综述:受激拉曼散射显微技术进行分子振动成像

受激拉曼散射(stimulated Raman scattering, SRS)显微技术发明距今已有十二载,现已成为研究分子振动成像的强大工具。研究人员很早就想方设法去探索如何利用激光脉冲来观察生物标本,然而直到20世纪90年代才出现诸多显微技术,如多光子显微技术、三次谐波(THG)显微技术、二次谐波(SHG)显微技术以及相干反斯托克斯拉曼散射(CARS)显微技术等,并且直到21世纪它们才得到广泛应用。

CARS显微技术具有可提供分子振动信息和远超自发拉曼散射显微技术的成像速度的优点,从而成为了重点研究对象。然而,CARS显微技术也存在一些缺陷,如CARS是一个涉及分子振动共振的四波混频过程,其散射信号中容易混入环境中的非共振四波混频的背景噪声。这种背景噪声不仅限制了灵敏度的提升,而且会干扰CARS使光谱失真,导致其难以被解读。因此,研究人员致力于研究如何避免CARS显微技术中存在的这些问题。

2012年,Yasuyuki Ozeki副教授所在的大阪大学K. Itoh教授课题组提出了使用SRS进行分子振动成像的想法(另外两个独立提出该想法的小组分别是:哈佛大学X. S. Xie教授课题组以及斯图加特大学A. Volkmer教授课题组)。

近日,Yasuyuki Ozeki副教授对SRS显微技术和多色SRS显微技术成像方法的原理以及SRS显微技术的最新应用进行了总结和展望,相关综述作为封面文章发表在Chinese Optics Letters 2020年第12期上(Yasuyuki Ozeki. Molecular vibrational imaging by stimulated Raman scattering microscopy: principles and applications [Invited][J]. Chinese Optics Letters, 2020, 18(12): 121702)。

利用双色光脉冲观测生物分子样本振动光谱特征的受激拉曼散射显微技术所成的分子振动图像

SRS本质上是双色光与分子间的一种相互作用。当双色光(泵浦光和斯托克斯光)的频率差等于分子振动共振频率时,此时由于SRS,泵浦光强度衰减、斯托克斯光强度增强。为了在显微镜下获得SRS信号,首先需要产生同步的泵浦脉冲序列和斯托克斯脉冲序列,并且及时对斯托克斯脉冲序列进行强度调制;然后将两者结合,聚焦于同一个样本上;当SRS发生时,利用锁定检测技术获取斯托克斯脉冲序列转移到泵浦脉冲序列上的强度调制,从而可以获取SRS信号;最后,再通过扫描平台或激光焦斑,即可获得图像。

与CARS不同,SRS并不会受到非共振背景噪声的影响(虽然SRS也存在其他类型的背景噪声),因此基于SRS信号强度与分子浓度成正比的假设,很容易从SRS图像中获取信息。

在过去的十年间,SRS显微技术飞速发展。首先,各种光谱成像方法如雨后春笋般涌现,其中通过观测不同振动频率处SRS的成像可分辨分子类别;其次,各类无需标记的成像手段面世;再者,拉曼探针成像已经问世,只需将氘、炔、腈等分子标记附着于想要研究的生物分子上,即可监测和跟踪各种代谢活动,如生物分子在细胞中的掺入、生化反应和消化等,同时拉曼探针的多路成像也受到了越来越多的关注;最后,成像方式并不局限于显微技术,SRS技术已成功用于内窥镜和流式细胞仪中,这大大扩展了振动成像的应用。

尽管SRS显微技术看似已经发展成熟,但在技术方面仍然有许多问题有待解决,如功能化激光源、功能化拉曼探针、以及诸如量子光学等可提高灵敏度途径的开发。但我们相信通过光学、激光工程、量子光学、化学、生物以及医学等多学科的共同力量,假以时日定可将SRS用于生物医学成像领域。

Molecular imaging with laser pulses

Twelve years have passed since the emergence of stimulated Raman scattering (SRS) microscopy, which is now regarded as a powerful method of molecular-vibrational imaging. At that time, researchers were struggling to explore how we can utilize laser pulses to look into biological specimens. Indeed, multiphoton microscopy, third-harmonic generation (THG) microscopy, second-harmonic generation (SHG) microscopy, and coherent anti-Stokes Raman scattering (CARS) microscopy appeared in the 1990s and their applications were expanding in the 2000s.

In particular, CARS microscopy was extensively investigated in that era because it gives molecular vibrational information while the imaging speed is much faster than spontaneous Raman scattering microscopy. Nevertheless, CARS microscopy has suffered from several limitations, because CARS is a four-wave mixing process involving molecular vibrational resonance and CARS signal is accompanied by a background signal from non-resonant four-wave mixing which is emitted from any substance. This background not only limits the achievable sensitivity but also interferes and distorts the CARS spectrum, which makes it difficult to interpret CARS image. Researchers were seeking ways to overcome these issues of CARS microscopy.

The idea of using SRS for molecular vibrational imaging was independently conceived by three groups (Prof. X. S. Xie's group at Harvard University, Prof. A. Volkmer's group at Stuttgart University, and Yasuyuki Ozeki in Prof. K. Itoh's group at Osaka University). It was published in Chinese Optics Letters, Vol. 18, Issue 12, 2020 (Yasuyuki Ozeki. Molecular vibrational imaging by stimulated Raman scattering microscopy: principles and applications [Invited] [J]. Chinese Optics Letters, 2020, 18(12): 121702).

Caption: Molecular-vibrational imaging by stimulated Raman scattering microscopy utilizes two-color optical pulses to provide the vibrational spectroscopic signature of biomolecules

SRS is an interaction between two-color light and molecules. When the optical frequency difference of the two-color light (called as pump and Stokes) matches the molecular vibrational resonance frequency, the pump light is attenuated and the Stokes light is amplified as a result of SRS. To acquire SRS signal under a microscope, synchronized pump and Stokes pulse trains are generated, and Stokes pulse train is intensity-modulated in time. Then the pump and Stokes pulse trains are combined and focused on a sample. When SRS occurs, the intensity modulation of the Stokes pulse train is transferred to the pump pulse train. The transferred intensity modulation is detected by the lock-in detection technique to obtain the SRS signal. Images are acquired by scanning the stage or laser focal spot.

Different from CARS, SRS does not suffer from the non-resonant background (although there remain different mechanisms that contribute to a background of SRS), and therefore SRS image is easily interpretable because we can assume that SRS signal intensity is proportional to molecular concentration.

In the last decade, there has been tremendous progress on SRS microscopy. First, various methods of spectral imaging have been developed, where SRS images at various vibrational frequencies are acquired to discriminate different molecules. Second, various label-free imaging applications have been explored. Third, Raman probe imaging has been developed, where molecular tags are attached to biomolecules of interest, such as deuterium, alkyne and nitrile. So that we can monitor and track the metabolic activity, such as incorporation, biochemical reaction, and digestion of these biomolecules in cells. Super-multiplex imaging using Raman probes is also attracting attention. Lastly, the imaging modality is not limited to microscopy. The applications of SRS to endoscopy and flow cytometry have been successfully demonstrated, expanding the applications of vibrational imaging.

Although SRS microscopy may look mature, there remain various technical challenges such as the development of functional laser sources, functional Raman probes, and other methods for enhancing the sensitivity such as quantum optics. Such developments require interdisciplinary research including optics, laser engineering, quantum optics, chemistry, biology and medicine, and will lead to unexplored applications in biomedical imaging.