Label-free DNA imaging in vivo with stimulated Raman scattering microscopy

Label-free DNA imaging is highly desirable in biology and medicine to perform live imaging without affecting cell function and to obtain instant histological tissue examination during surgical procedures. Here we show a label-free DNA imaging method with stimulated Raman scattering (SRS) microscopy for visualization of the cell nuclei in live animals and intact fresh human tissues with subcellular resolution. Relying on the distinct Raman spectral features of the carbon-hydrogen bonds in DNA, the distribution of DNA is retrieved from the strong background of proteins and lipids by linear decomposition of SRS images at three optimally selected Raman shifts. Based on changes on DNA condensation in the nucleus, we were able to capture chromosome dynamics during cell division both in vitro and in vivo. We tracked mouse skin cell proliferation, induced by drug treatment, through in vivo counting of the mitotic rate. Furthermore, we demonstrated a label-free histology method for human skin cancer diagnosis that provides comparable results to other conventional tissue staining methods such as H&E. Our approach exhibits higher sensitivity than SRS imaging of DNA in the fingerprint spectral region. Compared with spontaneous Raman imaging of DNA, our approach is three orders of magnitude faster, allowing both chromatin dynamic studies and label-free optical histology in real time.In vivo imaging of chromatin or chromosome structures and dynamics during vital cellular processes, such as cell division, differentiation, apoptosis, and carcinogenesis, generally relies on the use of either exogenous or endogenous fluorescent labels, the latter of which often involves complicated transgenic organisms (1, 2). A label-free approach, however, allows the visualization of these processes in a noninvasive way in live organisms. In medicine, visualization of nuclear morphology, architecture, size, shape, and mitotic figures provide the most important cytologic features for rendering histologic diagnosis (3, 4). Conventional histology is heavily reliant on tissue biopsies and staining (such as H&E or immunohistochemistry), whereas label-free imaging is able to reveal similar information as that from the stained tissue, and in addition, it allows for a noninvasive characterization and diagnosis of human tissue in real time in vivo.Stimulated Raman scattering (SRS) microscopy offers a contrast mechanism based on Raman spectroscopy, probing the intrinsic vibrational frequencies of chemical bonds or groups (5–8). In SRS microscopy, the collinear pump and Stokes laser beams, at frequencies of ω_p and ω_s, respectively, are tightly focused onto the sample (Fig. 1A). When the frequency difference, ω_p − ω_s, matches a Raman-active molecular vibration, the SRS signal (attenuation to the pump beam or increase on the Stokes beam) is generated through a nonlinear process similar to the stimulated emission. With a highly sensitive detection scheme, involving megahertz modulation transfer, SRS microscopy exhibits orders of magnitude of shorter acquisition time than conventional Raman microscopy (5). Being a nonlinear optical microscopy, it offers 3D sectioning capability with a diffraction-limited spatial resolution. SRS microscopy has been extensively applied to image biomolecules in cells and tissues (9–15).Open in a separate windowFig. 1.Label-free SRS imaging of DNA (magenta), protein (blue), and lipids (green) in live cells. SRS images at three selected Raman shifts in the CH stretching vibrational band were acquired. Linear decomposition was performed with a premeasured calibration matrix to retrieve the distribution of DNA, protein, and lipids. (A) Setup of the SRS microscopy, capable of automatically acquiring images at multiple Raman shifts. This was achieved by synchronizing the tuning of the laser frequency (Lyot filter) to the imaging frame trigger of the microscope. (Inset) Time-lapse images of a HeLa cell undergoing cell division (Movie S1). (B) Raman spectra of DNA, cellular protein, and cellular lipids extracted from HeLa cells. (C) Raman spectrum of the cell pellet. Linear fitting demonstrated that the three compounds in B accounted for ∼90% of the total CH stretching vibration of the cells. (D) SRS images of a live cell in mitotic phase (prophase) at 2,967, 2,926, and 2,850 cm⁻¹, respectively, and the decomposed distribution of DNA, protein, lipids, and the overlay. Chromosomes were visualized with both high contrast and high signal-to-noise ratio. (E) SRS images of a live cell in interphase and the decomposed distribution of DNA, protein, lipids, and the overlay. Detailed internal nuclear features were revealed clearly. (F) Images with SRS and TPEF of a mitotic cell stained with DRAQ5, correlated very well with each other. (Scale bar, 10 μm.)SRS imaging was initially carried out at one Raman shift at a time (5). Recent developments on multiplex detection allow for distinguishing various chemical species with overlapping Raman bands by either broadband excitation (16, 17) or narrowband scanning (18, 19). SRS at two specific Raman shifts within the broadband of the carbon-hydrogen (CH) stretching vibrational mode (2,800–3,050 cm⁻¹) has been used to simultaneously map protein and lipid distribution in cells and tissues (20, 21). In particular, protein and lipid imaging has been applied to delineate brain tumor margins, providing images similar to conventional H&E staining (11). However, SRS does not offer detailed nuclear morphology and architecture, compared with the conventional histology, due to the lack of imaging contrast for DNA.SRS has been demonstrated to be valuable for DNA imaging in cultured cells based on detection of the phosphate peaks within the fingerprint spectral region (22). However, imaging of DNA in this spectral region is difficult for cells in interphase because of the lower DNA density, especially in live tissue. This challenge is also the case for spontaneous Raman imaging (SI Text) (23).Here we demonstrate that, relying on the unique and distinct spectral features of DNA in the CH stretching vibrational region (the high wavenumber range), the distribution of DNA, together with those of protein and lipids, can be mapped by the linear decomposition of images at three optimally selected Raman shifts. This approach offers much higher sensitivity than that of DNA imaging in the fingerprint region, making dynamic imaging of DNA feasible for both mitotic phase and interphase cells in vitro and in vivo.