|
1.IntroductionThe myelin sheath is an essential component of the vertebrate nervous system composed of glial membrane layers tightly wrapped around axonal fibers. Its high lipid content allows electrical insulation of the axonal membrane and enables saltatory conduction over long distances ensuring rapid, reliable, and energetically efficient conduction of action potentials. Disruption of the myelin sheath is involved in central nervous system pathologies, such as multiple sclerosis and leukodystrophies, and peripheral nervous system (PNS) neuropathies, such as Charcot-Marie-Tooth disease.1, 2 Quantitative myelin imaging is a necessary tool to follow demyelination or to evaluate eventual remyelinating treatments. Myelin in the white matter can be visualized in vitro and in vivo with magnetic resonance imaging, positron emission tomography, and near-infrared imaging with a (0.1–1)-mm resolution.3, 4, 5, 6 This resolution is insufficient to detect single myelinated axons, thus preventing the study of demyelination in gray matter, where myelinated fibers are sparsely organized. However demyelinating lesions in the gray matter are known to be an important hallmark of pathologies such as multiple sclerosis.7 Different approaches have been recently designed to overcome this limitation. Coherent anti-Stokes–Raman scattering microscopy allows the visualization of single fibers without exogenous labeling.8, 9 This technique has a penetration depth of only a few tens of micrometers and cannot provide quantitative information on large volumes of nervous tissue. Third harmonic generation provides signals from myelin10 with an improved penetration but is still a costly and complicated technique, and its use for in vivo recording in the cortex has not yet been demonstrated. A noninvasive method allowing imaging of myelin with individual fiber resolution is therefore still lacking. Optical coherence tomography (OCT) is a simple yet powerful technique based on interference that allows noninvasive imaging of biological samples and provides optical sectioning deep into scattering tissues.11, 12 Full-field OCT (ff-OCT) is a high spatial resolution version of OCT, the only providing micron resolution in three dimensions.13 OCT uses the backscattered light due to refractive index (n) variations in the sample as a source of endogenous contrast. The lipid-rich myelin sheath has a high refractive index (up to n ∼ 1.455 ± 0.015 in the visible14) and lies in an aqueous background (n ∼ 1.33 in the visible15). Here we show that myelin produces a strong endogenous backscattering signal. To take advantage of this signal, we have designed a prototype of a high-numerical aperture infrared ff-OCT microscope with automatic correction of defocus aberration.16, 17 This prototype, designated as deep-OCM, allows high-resolution myelin imaging deep in tissue without staining. We present two applications of myelin imaging in large volumes of tissue with single fiber resolution: the myelination of individual fibers during development in the rat cortex and the comparison of sciatic nerve structure in wild-type and mutant mice in which myelination in the PNS is impaired. 2.Materials and Methods2.1.AnimalsAll animal manipulation and surgical procedures were in accordance with the European Community guidelines on the care and use of animals,18 French legislation,19 and the recommendations of the CNRS. Male Wistar rats, female Swiss CNP-EGFP mice20 [cyclic nucleotide phosphodiesterase (CNP)], C57BL/6/DBA2 wild-type, and Krox20 −/− mice21, 22 were used in this study. The age of the animals used in the different experiments is specified in the next paragraphs. 2.2.Slice PreparationFor in vivo fixation, animals were deeply anesthetized with urethane (1.5 g/kg) and perfused transcardially with cold phosphate buffer solution (PBS) followed by 4% paraformaldehyde in PBS solution. In the somato-sensory cortex of rats (P19 and P85) and CNP-EGFP mice (P120), 100 μm thick slices were cut tangentially to the flattened cortical surface. In the cerebellum of a P19 rat, 300-μm thick sagittal slices were cut. All slices were stored in PBS before imaging. For myelin immunofluorescence labeling, after bleaching (10% H2O2, 13.3% Dimethyl sulfoxide, 53.3% methanol), cortical slices were incubated with rat monoclonal anti–myelin basic protein (MBP) antibodies (1:100, Millipore Bioscience Research, Billerica, MA) followed by goat anti-rat Cy3 secondary antibodies (1:500, Jackson Immuno Research, Newmarket, UK). All fluorescence images were acquired using a Leica SP2 confocal microscope (Leica Microsystems GmbH, Wetzlar, Germany) with a 10× objective. 2.3.Sciatic Nerve PreparationFor in vitro imaging, the sciatic nerves of two mice (wild-type and Krox20 −/− mutant, P14) were extracted and laid on glass slides in PBS for immediate imaging. The nerves were not held by a coverslip to prevent them from being crushed. For in vivo imaging, a wild-type mouse (P60) was anesthetized with ketamine (100 mg/kg) and xylazine (10 mg/kg). A small incision was performed in the left hind limb and the sciatic nerve was exposed. A plastic holder was slipped under the nerve to maintain it and prevent movements. Imaging was performed in PBS without coverslip. 2.4.Surgical Procedure for In Vivo Cortex ImagingThree three-week old (P20-21; weight ∼80 g), three six-week old (P45-46; weight ∼250 g), and two 12-week old (P86-87; weight ∼400 g) male Wistar rats were used for in vivo somato-sensory cortex imaging. Rats were anesthetized by urethane injection (1.5 g/kg). Supplementary doses of urethane were applied when necessary. The body temperature of the animal was held at 38 °C on a heating blanket during the whole experiment. Animals were maintained in a stereotaxic frame during the surgical procedure. A craniotomy (∼3 mm diameter) centered on the somato-sensory cortex was performed on the rat's left hemisphere. In some cases, the dura mater was removed, as indicated in the figure legends. The craniotomy was sealed with a coverslip with 5 mm diameter that was glued onto the thinned skull to be in close contact with the brain, in order to prevent brain movements during the imaging procedure. For ff-OCT imaging, the animal was maintained with a metal holder glued on its skull to ensure optical stability. At the end of the experiments, rats were injected with a lethal dose of urethane. 2.5.Deep Optical Coherence Microscopy SetupThe microscope [Fig. 1a] consists of a low spatial and temporal coherence light source (250-W halogen lamp) with the standard Köhler microscope illumination, a Linnik interferometer (i.e., a Michelson interferometer with an objective in each arm), a tube lens, and a camera. The water immersion objectives used were either Zeiss IR-Achroplan 40×/0.8 W (to maximize lateral resolution) or Olympus UMPlanF 10×/0.3 W (for the observation of large fields of view). For dispersion correction, a coverslip identical to the one used to cover the sample was placed in the reference arm. To avoid the backreflection from the coverslip on the camera, the former was slightly tilted. A 50-cm doublet was used as a tube lens to image the sample and the reference arm mirror onto the camera. Two different cameras were used: an infrared InGaAs camera (Xeva-1.7-320, Xenics Infrared Solutions, Leuven, Belgium) and a silicon camera (Pantera 1M60, Dalsa, Waterloo, Canada). Myelin fibers were detected with both sensors. Deep-OCM images taken with the InGaAs camera are noted as “IR” in the text (wavelength range 1000–1170 nm, measured at half the peak height), whereas images from the silicon camera are described as deep-red (wavelength range 700–850 nm). Deep-red imaging maximized the lateral resolution, whereas IR imaging increased further the depth penetration due to decreased scattering at longer wavelengths.23 The InGaAs sensor chip had 320 × 256 pixels (pixel size: 30 μm) and the silicon chip had 1024 × 1024 pixels (pixel size: 12 μm). Fast (∼30 Hz) image acquisition was performed in both cases to prevent motion artifacts and loss of contrast due to sample motion. The InGaAs and silicon cameras were operated at 66 and 57 Hz, respectively, triggered by a NI 6722 digital I/O board (National Instruments, Austin, Texas) controlled by the software Light-CT (LLTech, Paris). A synchronized rectangular 33 or 28.5 Hz signal drove the piezo linear stage actuator (PI P-753.11C, Physik Instrumente GmbH, Karlsruhe, Germany) onto which the reference arm mirror was mounted. The piezoactuator amplitude was adapted to produce a change in reference arm length of λ/2, so that the relative phase between both arms of the interferometer would change by π. Taking the difference of subsequent camera images would therefore separate the interference signal from the background light, implementing a two-step phase-stepping protocol. The software Light-CT (LLTech, Paris, France) running on a standard PC allowed real-time display of the infrared images at 33 Hz, while the higher pixel number of the silicon camera limited display speed to ∼15 Hz. In vivo images were taken with the IR camera without any averaging to profit from the 30-ms exposure time. When imaging mechanically stable samples, such as tissue slices, 30 images were often averaged to improve the signal-to-noise ratio, leading to per-frame exposure times of 1.05 and 0.9 s for deep-red and IR images, respectively. Both the sample z position and the relative arm length could be changed with two motorized linear stages (M-VP-25XA, Newport, Irvine, CA). With this setup, the axial resolution (sectioning ability) is mainly due to the coherence length of the source, whereas the lateral resolution is due to the numerical aperture (NA) of the microscope objectives. In the lowest resolution configuration, with 10× objectives and the InGaAs camera (peak of the effective spectrum: λmax = 1100 nm), the lateral sampling of the system was 1.07 μm, the lateral resolution in water 2.2 μm (calculated using the Rayleigh formula 0.6*λmax/NA), and the axial resolution in water 3.3 μm (measured as the full width at half maximum of the interferogram). In the highest-resolution configuration, with 40× objectives and the silicon camera (λmax = 770 nm), the lateral sampling of the system was 0.1 μm, the lateral resolution in water 0.5 μm, and the axial resolution in water 1.15 μm. 2.6.Automatic Compensation of Defocus Aberration Due to Refractive Index MismatchLarge NA makes the setup highly sensitive to depth-dependent defocus induced by refractive index mismatch between the immersion medium in the sample arm and the tissue.16, 17 Refraction at the brain surface indeed causes the actual focus z A of the objective to be shifted deeper into the sample with respect to the nominal focus z N [Fig. 1b], whereas the coherence volume penetrates the tissue slower than the nominal focus [Fig. 1b].17 This spatial separation of coherence volume and actual focus results in a loss in OCT signal: even the small refractive index mismatch between immersion water and brain tissue of ∼2% is sufficient for depth penetration not to be limited by scattering-induced exponential signal decrease, as expected from Beer–Lambert law, but by signal loss due to defocus.17 Refractive index–mismatch compensation was performed online as described previously.16, 17 Briefly, compensation is obtained by keeping the coherence volume at constant depth inside the sample and moving the actual focus into this same position [Fig. 1b]. Because of the aqueous immersion (group index n g), a movement ℓ of the sample objective away from the sample increases the (paraxial) optical path in water by 2n gℓ, while the optical path in air only decreases by 2ℓ. To keep the coherence volume at constant depth in the sample, a displacement ℓ of the sample objective should be therefore associated with a displacement d(ℓ) = ℓ(n g − 1) of the reference arm. The metric used to determine the optimal positions ℓ and d(ℓ) was the total image intensity.17 In the following, we systematically applied automated defocus compensation, increasing the penetration depth by typically a factor of 2.5.17 In these conditions, deep-OCM provides micrometer resolution in the three dimensions and in the depth of the tissue. 2.7.StitchingFor images with a large field of view, a prototype of an ff-OCT microscope (LLTech, Paris, France) was used. It used a 640 × 512 pixel InGaAs Xenics Xeva camera, Olympus UMPlanF 10×/0.3W objectives, and a Märzhäuser xy-translation stage (Märzhäuser Wetzlar GmbH & Co. KG, Wetzlar, Germany). Stitching of individual images with 20% overlap was performed using the ImageJ plugin “Stitch Image Collection” (Stephan Preibisch, Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany). 2.8.QuantificationFor myelin quantification, vertical stacks of images were acquired in confocal immunofluorescence and in deep-OCM (both in vitro and in vivo). The stacks had a total depth of 60–90 μm and were acquired either just below the brain surface (in vivo) or in the first tangential slice from the brain surface (in vitro). The total length of myelin filaments per unitary surface was computed as follows. The length of myelin fibers contained within the depth of field in each independent frame was first measured manually using ImageJ. The average and standard error of the mean of fiber length density per unitary surface was then calculated for 23–37 frames chosen in at least four different stacks in each imaging condition. Images from stacks obtained in rats of a given age were pooled together. Frame sizes were 120 × 120 μm (immunofluorescence confocal), 105 × 105 μm (in vitro 40× deep-red deep-OCM) and 79 × 63 μm (in vivo 40× IR deep-OCM). The volume density of fiber length was obtained from the surface density by taking into account the thickness of each optical section (5.6 μm in confocal microscopy, 1.15 μm in in vitro deep-OCM, and 3.3 μm in in vivo deep-OCM). Fiber diameters, lengths, and interfringe spacings were determined manually using the ImageJ Plot Profile tool. Fiber inclination A was calculated from the fibers’ apparent length L via tan(A) = L/l z, where l z is defined as the full width of the interferogram measured at 10% of the maximum (l z = 2.3 μm for 40× deep-red deep-OCM). Statistical comparisons were performed by standard Student's t-tests. 3.Results3.1.In Vitro Imaging of Individual Myelinated FibersUsing deep-OCM, we were able to observe fibers in the rodent brain tissue. They show up with very high endogenous contrast due to backscattering of the incident light [Figs. 1b, 1c]. Compensation of index-mismatch-induced defocus was mandatory to observe the fibers with high contrast and resolution [Fig. 1b]. In 40× deep-red images of rat cortical slices [Fig. 1c], the observed diameter of the fibers was 0.6 ± 0.1 μm and was therefore at the diffraction limit (microscope lateral resolution: 0.56 μm), in good agreement with the diameter of individual myelinated fibers in the rat brain.24 Note that these fibers appear discontinuous in the images [see, for instance, Fig. 1c]. Three reasons could explain these discontinuities: (i) structural disruptions of the fiber, (ii) speckled aspect of the deep-OCM images, (iii) interference fringes along the fiber's length due to interferences with the mirror in the reference arm. Indeed, two plane waves give interference fringes, with a fringe period directly proportional to the sine of the angle between the two plane waves. In this third case, the inclination A of the fiber relative to the imaging plane should be related to the fringe period I [Figs. 2a, 2b]. The inclination of the fiber can be estimated from its total visible length L over the coherence volume thickness l z (see Sec. 2). We show in Fig. 2c that fringe period I and inclination A are indeed related. Therefore, the discontinuous appearance of the fibers does not have a structural origin, and the fringe spacing allows a quantitative determination of the local fiber inclination. Imaging protocols such as four-phase imaging13 (instead of the two-phase imaging carried out here) should suppress this effect at the cost of slower frame rates, which makes it less appropriate for in vivo imaging. To confirm that the observed fibers were myelinated axons, we compared deep-OCM and confocal fluorescence images of fiber bundles in the same slice of the corpus callosum of a mouse expressing the enhanced green fluorescent protein (EGFP) under the control of the 2′-3′-cyclic nucleotide 3′-phosphodiesterase (CNP) promoter, which causes specific labeling of oligodendrocytes. We observed the same myelin bundles in both images, whereas the cellular bodies of oligodendrocytes were only visible in the fluorescence image [Fig. 1d]. In order to quantify the proportion of myelinated fibers detected by deep-OCM imaging, 40× deep-red images of a fixed rat horizontal cortical slice [Fig. 3b] were compared to confocal images of another cortical slice of the same animal where the myelin basic protein MBP had been immunostained [Fig. 3a]. This quantization was performed for rats of two different ages (P19 and P85). The fiber densities measured with the two different methods were not significantly different for 12-week-old animals [Student's t-test, P = 0.39, Fig. 3c]. In the case of three-week-old animals, no fibers were detected by deep-OCM imaging, whereas a very small number could be detected after immunostaining (∼3% of the fiber density in adults). These few patchy myelinated zones may have been missed due to the smaller field of view of the deep-OCM images. These results indicate that this technique can be used for a quantitative analysis of myelinated axons. Myelinated fibers could also be detected by deep-OCM in other brain structures over large fields of view. We reconstructed a complete image of a fixed rat cerebellum sagittal slice with micron resolution by stitching consecutive fields of view [10× IR, 8 × 8 mm, Fig. 4a]. The stereotyped organization of the cerebellar cortex allows unequivocal identification of the different structures. Because of its rich myelin content, the white matter was the most visible structure in the cerebellum. The molecular and granular layers appeared clearly distinct, and the Purkinje cell somata were visible as hollow structures [Fig. 4c]. Individual myelinated fibers could be readily observed in the cerebellar white matter [Fig. 4d], beyond the depth sectioning ability of standard bright-field microscopy. In fact, a small z projection of deep-OCM images [Fig. 4b] is necessary to visualize the full length of fibers, which can be followed from one image frame to the next in 3-D stacks of deep-OCM images (Video 1). 10.1117/1.3650770.110.1117/1.3650770.2Myelin imaging was also performed in the peripheral nervous system. Deep-OCM 3-D reconstruction was used to obtain the structure of an acutely dissected mouse sciatic nerve in vitro, in which individual myelinated fibers could be discriminated [Fig. 5a, Video 2]. Deep-OCM imaging allows optical reconstruction of a large fraction (about one-half) of the whole nerve [Fig. 5a] with micrometer resolution. In order to check our ability to evaluate pathological demyelinating situations, we imaged the sciatic nerves of Krox20 mutant mice. Krox20 is a transcription factor whose inactivation in Schwann cells results in impaired myelination of the PNS.21, 22 We observed a strong difference between nerves of Krox20 mutant and wild-type mice [Fig. 5b]. Indeed, individual fibers could not be discriminated in mutant animals. Moreover, in identical imaging conditions, the reflected light amplitude was ∼40% stronger in wild-type nerves than in mutant nerves. Thus, deep-OCM can be used to evaluate pathological myelin alteration in peripheral nerves. 3.2.In Vivo Imaging of Individual Myelinated FibersWe tested whether sciatic nerve deep-OCM imaging could be performed in vivo in an anesthetized wild-type mouse, using minimal surgery to expose the nerve and stabilize it [Fig. 6a]. This was necessary to prevent respiratory or heart beat movements. In vitro and in vivo images were of comparable quality [Fig. 6b versus 5a]. Finally, we performed in vivo deep-OCM myelin imaging in the somato-sensory cortex of anesthetized rats. The animals were maintained in the imaging system by a tight head fixation to prevent movement during image acquisition [Fig. 7a]. The use of IR imaging and of refractive index–mismatch correction is particularly important in deep in vivo imaging, where they considerably improve the contrast and resolution of images. Two configurations were tested. In the first situation, the dura was removed, which potentially increases brain motion artifacts. In the second situation, the dura was preserved, despite its large backscattering. Detection of myelin fibers in the rat cortex was possible in both conditions [Figs. 7b, 7c]. In the absence of dura, we tested the maximum depth at which imaging could be performed. Between 10 and 640 μm depth, four OCT images were taken every 10 μm and then averaged. Figure 7c shows the normalized images at depths 20, 100, 180, 260, and 340 μm. Individual filaments could be observed up to a depth of ∼340 μm. Figure 7d shows the maximum OCT signal as a function of depth after applying a Gaussian smoothing filter with σ = 1 μm to reduce noise. Noise level is reached at depths of >380 μm, which could be either due to the sensitivity of the microscope or to the absence of horizontal myelinated fibers at this depth in cortex. Using deep-OCM, we followed the developmental profile of myelinated fibers from 3 to 12 weeks [Figs. 8a, 8b]. Myelinated axons were observed in vivo up to 340 μm deep into the somato-sensory cortex of 6- and 12-week-old animals [Fig. 8a, middle and right], whereas no fibers were observed in the cortex of 3-week-old animals [Fig. 8a, left]. The density of tangential fibers in the first 150 μm below the brain surface was consistent with the one measured in vitro [Figs. 8b, 3]. It was similarly high for 6- and 12-week-old rats [Student's t-test, P = 0.42, Fig. 8b], and was reduced in both cases at levels deeper than 150 μm, in good agreement with previous histological studies (Refs. 25, 26, 27). 4.DiscussionIn this study, we have shown that deep-OCM allows quantitative imaging of individual myelinated fibers in vitro and in vivo in the rodent nervous system, with micron-size resolution, without labeling and over large volumes of tissue, up to 340 μm deep into the tissue. Deep-OCM combines several technical advances compared to standard ff-OCT microscopy, such as fast acquisition rate, IR imaging, large numerical aperture microscope objectives, and real-time defocus correction.16, 17, 23 Deep-OCM imaging of myelinated fibers benefits from their large backscattered signal. This may arise from the high local refractive index of fibers due to their high lipid content. However, whereas imaging at moderate depth was possible in a standard configuration using deep-red light, imaging at a greater depth was only possible by combining infrared detection with the correction of defocus aberration due to refractive index mismatch between immersion medium and brain tissue. Without defocus correction, both signal strength and resolution quickly degrade with depth. Finally, imaging at high temporal resolution with fast cameras in addition to a tight tissue mechanical fixation was required to avoid motion artifacts due to breathing or heart beat. In the PNS, fast image acquisition of the myelin structure of sciatic nerves at high resolution was achieved both in vitro and in vivo, and defects of the myelin sheath observed in the raw images could be easily quantified from the reflected amplitude. Deep-OCM also allowed us to record images of individual myelinated fibers in vivo within the somato-sensory cortex of anesthetized rats, without exogenous labeling, and either when the dura matter was removed or directly below the intact dura mater. Quantification of myelinated fiber density was obtained from the image analysis. Similar axonal densities have been measured in the superficial layers of the rat cortex.28 By using this approach, we could observe the developmental appearance of myelin structures in the adult rat brain. Compared to other techniques that work in this resolution and penetration depth regime, such as two-photon microscopy of genetically labeled oligodendrocytes, deep-OCM has the advantage that it directly images the local increase of refractive index caused by the high lipid content of the myelin sheath. Scanned OCM or confocal reflection microscopy (two techniques relying also on endogenous backscattering) could be similarly relevant for myelin imaging. However, confocal microscopy alone has inferior rejection of light reflected back far from the focal plane when compared to an OCT or OCM system29, 30 and is therefore unlikely to provide comparable penetration depths. Only a combination of confocal detection with scanning OCM improves depth discrimination,29, 30 at the expense of scanning in all three dimensions, which might make such a system prohibitively slow for in vivo applications. In comparison, deep-OCM is intrinsically fast as it only requires mechanical z scanning. Deep-OCM should allow extensive and even chronic imaging without the risk of signal deterioration due to bleaching, as would be the case in two-photon microscopy. Such chronic imaging of individual myelinated fibers in the upper layers of the rat cortex could be of considerable interest in neuropathies progression studies. Whole-brain imaging techniques, such as MRI, are too limited in their resolution to provide any information on the single-fiber level. Our observations open perspectives of application to chronic experiments in rodents, notably for the longitudinal study of cortical demyelination and remyelination. To our knowledge, no other existing technique gives access to this information. The noninvasiveness, the speed, and the depth of penetration of deep-OCM could make it a potentially interesting technique for human diagnostics. Because the access to the peripheral nervous system is quite simple and requires only minor surgery, this method may be used, in the future, in human patients with peripheral neuropathies. Deep-OCM-based myelin imaging of the sciatic nerve could replace nerve biopsy, which is a standard procedure for diagnosing certain neuropathies. It could thereby help avoid neuropathic side effects associated with nerve biopsies. More studies will be necessary to demonstrate that myelinated fibers in the PNS of human patients are as readily accessible to deep-OCM imaging as shown for rats here, and that mechanically stable recordings can be performed. Investigations of the human central nervous system with single-fiber resolution will probably require further development to attain interesting depth, for instance, using endoscopy. In conclusion, deep-OCM makes a fine-scale analysis of the myelination processes possible, both in the cerebral cortex and in the peripheral nervous system. Such analysis currently requires time-consuming and invasive histological methods. Our technique is relatively noninvasive, affordable, and fast, and provides real-time imaging without complex off-line image treatment. It opens promising possibilities for in vivo acute or chronic myelin imaging as well as for the diagnosis of myelin pathologies. AcknowledgmentsWe thank Marie-Stéphane Aigrot, Bruno Stankoff, and Catherine Lubetzki for their helpful comments and for transferring the transgenic CNP-GFP mice. We also thank Vittorio Gallo for authorizing us to use the mice. We thank Charles Brossollet (LLtech, Paris, France), Benjamin Mathieu, and Gaspard Gerschenfeld for their help and Shawn Mikula for critical discussions. This work was supported by ANR RIB Grants No. MICADO ANR-07-RIB-010-02 and No. ANR-07-RIB-010-04. J.B. was funded by a PhD fellowship from Fondation Pierre-Gilles de Gennes. ReferencesS. S. Scherer and
L. Wrabetz,
“Molecular mechanisms of inherited demyelinating neuropathies,”
Glia, 56
(14), 1578
–1589
(2008). https://doi.org/10.1002/glia.20751 Google Scholar
N. Baumann and
D. Pham-Dinh,
“Biology of oligodendrocyte and myelin in the mammalian central nervous system,”
Physiol. Rev., 81
(2), 871
–927
(2001). Google Scholar
C. Laule,
I. M. Vavasour,
S. H. Kolind,
D. K. Li,
T. L. Traboulsee,
G. R. Moore, and
A. L. MacKay,
“Magnetic resonance imaging of myelin,”
Neurotherapeutics, 4
(3), 460
–484
(2007). https://doi.org/10.1016/j.nurt.2007.05.004 Google Scholar
S. Mori and
J. Zhang,
“Principles of diffusion tensor imaging and its applications to basic neuroscience research,”
Neuron, 51
(5), 527
–539
(2006). https://doi.org/10.1016/j.neuron.2006.08.012 Google Scholar
B. Stankoff,
Y. Wang,
M. Bottlaender,
M. S. Aigrot,
F. Dolle,
C. Wu,
D. Feinstein,
G. F. Huang,
F. Semah,
C. A. Mathis,
W. Klunk,
R. M. Gould,
C. Lubetzki, and
B. Zalc,
“Imaging of CNS myelin by positron-emission tomography,”
Proc. Natl. Acad. Sci. U.S.A., 103
(24), 9304
–9309
(2006). https://doi.org/10.1073/pnas.0600769103 Google Scholar
C. Wang,
C. Wu,
D. C. Popescu,
J. Zhu,
W. B. Macklin,
R. H. Miller, and
Y. Wang,
“Longitudinal near-infrared imaging of myelination,”
J. Neurosci., 31
(7), 2382
–2390
(2011). https://doi.org/10.1523/JNEUROSCI.2698-10.2011 Google Scholar
J. J. Geurts,
S. D. Roosendaal,
M. Calabrese,
O. Ciccarelli,
F. Agosta,
D. T. Chard,
A. Gass,
E. Huerga,
B. Moraal,
D. Pareto,
M. A. Rocca,
M. P. Wattjes,
T. A. Yousry,
B. M. Uitdehaag, and
F. Barkhof,
“Consensus recommendations for MS cortical lesion scoring using double inversion recovery MRI,”
Neurology, 76
(5), 418
–424
(2011). https://doi.org/10.1212/WNL.0b013e31820a0cc4 Google Scholar
Y. Fu,
T. B. Huff,
H. W. Wang,
H. Wang, and
J. X. Cheng,
“Ex vivo and in vivo imaging of myelin fibers in mouse brain by coherent anti-Stokes Raman scattering microscopy,”
Opt. Express, 16
(24), 19396
–19409
(2008). https://doi.org/10.1364/OE.16.019396 Google Scholar
J. Imitola,
D. Cote,
S. Rasmussen,
X. S. Xie,
Y. Liu,
T. Chitnis,
R. L. Sidman,
C. P. Lin, and
S. J. Khoury,
“Multimodal coherent anti-Stokes Raman scattering microscopy reveals microglia-associated myelin and axonal dysfunction in multiple sclerosis-like lesions in mice,”
J. Biomed. Opt., 16
(2), 021109
(2011). https://doi.org/10.1117/1.3533312 Google Scholar
M. J. Farrar,
F. W. Wise,
J. R. Fetcho, and
C. B. Schaffer,
“In vivo imaging of myelin in the vertebrate central nervous system using third harmonic generation microscopy,”
Biophys. J., 100 1362
–1371
(2011). https://doi.org/10.1016/j.bpj.2011.01.031 Google Scholar
J. G. Fujimoto,
“Optical coherence tomography for ultrahigh resolution in vivo imaging,”
Nat. Biotechnol., 21
(11), 1361
–1367
(2003). https://doi.org/10.1038/nbt892 Google Scholar
J. G. Fujimoto,
M. E. Brezinski,
G. J. Tearney,
S. A. Boppart,
B. Bouma,
M. R. Hee,
J. F. Southern and
E. A. Swanson,
“Optical biopsy and imaging using optical coherence tomography,”
Nat. Med., 1
(9), 970
–972
(1995). https://doi.org/10.1038/nm0995-970 Google Scholar
L. Vabre,
A. Dubois, and
A. C. Boccara,
“Thermal-light full-field optical coherence tomography,”
Opt. Lett., 27
(7), 530
–532
(2002). https://doi.org/10.1364/OL.27.000530 Google Scholar
I. P. Antonov,
A. V. Goroshkov,
V. N. Kalyunov,
I. V. Markhvida,
A. S. Rubanov, and
L. V. Tanin,
“Measurement of the radial distribution of the refractive index of the Schwann's sheath and the axon of a myelinated nerve fiber in vivo,”
J. Appl. Spectrosc., 39
(1), 822
–824
(1983). https://doi.org/10.1007/BF00662830 Google Scholar
M. Daimon and
A. Masumura,
“Measurement of the refractive index of distilled water from the near-infrared region to the ultraviolet region,”
Appl. Opt., 46
(18), 3811
–3820
(2007). https://doi.org/10.1364/AO.46.003811 Google Scholar
S. Labiau,
G. David,
S. Gigan, and
A. C. Boccara,
“Defocus test and defocus correction in full-field optical coherence tomography,”
Opt. Lett., 34
(10), 1576
–1578
(2009). https://doi.org/10.1364/OL.34.001576 Google Scholar
J. Binding,
J. Ben Arous,
J.-F. Léger,
S. Gigan,
C. Boccara, L. Bourdieu,
“Brain refractive index measured in vivo with high-NA defocus-corrected full-field OCT and consequences on two-photon microscopy,”
Opt. Express, 19
(6), 4833
–4847
(2011). https://doi.org/10.1364/OE.19.004833 Google Scholar
Directive 86/609/CEE, CE official journal L358 (December 18,
(1986) Google Scholar
Décret No. 97/748, 19 October, 1987, J. O. République française (October 20,
(1987) Google Scholar
X. Yuan,
R. Chittajallu,
S. Belachew,
S. Anderson,
C. J. McBain, and
V. Gallo,
“Expression of the green fluorescent protein in the oligodendrocyte lineage: a transgenic mouse for developmental and physiological studies,”
J. Neurosci. Res., 70
(4), 529
–545
(2002). https://doi.org/10.1002/jnr.10368 Google Scholar
P. Topilko,
S. Schneider-Maunoury,
G. Levi,
A. Baron-Van Evercooren,
A. B. Chennoufi,
T. Seitanidou,
C. Babinet, and
P. Charnay,
“Krox-20 controls myelination in the peripheral nervous system,”
Nature, 371
(6500), 796
–799
(1994). https://doi.org/10.1038/371796a0 Google Scholar
F. Coulpier,
L. Decker,
B. Funalot,
J.-M. Vallat,
F. Garcia-Bragado,
P. Charnay, and
P. Topilko,
“CNS/PNS boundary transgression by central glia in the absence of Schwann cells or Krox20/Egr2 function,”
J. Neurosci., 30
(17), 5958
–5967
(2010). https://doi.org/10.1523/JNEUROSCI.0017-10.2010 Google Scholar
D. Sacchet,
J. Moreau,
P. Georges, and
A. Dubois,
“Simultaneous dual-band ultra-high resolution full-field optical coherence tomography,”
Opt. Express, 16
(24), 19434
–19446
(2008). https://doi.org/10.1364/OE.16.019434 Google Scholar
G. Partadiredja,
R. Miller, and
D. E. Oorschot,
“The number, size, and type of axons in rat subcortical white matter on left and right sides: a stereological, ultrastructural study,”
J. Neurocytol., 32
(9), 1165
–1179
(2003). https://doi.org/10.1023/B:NEUR.0000021910.65920.41 Google Scholar
S. Mikula,
I. Trotts,
J. M. Stone, and
E. G. Jones,
“Internet-enabled high-resolution brain mapping and virtual microscopy,”
NeuroImage, 35
(1), 9
–15
(2007). https://doi.org/10.1016/j.neuroimage.2006.11.053 Google Scholar
I. Trotts,
S. Mikula, and
E. G. Jones,
“Interactive visualization of multiresolution image stacks in 3D,”
NeuroImage, 35
(3), 1038
–1043
(2007). https://doi.org/10.1016/j.neuroimage.2007.01.013 Google Scholar
J. T. Eayrs and
B. Goodhead,
“Postnatal development of the cerebral cortex in the rat,”
J. Anat., 93
(Pt 4), 385
–402
(1959). Google Scholar
J. A. Izatt,
M. R. Hee,
G. M. Owen,
E. A. Swanson, and
J. G. Fujimoto,
“Optical coherence microscopy in scattering media,”
Opt. Lett., 19
(8), 590
–592
(1994). https://doi.org/10.1364/OL.19.000590 Google Scholar
J. Mertz, Introduction to Optical Microscopy, Roberts, Greenwood Village, CO
(2010). Google Scholar
|