. Y. Ozeki, W. Umemura, Y. Otsuka, S. H. Satoh, Hiroyuki Sumimura, N. F. Kazuhiko, Nishizawa, K. Kiichi, and Itoh, “High-speed molecular spectral imaging of tissue with stimulated Raman scattering,” nature photonics (2021).
19. S. Fredrik, B. Christian, H. Thomas, and M. E. Annika, “Nonlinear microscopy with fiber laser continuum excitation,” J. Biomed. Opt. 15(01), 1–4 (2010). [CrossRef]
20. K. Kieu, B. G. Saar, G. R. Holtom, F. W. Wise, and X. S. Xie, “High power all-fiber picosecond laser system for coherent Raman microscopy,” Opt. Lett. 34(13), 2051–2053 (2009). [CrossRef]
21. M. Baumgartl, T. Gottschall, J. Abreu-Afonso, A. Diez, T. Meyer, B. Dietzek, M. Rothhardt, J. Popp, J. Limpert, and A. Tunnermann, “Alignment-free, all-spliced fiber laser source for CARS microscopy based on four-wave-mixing,” Opt. Express 20(19), 21010–21018 (2012). [CrossRef]
22. L. A. Sordillo, Y. Pu, S. Pratavieira, Y. Budansky, and R. R. Alfano, “Deep optical imaging of tissue using the second and third near-infrared spectral windows,” J. Biomed. Opt 19(5), 056004 (2014). [CrossRef]
23. K. Kieu and M. Mansuripur, “Femtosecond laser pulse generation with a fiber taper embedded in carbon nanotube/polymer composite,” Opt. Lett. 32(15), 2242–2244 (2007). [CrossRef]
24. N. G. Horton, K. Wang, D. Kobat, C. G. Clark, F. W. Wise, C. B. Schaffer, and C. Xu, “In vivo three-photon microscopy of subcortical structures within an intact mouse brain,” Nat. Photonics 7(3), 205–209 (2013). [CrossRef]
25. K. Makihara, D. Kaneta, T. Iwamura, A. Sugita, P. Leproux, V. Couderc, and H. Kano, “Photo-induced meta-stable polar conformations in polystyrene microspheres revealed by time-resolved SHG microscopy,” Appl. Phys. Express 13(5), 052003 (2020). [CrossRef]
26. L. Huang, X. Zhou, Q. Liu, C. E. MacAulay, and S. Tang, “Miniaturized multimodal multiphoton microscope for simultaneous two-photon and three-photon imaging with a dual-wavelength Er-doped fiber laser,” Biomed. Opt. Express 11(2), 624–635 (2020). [CrossRef]
27. D. Kaneta, M. Goto, M. Hagihara, P. Leproux, V. Couderc, M. Egawa, and H. Kano, “Visualizing intra-medulla lipids in human hair using ultra-multiplex CARS, SHG, and THG microscopy,” Analyst 146(4), 1163–1168 (2021). [CrossRef]
28. R. Cisek, D. Tokarz, M. Steup, I. J. Tetlow, M. J. Emes, K. H. Hebelstrup, A. Blennow, and V. Barzda, “Second harmonic generation microscopy investigation of the crystalline ultrastructure of three barley starch lines affected by hydration,” Biomed. Opt. Express 6(10), 3694–3700 (2015). [CrossRef]
29. G. Mizutani, Y. Sonoda, H. Sano, M. Sakamoto, T. Takahashi, and S. Ushioda, “Detection of starch granules in a living plant by optical second harmonic microscopy,” J. Lumin. 87-89, 824–826 (2000). [CrossRef]
30. S. Brenner, “The genetics of Caenorhabditis elegans,” Genetics 77(1), 71–94 (1974). [CrossRef]
31. H. Segawa, N. Fukutake, P. Leproux, V. Couderc, T. Ozawa, and H. Kano, “Electronically resonant third-order sum frequency generation spectroscopy using a nanosecond white-light supercontinuum,” Opt. Express 22(9), 10416–10429 (2014). [CrossRef]
32. H. Yoneyama, P. Leproux, V. Couderc, A. Inoko, and H. Kano, “CARS molecular fingerprinting using sub-100-ps microchip laser source with fiber amplifier,” APL Photonics3(9), 092408 (2018). [CrossRef]
33. D. Yelin and Y. Silberberg, “Laser scanning third-harmonic-generation microscopy in biology,” Opt. Express 5(8), 169–175 (1999). [CrossRef]
34. T. Hardy, M. Lee, R. S. Hames, S. L. Prosser, D. M. Cheary, M. D. Samant, F. Schultz, J. E. Baxter, K. Rhee, and A. M. Fry, “Multisite phosphorylation of C-Nap1 releases it from Cep135 to trigger centrosome disjunction,” J Cell Sci 127, 2493–2506 (2014). [CrossRef]
35. R. Vlijm, X. Li, M. Panic, D. Ruthnick, S. Hata, F. Herrmannsdorfer, T. Kuner, M. Heilemann, J. Engelhardt, S. W. Hell, and E. Schiebel, “STED nanoscopy of the centrosome linker reveals a CEP68-organized, periodic rootletin network anchored to a C-Nap1 ring at centrioles,” Proc Natl Acad Sci USA 115(10), E2246–E2253 (2018). [CrossRef]
36. G. Conley Natividad, S. K. Theodossiou, N. R. Schiele, G. K. Murdoch, A. Tsamis, B. Tanner, G. Potirniche, M. Mortazavi, D. A. Vorp, and B. A. Martin, “Ex-vivo quantification of ovine pia arachnoid complex biomechanical properties under uniaxial tension,” Fluids Barriers CNS 17(1), 68 (2020). [CrossRef]
37. J. DeSisto, R. O’Rourke, H. E. Jones, B. Pawlikowski, A. D. Malek, S. Bonney, F. Guimiot, K. L. Jones, and J. A. Siegenthaler, “Single-cell transcriptomic analyses of the developing meninges reveal meningeal fibroblast diversity and function,” Dev Cell 54(1), 43–59.e4 (2020). [CrossRef]
38. D. G. McLone and W. Bondareff, “Developmental morphology of the subarachnoid space and contiguous structures in the mouse,” Am. J. Anat. 142(3), 273–293 (1975). [CrossRef]
39. J. Trägårdh, G. Robb, K. K. E. Gadalla, S. Cobb, C. Travis, G.-L. Oppo, and G. McConnell, “Label-free imaging of thick tissue at 1550 nm using a femtosecond optical parametric generator,” Opt. Lett. 40(15), 3484–3487 (2015). [CrossRef]
40. D. A. Dombeck, K. A. Kasischke, H. D. Vishwasrao, M. Ingelsson, B. T. Hyman, and W. W. Webb, “Uniform polarity microtubule assemblies imaged in native brain tissue by second-harmonic generation microscopy,” Proc Natl Acad Sci U S A 100(12), 7081–7086 (2003). [CrossRef]
41. J. Yang, X. Q. Liu, G. H. Yue, M. Adamian, O. Bulgakav, and T. S. Li, “Rootletin, a novel coiled-coil protein, is a structural component of the ciliary rootlet,” J Cell Biol 159(3), 431–440 (2002). [CrossRef]
42. V. Van Steenbergen, W. Boesmans, Z. Li, Y. de Coene, K. Vints, P. Baatsen, I. Dewachter, M. Ameloot, K. Clays, and P. Vanden Berghe, “Molecular understanding of label-free second harmonic imaging of microtubules,” Nat Commun 10(1), 3530 (2019). [CrossRef]
43. Marjo S. van der Knaap and Jaap Valk, “Magnetic Resonance of Myelination and Myelin Disorders, 3rd ed,” Springer (2005).
44. S. Miyazaki, P. Leproux, V. Couderc, Y. Hayashi, and H. Kano, “Multimodal nonlinear optical imaging of Caenorhabditis elegans with multiplex coherent anti-Stokes Raman scattering, third-harmonic generation, second-harmonic generation, and two-photon excitation fluorescence,” Appl. Phys. Express 13(7), 072002 (2020). [CrossRef]
45. P. Mahou, N. Olivier, G. Labroille, L. Duloquin, J. M. Sintes, N. Peyrieras, R. Legouis, D. Debarre, and E. Beaurepaire, “Combined third-harmonic generation and four-wave mixing microscopy of tissues and embryos,” Biomed. Opt. Express 2(10), 2837–2849 (2011). [CrossRef]
46. G. V. Clokey and L. A. Jacobson, “The autofluorescent lipofuscin granules in the intestinal-cells of caenorhabditis-elegans are secondary lysosomes,” Mech Ageing Dev 35(1), 79–94 (1986). [CrossRef]