Higher harmonic generation microscopy of in vitro cultured mammal oocytes and embryos

Oocyte and embryo selection governs the success of assisted reproductive technologies. The imaging tools applied for selecting embryos may need to contain several key properties: noninvasiveness, high 3D resolution, and the contrast capability to provide as much information about the embryos as possible, such as spindle fibers, zona pellucida, and organelles. Currently adopted imaging techniques can only provide one or two of these desired properties and are with limited contrast of the embryos. Some image techniques can even damage the embryos. Previous studies have shown that harmonic generation microscopy (HGM), a virtualtransition based technology, can provide noninvasive imaging in zebrafish embryos with a sub-cellular 3D resolution and a millimeter penetration depth, and thus could be a suitable tool for future oocyte and embryo selection of assisted reproductive technologies. However to evaluate HGM in clinical use, the intrinsic contrast origin of the second harmonic generation (SHG) and third harmonic generation (THG) inside the mammal embryos has to be studied. In this work we performed HGM studies on the in vitro cultured mouse oocytes and embryos by combining the SHG and THG modalities, with a focus on the contrast origin evaluation. Through the noninvasive HGM imaging, we can clearly identify various structures in the whole oocytes and embryos, including spindle fibers, zona pellucida, polar bodies, cell membranes, and the laminated organelles in the cells. The origin of the THG contrast was further confirmed through the standard staining studies. Through SHG signals, we could not only observe the spindle fibers when the oocytes were arrested at metaphase II or during the cleavage of the embryos, but can also distinguish and analyze the thickness of the three layers of the zona pellucida. Combining two different higher-harmonic generation modalities, SHG and THG, HGM successfully revealed the subcellular structures of the whole mouse embryos with a high 3D spatial resolution. ©2008 Optical Society of America OCIS codes: (170.3880) Medical and biological imaging; (170.5810) Scanning microscopy; (170.6900) Three-dimensional microscopy References and links 1. L. Liu and D. L. Keefe, “Ageing-associated aberration in meiosis of oocytes from senescence-accelerated mice,” Hum. Reprod. 17, 2678-2685 (2002). 2. S. Cooke, J. P. P. Tyler, G. L. Driscoll, “Meiotic spindle location and identification and its effect on embryonic cleavage plane and early development,” Hum. Reprod. 18, 2397-2405 (2003) 3. P. Saldeen and P. Sundström, “Nuclear status of four-cell preembryos predicts implantation potential in in vitro fertilization treatment cycles,” Fertil. Steril. 84, 584-589 (2005). #93726 $15.00 USD Received 11 Mar 2008; revised 17 May 2008; accepted 13 Jun 2008; published 18 Jul 2008 (C) 2008 OSA 21 July 2008 / Vol. 16, No. 15 / OPTICS EXPRESS 11574 4. J. Holte, L. Berglund, K. Milton, C. Garello, G. Gennarelli, A. Revelli, T. Bergh, “Construction of an evidencebased integrated morphology cleavage embryo score for implantation potential of embryos scored and transferred on day 2 after oocyte retrieval,” Hum. Reprod. 22, 548-557 (2007) 5. W. T. Garside, J. R. L. D. Mola, J. A. Bucci, R. W. Tureck, S. Heyner, “Sequential analysis of zona thickness during in vitro culture of human zygotes: Correlation with embryo quality, age, and implantation,“ Mol. Reprod. Dev. 47, 99-104 (1997). 6. A. Gabrielsen, P.R. Bhatnager, K. Petersen and S. Lindenberg, “Influence of zona pellucida thickness of human embryos on clinical pregnancy outcome following in vitro fertilization treatment,” J. Assist. Reprod. Genet. 17, 323–328 (2000). 7. C. Pelletier, D. L. Keefe, and J. R. Trimarchi, “Noninvasive polarized light microscopy quantitatively distinguishes the multilaminar structure of the zona pellucida of living human eggs and embryos,” Fertil. Steril. 81, 850-856 (2004). 8. Y. Shen, T. Stalf, C. Mehnert, U. Eichenlaub-Ritter, and H.-R. Tinneberg, “High magnitude of light retardation by the zona pellucida is associated with conception cycles,” Hum. Reprod. 20, 1596-1606; (2005). 9. B. S. Weakley, P. Webb, and J. L. James, “Cytochemistry of the Golgi apparatus in developing ovarian germ cells of the Syrian hamster,” Cell Tissue Res. 220, 349-372 (1981). 10. R. D. Moreno, G. Schatten, and J. Ramalho-Santos, “Golgi apparatus dynamics during mouse oocyte in vitro maturation: effect of the membrane trafficking inhibitor brefeldin A,” Biol. Reprod. 66, 1259-1266 (2002) 11. L. M. Mehlmann, M. Terasaki, L. A. Jaffe, D. Kline, “Reorganization of the endoplasmic reticulum during meiotic maturation of the mouse oocyte,” Dev. Biol. 170, 607-615 (1995). 12. M. Terasaki, L. L. Runft, and A. R. Hand, “Changes in organization of the endoplasmic reticulum during Xenopus oocyte maturation and activation,” Mol. Biol. Cell 12, 1103-1116 (2001) 13. G. Fitzharris, P. Marangos, and J. Carroll, “Cell cycle-dependent regulation of structure of endoplasmic reticulum and inositol 1, 4, 5-trisphosphate-induced Ca release in mouse oocytes and embryos,” Mol. Biol. Cell 14, 288-301 (2003). 14. G. FitzHarris, P. Marangos, and J. Carroll, “Changes in endoplasmic reticulum structure during mouse oocyte maturation are controlled by the cytoskeleton and cytoplasmic dynein,” Dev. Biol. 305, 133-144 (2007). 15. T. Tokura, Y. Noda, Y. Goto, and T. Mori, “Sequential observation of mitochondrial distribution in mouse oocytes and embryos,” J. Assist. Reprod. Genet. 10, 417-426 (1993). 16. J. V. Blerkom, M. N. Runner, “Mitochondrial reorganization during resumption of arrested meiosis in the mouse oocyte,” Am. J. Anat. 171, 335-355 (1984). 17. J. V. Blerkom, P. Davis, V. Mathwig, and S. Alexander, “Domains of high-polarized and low-polarized mitochondria may occur in mouse and human oocytes and early embryos” Hum. Reprod. 17, 393–406 (2002). 18. M. Shribak, J. LaFountain, D. Biggs, and S. Inoué, “Quantitative orientation-independent differential interference contrast (DIC) microscopy coupled with orientation-independent polarization microscopy,” Microsc. Microanal. 13, 10-11 (2007). 19. S. Inoué, and R. Oldenbourg, “Microtubule dynamics in mitotic spindle displayed by polarized light microscopy,” Mol. Biol. Cell 9, 1603-1607 (1998). 20. J. V. Blerkom and G. Henry, “Oocyte dysmorphism and aneuploidy in meiotically mature human oocytes after ovarian stimulation,” Hum. Reprod. 7, 379-390 (1992). 21. J. M. Squirell, D. L. Wokosin, J. G. White, and B. D. Bavister, “Long-term two-photon fluorescence imaging of mammalian embryos without compromising viability,” Nat. Biotechnol. 17, 763-767 (1999). 22. S.-Y. Chen, C.-S. Hsieh, S.-W. Chu, C.-Y. Ko, Y.-C. Chen, H.-J. Tsai, C.-H. Hu, C.-K. Sun, “Noninvasive harmonics optical microscopy for long-term observation of embryonic nervous system development in vivo. “J. Biom. Opt. 11, 054022 1-8 (2006). 23. C.-K. Sun, S.-W. Chu, S.-Y. Chen, T.-H. Tsai, T.-M. Liu, C.-Y. Lin, and H.-J. Tsai, “Higher harmonic generation microscopy for developmental biology,” J. Struct. Bio. 147, 19-30 (2004). 24. S.-W. Chu, S.-Y. Chen, T.-H. Tsai, T.-M. Liu, C.-Y. Lin, H.-J. Tsai, and C.-K. Sun, “In vivo developmental biology study using noninvasive multi-harmonic generation microscopy,” Opt. Express 11, 3093-3099 (2003). 25. V. Barzda, C. Greenhalgh, J. A. D. Au, S. Elmore, J. H. van Beek, and J. Squier, "Visualization of mitochondria in cardiomyocytes by simultaneous harmonic generation and fluorescence microscopy," Opt. Express 13, 82638276 (2005). 26. K. Konig, P. T. C. So, W. W. Mantulin, E. Gratton, “Cellular response to near-infrared femtosecond laser pulses in two-photon microscopes,” Opt. Lett. 22, 135-136 (1997). 27. S.-U. Chen, Y.-L. Lien, and H.-F. Chen, “Open pulled straws for vitrification of mature mouse oocytes preserve patterns of meiotic spindles and chromosomes better than conventional straws,” Hum. Reprod. 15, 2598–2603 (2000). 28. S.-U. Chen, Y.-R. Lien, Y.-Y. Cheng, H.-F. Chen, H.-N. Ho, and Y.-S. Yang, “Vitrification of mouse oocytes using closed pulled straws (CPS) achieves a high survival and preserves good patterns of meiotic spindles, compared with conventional straws, open pulled straws (OPS) and grids,” Hum. Reprod. 16, 2350 – 2356 (2001). 29. S.-U. Chen, T.-H. Lee, Y.-R. Lien, Y.-Y. Tsai, L.-J. Chang, and Y.-S. Yang, “Microsuction of blastocoelic fluid before vitrification increased survival and pregnancy of mouse expanded blastocysts, but pretreatment with the cytoskeletal stabilizer did not increase blastocyst survival,” Fertil. Steril. 84, 1156-1162 (2005). 30. T. Y. F. Tsang, “Optical 3rd-harmonic generation at interfaces,” Phys. Rev. A 52, 4116-4125 (1995). #93726 $15.00 USD Received 11 Mar 2008; revised 17 May 2008; accepted 13 Jun 2008; published 18 Jul 2008 (C) 2008 OSA 21 July 2008 / Vol. 16, No. 15 / OPTICS EXPRESS 11575 31. S.-W. Chu, I.-H. Chen, T.-M. Liu, C.-K. Sun, S.-P. Lee, B.-L. Lin, P.-C. Cheng, M.-X. Kuo, D.-J. Lin, and H.L. Liu, “Nonlinear bio-photonic crystal effects revealed with multimodal nonlinear microscopy,” J. Microsc.Oxford 208, 190-200 (2002). 32. 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, 493-508 (2002). 33. J. A. Squier, M. Muller, and G. J. Brakenhoff, and K. R. Wilson, “Third harmonic generation microscopy,” Opt. Express 3, 315-324 (1998). 34. Y. Barad, H. Eisenberg, M. Horowitz, and Y. Silberberg, “Nonlinear scanning laser microscopy by third harmonic generation,” App. Phy. Lett. 70, 922-924 (1997). 35. M. Muller, J. Squier, K. R. Wilson, and G. J. Brakenhoff, “3D microscopy of transparent objects using thirdharmonic generation,” J. Microsc.-Oxford 191, 266-274 (1998). 36. S.-W. Chu, S.-P. Tai, C.-L. Ho, C.-H. Lin, and C.-K Sun, “High-resolution simultaneous three-photon fluorescence and third-harmonic-gene


Introduction
In order to improve the implantation and pregnancy potential, the selection of embryos has become a key project in the assisted reproductive technologies (ART), which include in vitro fertilization (IVF) and intra cytoplasmic sperm injection (ICSI).There are several measurable indicators to qualify oocytes and embryos: the alignment and shape of MI and MII spindles [1], the relative location of the first polar body, second polar body, and the meiotic spindle [2], the blastomere number, fragmentation, blastomere size variation, symmetry of the cleavage and mononuclearity in the blastomeres [3,4], the thickness, retardance, and degree of irregularity of zona pellucida [5][6][7][8], as well as the distribution of organelles, including Golgi apparatus [9,10], endoplasmic reticulum (ER) [11][12][13][14], and mitochondria [15][16][17].Many live imaging systems are used to study embryogenesis of embryology and evaluate the quality of oocytes and embryos for clinical use.Differential Interference Contrast Microscopy (DIC) can observe structure of unstained, transparent living cells and organelles, producing an optical phase gradient image [18].However, DIC does not provide depth-resolved images and can not show spindle fibers due to their slight variation in refractive index with the surrounding cytoplasm.Polar scope, which can observe the specimens that are visible primarily due to their optically anisotropic characteristic, reveals the spindle fibers in the oocyte [2,19,20] but is not sensitive to the detailed inner structure of the examined embryo and does not have a high depth resolution, and therefore is unable to provide 3D reconstruction of the cell distribution.Compared with the commonly adopted 2D imaging tools, the challenge of using 3D confocal microscopy to observe mammalian embryos is to maintain embryo viability following continuous exposure to excitation illumination [21] and the invasiveness of dye staining.To the best of our knowledge, there is also no clinically-approved staining dye for embryo and oocyte selection in IVF.Harmonic Generation Microscopy (HGM), a virtualtransition based technique, has previously provided noninvasive cellular images deep inside the live zebrafish embryos [22] with high cell viability and could be an ideal imaging system to assist mammalian embryo selection due to its high penetration capability, high 3D spatial resolution, and minimized photodamage and photobleaching effects.However to evaluate HGM in clinical use, the intrinsic contrast origin of the second harmonic generation (SHG) and third harmonic generation (THG) inside the mammal embryos has to be studied.
In this paper we report a HGM study of in vitro cultured mouse oocytes and embryos.Similar to the previous zebrafish embryonic study [23], HGM successfully revealed spindle fibers through the SHG modality and cell membranes through the THG modality [24,25].Through the SHG modality, we also observed the three layers of the zona pellucida, with two layers of strong SHG signals with one dark layer in between, while the SHG polarization anisotropy of the zona pellucida was also investigated to study its contrast origin.Through fluorescence dye labeling, we also successfully unraveled the contrast origins of the THG observed organelles to be dominated by Golgi apparatus, endoplasmic reticulum (ER), and mitochondria.Through high-resolution imaging the embryos at different stages, including the stages of oocyte, 2P, 2-cell, 4-cell, 8-cell uncompacted, 8-cell compacted, morula, and blastocyst, we can observe the developmental status of the whole mouse embryos, such as the thinning of zona pellucida, the expression of cell adhesion, and the cleavage of cells with a high 3D spatial resolution without the assist of the fluorescence dyes.Our study indicates that HGM can provide the much desired contrasts for oocyte and embryo selection with a high 3D resolution and could be an ideal imaging tool for ART.

Materials and methods
The experimental protocols were approved by the National Taiwan University Institutional Animal Care and Use Committee (NTU-IACUC).

Harmonic Generation Microscope
Our laser scanning harmonic generation microscope adapts an Olympus FV300 scanning unit along with an Olympus IX71 microscope.The collimated Cr:forsterite laser beam, whose center wavelength was 1230nm with a repetition rate of 110MHz and a pulse width of 140fs, was coupled into the scanning system as the excitation source.In previous femtosecond laser viability studies with a numerical aperture (NA) ~1 objective, by using a laser at 1230 nm [22,23], deep penetration and high cell viability can be provided even under a strong illumination power of 100 mW.In contrast, with a 800nm Ti:sapphire laser, less than 2-7 mW average power could be applied to live samples to prevent optical damage [26].As the excitation laser wavelength is increased to 1047 nm [21], the maximum average illumination power can only be increased to 13 mW to provide the cell viability.In order to observe the whole 3D structure of the mouse oocytes and embryos, we used a 2-mm-working-distance high NA infrared objective (LUMPlanF1/IR 60X/water/NA0.9,Olympus) to focus the laser beam onto the desired location of the specimen with a spot size close to its diffraction limit.The average power after the objective was about 150mW.Taking advantages of the transparence and the finite thickness of the live specimen, we used a NA0.9 air condenser to collect higher harmonic generation signals in the forward direction.The SHG and THG signals were separated by a beam splitter and were guided into two photomultiplier tubes (PMTs) (Hamamatsu) to record the filtered SHG and THG signals respectively.For the THG contrast origin study, the filtered fluorescence was collected using the focusing infrared objective in the epi-direction for PMT detection.For the observation of oocytes and embryos stained with BODIPY FL C 5 -ceramide (Molecular Probes, Eugene, OR), ER-Tracker Blue-White DPX (Molecular Probes, Eugene, OR), and LysoSensor Green DND189 (Molecular Probes, Eugene, OR), a filter with a central wavelength of 525nm and a full-width-half-maximum (FWHM) bandwidth of 50nm was used.For the observation of oocytes and embryos stained with MitoTracker Red CM-H 2 XRos (Molecular Probes, Eugene, OR) and LysoTracker Red Lysosomal Probe (Lonza Walkersville, Walkersville, MD), a filter with a central wavelength of 640nm and a FWHM bandwidth of 20nm was used.During observation, the mouse oocytes and embryos were placed in glass-bottom culture dishes (MatTek, Ashland, MA), and the dishes were placed in a CO 2 micro-incubator (MIU-IBC-I, Olympus).The environment inside the dish was kept at just below 37°C with a controlled 5% CO 2 concentration in air.

Preparation of oocytes and embryos
Female ICR mice aged 4-8 weeks and more than 8 months (only for the aging mouse test) were injected with 10 International Unit (IU) of pregnant mare's serum gonadotrophin (Sigma, St. Louis, MO) to induce them to superovulate [27].Later, these female mice were injected intraperitoneally with 10 IU of human chorionic gonadotropin (HCG) (Organon, Oss, Netherlands) to trigger ovulation.For HGM observation of oocytes, the oviducts were excised and the cumulus-oocyte complexes were collected in human tubal fluid (HTF) medium.By pipetting the cumulus-oocyte complexes in HTF medium containing 80 IU/ml hyaluronidase (Sigma-Aldrich, St. Louis, MO) and by washing them, the granulosa cells of the oocytes were removed.The matured oocytes with the first polar body were collected for the experiments [28].For HGM observation of embryos, each female mouse was placed with a male mouse for mating.The oviducts were then excised and the fertilized embryos were collected in HTF medium [29].The oocytes and embryos were cultured with HTF medium containing 0.5% bovine serum albumin (BSA; Sigma, St. Louis, MO) in an atmosphere with 5% CO 2 in air at 37°C.

Removal of zona pellucida
For the fluorescence staining experiments (to study the THG origin), the removal of zona pellucida is necessary.To remove the zona pellucida, the oocytes and embryos were transferred to M2 medium (Sigma, St. Louis, MO), then transferred to acid Tyrode media (Sigma, St. Louis, MO).After the dissolution of zona pellucida, the oocytes and embryos were transferred back to the M2 medium for the later staining.

Fluorescence staining
The zona-pellucida-removed oocytes and embryos were stained according to the dye's commercial protocol.For Golgi apparatus labeling, the naked oocytes and embryos were placed in the M2 medium containing 5μM of BODIPY FL C 5 -ceramide (Molecular Probes, Eugene, OR) and incubated in dark for 30 minutes.For ER labeling, the oocytes and embryos were placed in the M2 medium containing 1μM of ER-Tracker Blue-White DPX (Molecular Probes, Eugene, OR) and incubated in dark for 30 minutes.For mitochondria staining, the oocytes and embryos were placed in the M2 medium containing 500nM of MitoTracker Red CM-H 2 XRos (Molecular Probes, Eugene, OR) and incubated in dark for 45 minutes.For lysosome labeling, the naked oocytes and embryos were placed in 2 wells, one was filled with the M2 medium containing 75nM of LysoSensor Green DND189 (Molecular Probes, Eugene, OR), the other was filled with the M2 medium containing 75nM of LysoTracker Red Lysosomal Probe (Lonza Walkersville, Walkersville, MD).Both of them were incubated in dark for 2 hours.All the oocytes and embryos were then washed in fresh medium several times and incubated for an additional 0.5-1 hour.For correlation analysis, five data were collected from oocytes and four data were collected from embryos stained by each dye.

HGM imaging
Due to the localized excitation property of the nonlinear processes under tight focus and the high penetration capability of the 1230 nm incident light, HGM allows the least invasive observation of the whole mouse embryo with a diffraction-limited 3D spatial resolution.Fig. 1 shows a z-series of the depth-resolved HGM images (512x512 pixels) of the in vitro studied mouse embryo.The 3D embryo structure can be completely resolved from the surface of the embryo to the bottom.No significant HGM signal degradation between the images of difference depths can be found.Fig. 2 shows the optically sectioned HGM images of the in vitro cultured mouse oocyte and different embryos at various developing stages.The typical frame rate during acquisition was about 2 seconds per frame.Due to the relatively thin oocyte and embryo thickness, all specimens were found to be easily examined under the Cr:forsterite laser based HGM with a full penetration capability.Because of the Gouy phase shift effect with a tightly focused laser beam, THG is known to occur at optically thin layers or at the interface of two media with different linear (refractive index) or nonlinear optical susceptibilities, such as cell membranes, membrane organelles [30], and polar bodies (red arrow in Fig. 2(A)).Since SHG only arises from highly organized nano-structure with ordered arrangement of highly asymmetric biomolecules where the optical centrosymmetry is broken [31,32], it can reveal spindles (yellow arrow in Fig. 2(A)) [23] and zona pellucida (white arrow in Fig. 2(A)).It is known that before ovulation, the eggs enlarge, divide by meiosis, and mature in their ovarian follicles until they reach the metaphase II and release the first polar body, as revealed by the HGM sectioned image shown in Fig. 2(A).The matured oocyte, which was arrested at the metaphase of meiosis and kept the spindle (indicates by yellow arrow in Fig. 2(A)), was surrounded by a protective coat of noncellular materials (made of extracellular matrix and glycoproteins) called the zona pellucida (indicates by white arrow in Fig. 2(A)).For HGM images, the zona pellucida contains 2 layers of SHG signals, an inner For fertilization to occur, a sperm cell must bind to the zona pellucida, penetrate the zona pellucida and the cell membrane of the oocyte, and enter the oocyte cytoplasm, thus forming two pronuclei: one from the sperm and one from the oocyte.Fig. 2(B) and Fig. 2(C) show the HGM images of the same live mouse embryo at the 2P stage optically sectioned at different depths.By scanning from the surface of the embryo to the bottom, we can clearly observe these two nuclei through THG (indicate by white arrows) due to its sensitivity in membranes.
After fusion of the sperm and egg pronuclei, the cleavage begins.After fertilization, the zygote is known to move toward the uterus, a journey that takes three to four days in mice.As it travels, the zygote divides.The first cleavage produces two identical cells (corresponding to the imaged embryo stage in Fig. 2(D)) and then these cells divide again to produce four cells (corresponding to the imaged embryo stage in Fig. 2(E)).If these cells separate, genetically identical embryos will be resulted, which is the basis of identical twinning.Usually, these cells remain together, dividing asynchronously to produce 8 cells, 16 cells, and so on.At about the eight-cell stage (corresponding to the imaged embryo stage in Fig. 2(F)), the cell adhesion proteins are known to express and the embryo starts to compact, meaning that the formerly "loose" ball of cells (called the blastomere) begin to huddle together into a tight array that is interconnected by gap junctions, as been truthfully reflected by the sectioned HGM image shown in Fig. 2(G).By the 16-cell stage, the compacted embryo is termed a morula (corresponding to the imaged embryo stage in Fig. 2(H)).In mice, the first evidence that cells have become specialized is when the outer cells of the 16-cell morula divide and produce an outer rim of cells (called the trophectoderm) and an inner core of cells (called the inner cell mass, ICM).Ultimately, the cells of the ICM will give rise to all the tissues of the embryo's body.The trophectoderm, in turn, will generate the trophoblast cells of the chorion, which is the embryo's contribution to the extraembryonic tissue known as the placenta.
By embryonic day 3 (E3.0) in the development of ICR mice, the embryo will develop a cavity called the blastocoel (as revealed by HGM as indicated with the yellow arrow in Fig. 2(I)).As a result of cavitation and the physical separation and differentiation of the trophectoderm from the inner cell mass, the morula can be found to become a blastocyst, corresponding to the imaged embryo stage in Fig. 2(I).Its chief structural features revealed by THG signals are the outer sphere of the flattened trophectoderm cells (which became the trophoblast as indicated by the white arrow in Fig. 2(I)), the inner cell mass (as indicated by the red arrow in Fig. 2(I)), and the fluid-filled blastocoel.The thinning of the zona pellucida was also reflected by the SHG signals.At about E4.0 stage of embryogenesis in mice, embryonic stem cells will be derived from the inner cell mass of the blastocyst.

Organelle origins for THG contrast
Because of the Gouy phase shift effect in a focused laser beam, strong THG signals are expected to occur at the interface of two media with different optical properties [30,[33][34][35], and THG is reported to provide image contrast for cell membrane, mitochondria [36][37][38], lipid bodies [39], and elastin fibers [40].In this HGM study, THG can not only provide contrast for cell membranes and the polar body (indicates by red arrow in Fig. 2(A)), but also organelles in cells.Since the distribution of some organelles, including Golgi apparatus [9,10], ER [11][12][13][14], mitochondria [15][16][17], indicates the quality of oocytes and embryos, in order to evaluate the potential of HGM to serve as an oocyte and embryo selecting tool, the organelle origin of the THG contrast should be studied and clarified.
The major eukaryotic organelles are Golgi apparatus, ER, mitochondria, and lysosomes.We chose 5 multiphoton dyes to target these organelles in mouse oocytes.Using the Cr:forsterite laser with a central wavelength of 1230nm, we collected the two photon fluorescence (2PF) or three photon fluorescence (3PF) emitted from the dye stained organelles.Their corresponding 2PF and 3PF spectra after normalization are shown in Fig. 3  We quantitatively analyzed the correlation between the two signals by using the Pearson correlation [41]: where X is the THG signal intensity, Y is the fluorescence signal intensity, with means X , Y , and standard deviations S X , S Y respectively.For each dye, five data were collected from the horizontally sectioned images near the center of each stained oocyte and embryo.The results are shown in Fig. 5.We can see that the fluorescence emitted from the BODIPY FL C 5 -ceramide labeled Golgi apparatus, the ER-Tracker Blue-White DPX labeled endoplasmic reticulum, and the MitoTracker Red CM-H 2 XRos labeled mitochondria are better correlated with the THG than LysoSensor Green DND189 and LysoTracker Red Lysosomal Probe labeled lysosomes.The former three dye experiments have Pearson's correlation higher than 0.3, while the latter two dye experiments have Pearson's correlation lower than 0.2.The correlation coefficient ranges from 0-0.3 are usually viewed as little or no association [41].Therefore we suggest that the THG signals revealed in the mouse oocytes are primarily contributed from large sized membrane organelles, including Golgi apparatus, ER, and mitochondria.This is highly reasonable since THG prefers to arise from thin interfaces of two media with different optical properties, such as membranes.Forward THG intensity also nonlinearly increases with the source size as long as the size is comparable or less than the optical wavelength.A Golgi apparatus is ~1-3 μm in size with 4-7 membranes stacked Pearson's correlation Dyes together.ERs make up 50% of all membranes of a cell.Mitochondrias are double membrane bound organelles with size of 2-5μm in length and 0.5-1μm in diameter.These organelles are also ten times or more larger than lysosomes, and are much larger than other organelles.It is worth to mention that all these adopted staining processes did not enhance the THG intensities, while under the same PMT voltage all the THG intensities of the stained oocytes and embryos were measured to be equal to or slightly less than their corresponding THG intensities before staining, thus not alternating the THG contrast, very different from a previous report with absorption dye staining [42].Our contrast origin study also does not imply that these staining dyes are applicable to future clinical oocyte and embryo selection in assisted reproductive technologies due to dye toxicity issues.

SHG contrast study
Since SHG only arises from highly organized nano-structures with ordered arrangement of highly asymmetric biomolecules where the optical centrosymmetry is broken [31], it reveals spindles and zona pellucida in our studied specimens.From the previously reported polar scope images, the zona pellucida of mouse [43] and human [7,8] embryos can be divided into three layers: a greyish-appearing outer layer (OL), a dark-appearing middle layer (ML), and a highly birefringent inner layer (IL).From HGM images, the zona pellucida revealed by SHG signals can also be divided into three layers.Fig. 6(A) shows the in vitro SHG image of a mouse embryo at the morula stage.By analyzing the axial intensity distribution of SHG signals (Fig. 6(C)) in the selected area of the zona pellucida, we can estimate the thickness of the three layers.In Fig. 6(C), the dash line represents half of the sum value of the maximum intensity of the OL and the minimum intensity of the ML.By dividing the SHG signals from zona pellucida into three sections: the thickness ratio of the three layers can be estimated to be: ~55% for IL (~ 5μm), ~22% for ML (~ 2 μm), and ~ 23% for OL (~ 2.1 μm).These results agree well with the previous analyzed results by using polarized light microscope [43].
The zona pellucida is an extracellular glycoprotein coat and is revealed as an uniform layer under DIC and transmission electron microscope [44], but separates into three layers under a pol-scope [43], implying the similarity of molecular properties but difference in optical properties of the three layers.The measurement of birefringence showed the different retardance of the three layers [43].Within the zona pellucida, it was previously known that the projections of the OL form a hexagonal network with the zona fibrils [8,[45][46][47], while the filaments of the IL appear to course radically from the oocyte, forming a cylindrical like structure [43,48].For materials with a cylindrical symmetry, the general vector expression for the second-order nonlinear polarization is [49] ) where s ˆ represents the unit vector along the symmetry axis, E is the incident electric field with linear polarization, a, b, and c are parameters related to effective second-order SHG coefficient tensors, with d 31 = b, d 15 = c/2, and d 33 = (a + b + c) [49].Since the hexagonal symmetry and the cylindrical symmetry share the same form of the second order nonlinear susceptibility tensor, the SHG intensity distribution of the IL and OL should then be expected to be proportional to following Eq.( 2), where θ is the angle between s ˆ and E [50].Fig. 7 shows an example analysis on the SHG polarization anisotropy in zona pellucida.Similar results can also be found in other in vitro HGM images of mouse oocytes and embryos with an average d 33 /d 31 ratio of 2.71 for IL (out of 9 data) and an average d 33 /d 31 ratio of 2.17 for OL (out of 5 data).The agreements in the polarization dependency between our measurements and Eq. ( 3) suggest that the SHG contrast of zona pellucida should be contributed from the previously reported hexagonal network structure [8,[45][46][47] in the OL and the radial filament alignment in the IL [43,48].

Viability test
In order to show that the Cr:forsterite laser based HGM is truly noninvasive on the examined mouse embryos, we did a serious of viability tests.The first one is that we collected the mouse embryos at the 4-cell stage from the oviduct.Half of the embryos were used as the control set and were placed in the MIU-IBC-I incubator for 10 mins without laser illumination, and then were kept in the CO 2 incubator for 2.5 days.For the experimental set, another half of the embryos were continuously scanned without disruption (in contrast to the previous disrupted two photon work [21]) using HGM with 140mW laser power for 10 mins (total exposure = 29J per embryo) and were then placed in the CO 2 incubator for 2.5 days.The surviving rate were determined depending on those of which developed to the blastocyst stage in the correct time scale.The viability of experimental set was 67% (6 embryos developed normally to the blastocyst stage out of 9), which is very close to the 70% of the control set (7 embryos developed normally to the blastocyst stage out of 10).Our study confirms the noninvasive nature of the Cr:forsterite laser based HGM.
Another viability test is the fertility test.Nowadays doctors culture the embryos till the blastocyst stage before implantation, the selection of the embryos is thus performed right before their implantation back into the mothers.The morula stage, which is the stage just before blastocyst, could therefore be a proper time point to check the embryos' quality.To mimic this situation, we collected the embryos at the morula stage.After the same dosage of continuous illumination (29J) in HGM observation, the embryos were transferred back to the mother mouse.For the control set (with no illumination), 13 embryos were implanted with 8 mouse successfully born (62%).For the experiment set (with illumination), 18 embryos were implanted with 12 mouse successfully born (67%).Fig. 8 shows the 12 born mice of the experiment set.This study further supports our previous viability test that the HGM not only has negligible influence on the mammalian embryo development, but also would not affect pregnancy potential of embryos.A larger scale viability test should be further performed before HGM can be applied for human embryo selections.

Aging mouse test
The female fertility decreases with age due to poor oocyte quality, both in mice [51] and human [52,53].The dynamics of organelles of the oocytes greatly influence the quality of the oocyte.It is known that the presence of smooth endoplasmic reticulum clusters (sERCs) is related to the lower chance of successful pregnancy [54].It was also reported that the vacuole formation in the mitochondria distribution indicates damaged oocytes [55].Since THG images reveal the distribution of these organelles, HGM could thus be a useful tool in identifying oocyte quality.According to previous studies, 30-40 week-old mice can be viewed as aging mice.Fig. 9 are in vitro HGM images of oocytes obtained from female mice at different ages.

Summary
As the improvement of implantation and pregnancy potential has been concerned more and more in modern days, the proper selection of healthy oocytes and embryos plays an important role during IVF.Current live imaging tools have their limits in scoring oocytes and embryos.DIC lacks the ability to indicate spindle fibers and zona pellucida while polscope cannot image organelles, and both of them are unable to provide 3D resolutions on the examined 3D objects.Confocal and two-photon systems need dye staining and therefore are too invasive for clinical applications.With a high spatial resolution and the noninvasive nature, HGM could thus be a potential tool in selecting embryos.In this work, we study the contrast mechanisms of THG and SHG in mammalian oocytes and embryos to examine the potential of HGM for IVF applications.By using a light source at 1230nm, HGM can optically section the whole mammal embryos without compromising viability with a high 3D resolution.THG can provide the contrasts for cell membranes and laminated organelles, including Golgi apparatus, ER, and mitochondria.Endogenous THG signals also reveal the blastomere, the nucleus, and the polar body.On the other hand, SHG signals not only reveal the much required spindle fibers, but also indicate that the zona fibrils form a hexagonal network with a cylindrical symmetry.Through SHG modalities, we can estimate the thickness of the three layers of the zona pellucida, which is also an important indicator in selecting oocytes.Compared with other label-free techniques, HGM can provide combined contrasts of DIC and polar scope and is with a high 3D resolution.The viability test supported the noninvasive nature of the studied HGM.This work indicates that, combining the high 3D resolution and high cell viability properties, SHG and THG modalities can provide various desired contrasts in the whole mouse embryo and the combined HGM could serve a valuable tool in selecting mammalian oocytes and embryos in vitro.

Fig. 1 .
Fig. 1.Depth-resolved section series with combined SHG and THG signals inside a live mouse embryo at the 2-cells stage.(A) -(H) were acquired at imaging depths from 0μm to 63μm relative to (A) with a step of 9μm.The SHG and THG signals are denoted by green and blue colors, respectively.Scale bar: 60μm.

Fig. 2 .
Fig. 2. HGM sectioned images of the in vitro cultured mouse (A) oocyte and (B)-(I) embryos.(A) The SHG signals reveal the spindle fibers (indicated by the yellow arrow) and the zona pellucida (indicated by the white arrow) of the oocyte, while the THG signals reveals the cell membrane, the organelles, and the polar body (indicated by the red arrow).(B) and (C) are images of the same mouse embryo at the 2P stage taken at different depths.The white arrows indicate the two pronuclei.(D)-(H) are images of mouse embryos at the 2-cells, 4-cells, 8-cells uncompacted, 8-cells compacted, and morula stages, respectively.(I) After cavitation, the embryo developed the blastocoel (indicated by the yellow arrow), the trophectoderm (indicated by the white arrow), the inner cell mass (indicated by the red arrow), and turned into a blastocyst.The SHG and THG signals are denoted by green and blue colors, respectively.Scale bar: 60μm.layer with stronger signals and an outer layer with weaker signals, with a dark middle layer between them.For fertilization to occur, a sperm cell must bind to the zona pellucida, penetrate the zona pellucida and the cell membrane of the oocyte, and enter the oocyte cytoplasm, thus forming two pronuclei: one from the sperm and one from the oocyte.Fig. 2(B) and Fig. 2(C) show the HGM images of the same live mouse embryo at the 2P stage optically sectioned at different those of dye fluorescence.Fig.4shows the simultaneously acquired THG and multi-photon fluorescence images of stained oocytes.The first column shows the THG images detected by the forward direction PMT, while the second one is the multi-photon fluorescence images detected by the epi-direction PMT.The third column is the combined THG/fluorescence images.Different rows correspond to oocytes stained by different dyes.(A)-(C) are images of oocytes stained with BODIPY FL C 5 -ceramide, (D)-(F) are with ER-Tracker Blue-White DPX, (G)-(I) are with MitoTracker Red CM-H 2 XRos, (J)-(L) are with LysoSensor Green DND189, and (M)-(O) are with LysoTracker Red Lysosomal Probe, respectively.For the images of the first three rows, the THG signals are better correlated with the fluorescence signals than the latter two columns.

Fig. 6 .
Fig. 6. (A) SHG image of a mouse embryo at the morula stage.(B) Enlarged image corresponding to the squared area in (A), which can distinguish the three layers of zona pellucida.(C) The vertical position dependent SHG intensity after integration along the horizontal direction.Scale bar: 60μm.

Fig. 7 (
A) is an in vitro HGM image of a mouse oocyte optically sectioned through the center of the oocyte.The white arrow shows the excitation light (1230nm) polarization.Fig. 7 (B) and (C) show respectively the IL and OL SHG intensity dependencies as a function of the angle θ between excitation light polarization direction and s ˆ, while s ˆ is just normal to the surface of the zona pellucida .The red lines on top of the experimental data are fitting curves following Eq.(3) with d 33 /d 31 =2.76 in Fig. 7 (B) and d 33 /d 31 =2.05 in Fig. 7 (C).

Fig. 7 .
Fig. 7. (A) In vitro HGM image sectioned through the center of a mouse oocyte.Following (A), (B) and (C) show the SHG intensity of the (B) IL and (C) OL as a function of the angle θ between the laser polarization and the normal vector to the surface of the zona pellucida.Red lines are fitting curves following Eq.(3).Scale bar: 60μm.

Fig 8 .
Fig 8. Born mice of the experiment set.Among the implanted 18 embryos, 12 mice were hatched, with a ratio similar to the control set.

Fig. 9 (
A) shows the image of the oocyte collected from a pubertal mouse (4-6 weeks), which reveals the punctate THG signals with an even organelle distribution.Fig. 9(B) and 9(C) show the in vitro images of the oocytes obtained from two aging mice (> 8 months), where the THG signals reveal non-uniform clustered distribution (indicated by the yellow arrow) of organelles with vacuoles (indicated by the white arrow), which appear throughout the oocyte and are with a size significantly larger than the space between organelles inside the oocyte obtained from pubertal mice (Fig. 9(A)).The organelles even generate an agglomerate of THG signals (indicated by the red arrow) in Fig. 9 (C).Compared with the complete spindles shown in Fig. 9(A), the SHG signals which reveal the abnormal spindles (indicated by the green arrows) in Fig. 9(B) and (C) additionally indicate the poor quality of oocytes obtained from aging female mice [1].
. The peak at 410nm represents the THG signals, the intensities of which were comparable to Normalized nonlinear emission spectra from the organelles labeled by 5 different dyes.The black, red, and blue curves are the spectra measured from the BODIPY FL C 5 -ceramide labeled Golgi apparatus, ER-Tracker Blue-White DPX labeled ER, and MitoTracker Red CM-H 2 XRos labeled mitochondria, respectively.The green and light blue curves are spectra measured from lysosomes labeled by LysoSensor Green DND189 and LysoTracker Red Lysosomal Probe, respectively.The peak at 410nm represents the THG signals.