For in vivo deep tissue imaging, high order wavefront measurement and correction is needed for handling the severe wavefront distortion. Towards such a goal, we have developed the iterative multi-photon adaptive compensation technique (IMPACT). In this work, we explore using IMPACT to perform calcium imaging of neocortex through the intact skull of adult mice, and to image through the highly scattering white matter on the hippocampus surface.
© 2015 Optical Society of America
In vivo imaging has become a routine practice in many biology research fields [1, 2]. The rapid development of powerful fluorescence proteins has given optical imaging its unique advantages . For example, the continuous progress of calcium indicator in neuroscience has allowed the scientists to observe the dynamics of neural networks of awake and behaving animals, which provides insight that is difficult to acquire with other imaging techniques . However, the refractive index inhomogeneity in biological tissue fundamentally limits the achievable imaging depth and quality. Different from fixed sample imaging, one cannot rely on optical clearing to reduce the refractive index inhomogeneity and improve the imaging depth . How to noninvasively perform in vivo imaging at high spatial (sub-micron) and temporal (sub-second) resolutions at large depth remains an open question for the imaging tool developers.
Long optical wavelength has shown great promises for deep tissue imaging . The refractive index inhomogeneity is reduced at longer wavelength. As a result, light can travel further into tissue before being completely diffused. In practice, the available fluorescence proteins typically define the usable wavelengths. At a given wavelength, we can further improve the imaging performance by measuring and controlling the optical wavefront [6–23]. Inherently, the focus distortion including the reduced Strehl ratio originates from the optical wavefront distortion. Therefore, wavefront correction can help restore the focus quality and signal strength.
A variety of wavefront measurement and correction methods have been developed in the past [6, 7, 17–20, 24]. Many have borrowed the ideas from the research field of astronomical adaptive optics (AO). In conventional AO, a wavefront sensor directly measures the emission from the laser focus. However, in multi-photon microscopy, the emission and the excitation are typically far apart in spectrum and their correction wavefront could be rather different. A different scheme is to abandon the idea of wavefront sensor and use wavefront modulation to figure out the correct wavefront. In this way, the wavefront is determined regardless of the emission wavelength, which works well for multi-photon microscopy and is the also scheme we employed in our lab [17, 24].
In highly turbid tissue, the wavefront distortion is rather complex. High order wavefront measurement and correction are needed. Towards this goal, we have developed IMPACT that takes advantage of iterative feedback and the nonlinearity inherent in the multi-photon signals to rapidly measure and compensate wavefront distortions [17, 24]. In the recent development , we have shown that it is possible to directly image the neocortex through the intact skull of adult mice. It was the high order wavefront correction that made such a task possible. In this work, we took one step further and used IMPACT to enable calcium imaging through the intact skull of adult mice, which provides a noninvasive solution to study the neural activity at sub-micron resolutions. Hippocampus is a major component of mammalian brain and plays an important role in memory formation. Its surface has a dense layer of white matter composed of axons and is highly scattering. In this work, we also explored using IMPACT to image hippocampus through the scattering alveus layer.
Figure 1 shows the schematic drawing of the imaging system. Basically, we used relay lenses to image the galvo scanner onto a segmented MEMS mirror and then onto the pupil of a NA 0.8 16x water dipping objective. The principle and operation procedure have been thoroughly discussed in the previous reports [17, 24]. The laser source was a 80 MHz 935 nm 140 fs Ti:Sapphire oscillator (Chameleon, Coherent, CA, USA).
A major challenge of calcium imaging is that the fluorescence intensity fluctuates over time as a result of the neural activity. To provide a reliable fluorescence signal for the IMPACT measurement, we injected a mixture of two types of viruses to the animal. One is for the green calcium indicator and the other is for expressing a red fluorescence protein that is used for the IMPACT measurements. Specifically, we used Rbp4-cre mice (cre-dependent for layer 5 neurons) and the mixture of AAV2/1-Flex-Syn-GCaMP6f and AAV2/1-Flex-Syn-dsRed1.2 viruses. After anesthesia, we made an incision on the mouse skin around the target coordinates (1.5 mm lateral and 2.8 mm caudal to bregma) to expose the skull and injected a 30 nL mixed viruses (ratio = 1:1) to the S1 cortex at 0.7 mm depth. After 2-3 weeks’ viral expression, we used the mice for imaging. During the imaging, we exposed the skull and attached a head post to the skull surface using dental acrylic.
For the hippocampus imaging, we used transgenic Cx3Cr1-GFP × Thy1-YFP mice. We first anesthetized the mice and then implanted the hippocampal window (1.7 mm lateral, 2.1 mm caudal to bregma) . We typically allow 2-3 recovery weeks before the in vivo imaging. All procedures involving mice were approved by the Animal Care and Use Committees of HHMI Janelia Research Campus.
3. Calcium imaging of neocortex through the intact skull of adult mice
Noninvasively monitoring the brain activity at sub-cellular resolution is a challenging task. The typical procedure is to perform craniotomy to remove the skull and implant a cranial window. However, the craniotomy may disturb the cortex tissue and cause inflammation. An alternative solution is to thin the skull although the remaining skull may still cause wavefront distortion. Moreover, the bone tissue may continue to grow near the thinning site, which can cause problems for longitudinal studies. With the high order wavefront correction provided by IMPACT, we can directly image the neocortex through the intact skull of adult mice. Here we use IMPACT to perform through skull calcium imaging [Fig. 2(a)].
As the calcium signal from GCaMP fluctuates over time as a result of neural activity, we used the stable signal of a different color (dsRed) for IMPACT measurement. Both labels can be excited efficiently at 935 nm. The thickness of the skull was ~120 µm determined by the second harmonic signal from the bone. In Media 1, we show the images of dendrites and spines at 168-197 µm under the surface of the skull, with system correction and full correction (full correction = system correction + sample correction). With only system correction, the fine structural information was lost due to the strong wavefront distortion of the skull. With full correction, we could resolve the dendrites and spines, as shown in Fig. 2(b) and 2(c). The high order wavefront distortion was shown in Fig. 2(e). We imaged the spontaneous neural activity at 14 Hz with 168 mW laser power (Media 2), and the calcium dynamics are shown in Fig. 2(f). Signals from both dendrites (e.g. ROI 8) and spines (e.g. ROI 9) were recorded through the intact skull. The compensation pattern was effective for at least 5 hours during the imaging.
4. Hippocampus imaging
Hippocampus is typically more than 1 mm below the cortical surface in mice. Direct optical imaging remains a challenge. Although long excitation wavelength can directly image the top layer CA1 cells, the obtained resolution is still insufficient to resolve sub-micron neural structures. Moreover, the majority of the hippocampus remains inaccessible. For hippocampus imaging, the typical procedure is to aspirate the cortex tissue above the area of interest in the hippocampus and implant the window . However, the image quality still suffers from the tissue induced wavefront distortion, especially from the alveus that is mainly composed of white myelinated fibers running parallel to the surface.
We chose the transgenic Cx3Cr1-GFP × Thy1-YFP (H line) mice for this study, in which the neurons express YFP and the microglia express GFP. In experiments, we allowed 2-3 weeks for the mice to fully recover from the window implanting. The phenotype of the microglia appeared normal, suggesting that the hippocampus tissue remained healthy. We imaged the hippocampus from the alveus to the dentate gyrus, as shown in Fig. 3(a). The alveus was ~35 µm thick, the CA1 cells were at ~115-160 µm depth, and the dentate granule cells were at ~580-635 µm depth. The myelinated axons at alveus [e.g. shown in Fig. 3(b)] caused strong light attenuation. We employed IMPACT to image the dendrites and spines of the CA1 pyramidal neurons at ~391-400 µm below the hippocampal surface (Media 3). The dendritic spines were hardly resolvable [Fig. 3(e)] with only system compensation [Fig. 3(c)]. With full compensation [Fig. 3(d)], the image contrast and resolution [Fig. 3(f)] were both improved. A comparison is shown in Fig. 3(g) in which the signal along the dash line in Fig. 3(e) is plotted.
We also imaged the dentate granule cells of dentate gyrus at 570-588 µm depth (Media 4). The images with system compensation and full compensation are shown in Fig. 3(h) and 3(i), respectively. The full compensation profile is shown in Fig. 3(j). We compared the signal intensity along the dash line in Fig. 3(h) and show the plot in Fig. 3(k).
5. Discussion and conclusions
These in vivo imaging results demonstrate the benefit of our wavefront correction method. Especially in the case of imaging through intact skull, the structures only became visible when IMPACT was applied. It was the high order wavefront correction that made such an imaging task possible. The method provides a solution to image brain with minimal perturbation.
A major challenge for measuring wavefront in calcium imaging is that the fluorescence indicator may fluctuate over time. Besides the green fluorescence calcium indicator, we applied a second red label, which can be excited at the same wavelength of GCaMP, to provide a stable signal source for the wavefront measurement. With longer emission wavelength, the red label works even better than the typical GFP label as its emission suffers less scattering and blood absorption. As IMPACT is independent of the emission wavelength, the wavefront measured with the red label works equally well for the calcium indicator, which is different from the wavefront sensor based methods. This feature also allows IMPACT to be used for even longer excitation wavelength.
There are two limitations of the present system. One is the small field of view after a single wavefront correction. Tiling can help form a larger field of view . For neuroimaging, it is beneficial to have a simultaneous correction over a wide area. For studying isolated points such as spatially isolated neuronal somas, we can potentially combine IMPACT with the random access measurements based on acoustic optical deflectors . We will first define the ROIs in 3D and measure their corresponding correction wavefront. The switching of the wavefront takes ~100 µs. We can take the following 100 µs to sample a few points on the cell of interest before moving onto the next position. In this way, it is possible to sample ~500 cells at a 10 Hz rate. To achieve a continuous wide correction area, we need to abandon the widely used pupil plane correction and employ the multi-conjugate configuration, as discussed in our recent study . The other limitation is that we have only provided phase compensation. When imaging at large depth, especially with high NA objective (NA ≥ 1), the path length inside the tissue for the peripheral rays can be much longer than that of the central rays. Consequently, the peripheral rays experience much stronger attenuation and the effective NA at the focal plane decreases. For studying larger structures such as the soma, the reduced NA is acceptable. But for high resolution imaging of fine structures such as the dendritic spines, we should also apply amplitude compensation to restore the diffraction limited focus. In fact, during the IMPACT measurement, we obtained both the phase and amplitude at each pixel. Therefore, the amplitude profile is available and we just need to image the amplitude shaping device (e.g. a SLM configured for amplitude modulation) onto the MEMS to perform the simultaneous phase and amplitude correction.
In summary, we report the latest in vivo imaging results with IMPACT. The results suggest that IMPACT can significantly improve the imaging performance through highly turbid tissue. With dual labeling, we can now apply IMPACT to calcium imaging, which opens up new opportunities for deep tissue neuroimaging.
The research is supported by the Howard Hughes Medical Institute.
References and links
2. T. Lämmermann, P. V. Afonso, B. R. Angermann, J. M. Wang, W. Kastenmüller, C. A. Parent, and R. N. Germain, “Neutrophil swarms require LTB4 and integrins at sites of cell death in vivo,” Nature 498(7454), 371–375 (2013). [CrossRef] [PubMed]
3. H. Dana, T.-W. Chen, A. Hu, B. C. Shields, C. Guo, L. L. Looger, D. S. Kim, and K. Svoboda, “Thy1-GCaMP6 transgenic mice for neuronal population imaging in vivo,” PLoS ONE 9(9), e108697 (2014). [CrossRef] [PubMed]
5. 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] [PubMed]
7. X. Tao, J. Crest, S. Kotadia, O. Azucena, D. C. Chen, W. Sullivan, and J. Kubby, “Live imaging using adaptive optics with fluorescent protein guide-stars,” Opt. Express 20(14), 15969–15982 (2012). [CrossRef] [PubMed]
8. O. Katz, P. Heidmann, M. Fink, and S. Gigan, “Non-invasive single-shot imaging through scattering layers and around corners via speckle correlations,” Nat. Photonics 8(10), 784–790 (2014). [CrossRef]
10. O. Katz, E. Small, and Y. Silberberg, “Looking around corners and through thin turbid layers in real time with scattered incoherent light,” Nat. Photonics 6(8), 549–553 (2012). [CrossRef]
11. C. L. Hsieh, Y. Pu, R. Grange, G. Laporte, and D. Psaltis, “Imaging through turbid layers by scanning the phase conjugated second harmonic radiation from a nanoparticle,” Opt. Express 18(20), 20723–20731 (2010). [CrossRef] [PubMed]
14. Y. M. Wang, B. Judkewitz, C. A. Dimarzio, and C. Yang, “Deep-tissue focal fluorescence imaging with digitally time-reversed ultrasound-encoded light,” Nat. Commun. 3, 928 (2012). [CrossRef] [PubMed]
17. J. Tang, R. N. Germain, and M. Cui, “Superpenetration optical microscopy by iterative multiphoton adaptive compensation technique,” Proc. Natl. Acad. Sci. U.S.A. 109(22), 8434–8439 (2012). [CrossRef] [PubMed]
19. D. Débarre, E. J. Botcherby, T. Watanabe, S. Srinivas, M. J. Booth, and T. Wilson, “Image-based adaptive optics for two-photon microscopy,” Opt. Lett. 34(16), 2495–2497 (2009). [CrossRef] [PubMed]
20. S. G. Adie, B. W. Graf, A. Ahmad, P. S. Carney, and S. A. Boppart, “Computational adaptive optics for broadband optical interferometric tomography of biological tissue,” Proc. Natl. Acad. Sci. U.S.A. 109(19), 7175–7180 (2012). [CrossRef] [PubMed]
22. D. B. Conkey, A. M. Caravaca-Aguirre, and R. Piestun, “High-speed scattering medium characterization with application to focusing light through turbid media,” Opt. Express 20(2), 1733–1740 (2012). [CrossRef] [PubMed]
23. K. Wang, D. E. Milkie, A. Saxena, P. Engerer, T. Misgeld, M. E. Bronner, J. Mumm, and E. Betzig, “Rapid adaptive optical recovery of optimal resolution over large volumes,” Nat. Methods 11(6), 625–628 (2014). [CrossRef] [PubMed]
25. D. A. Dombeck, C. D. Harvey, L. Tian, L. L. Looger, and D. W. Tank, “Functional imaging of hippocampal place cells at cellular resolution during virtual navigation,” Nat. Neurosci. 13(11), 1433–1440 (2010). [CrossRef] [PubMed]
26. G. Katona, G. Szalay, P. Maák, A. Kaszás, M. Veress, D. Hillier, B. Chiovini, E. S. Vizi, B. Roska, and B. Rózsa, “Fast two-photon in vivo imaging with three-dimensional random-access scanning in large tissue volumes,” Nat. Methods 9(2), 201–208 (2012). [CrossRef] [PubMed]
27. M. Cui, “A numerical study of multi-conjugate large area wavefront correction for deep tissue microscopy ” submitted (2015).