Mapping Cancer Metabolism by Mid-infrared Photothermal Microscopy - Program Summary While altered cell metabolism is emerging as a hallmark of cancer, there is an unmet need for new tools for quantitation of metabolites. NMR spectroscopy, mass spectrometry, FTIR, and Raman spectroscopy are widely used for molecular detection in tissue extracts or intact tissues. Yet, these tools do not indicate the spatial locations of the analytes inside the cell. We address this unmet need via development of a lock-in free, wide- field mid-infrared photothermal (MIP) microscope. Our technology will enable quantitative vibrational imaging of metabolites in live tumor cells and intact biopsies. In MIP microscopy recently developed in the PI lab (Sci Adv 2016), a visible beam probes the thermal effect (e.g. change of refractive index and thermal expansion) induced by a pulsed infrared beam. The MIP signal is then extracted through a lock-in amplifier. To match the IR/visible illumination area, the PI lab further developed a wide-field MIP microscope in which a complementary metal– oxide–semiconductor (CMOS) camera and synchronization electronics are harnessed for whole-field lock-in detection (Sci Adv 2019). Despite these initial successes, the sensitivity of MIP microscopy is limited by the detection schemes. First, the golden standard lock-in detection misses all the harmonic frequencies in the MIP signal. Second, the well-depth of a typical CMOS camera seriously limits the probe power to 0.01 mW at sample. Thus, many averages are needed to reach a reasonable signal to noise ratio. We overcome these difficulties through two innovations. The first one is to digitize the probe photons received by a fast photodiode. Then, in the frequency domain, a match filter is used to extract all MIP signals at fundamental and harmonic frequencies. The second one is to perform patterned probe illumination and collect photons with a photodiode which has a saturation threshold of tens of mW. Then, a MIP image is recovered by matrix inversion. In this “single-pixel camera” approach, the probe power can be increased by 1000 times, which indicates that the speed can be improved 30 times to reach the same signal to noise ratio of wide field MIP at the shot noise limit. The goal of this R33 proposal is to develop a digital signal processing, single pixel camera MIP microscope and validate its potential for high-content cancer metabolic imaging. In particular, we aim to validate a metabolic switch from glucose-mediated lipogenesis to fatty acids uptake/oxidation in ovarian cancers that become resistant to cisplatin. By accomplishing the proposed studies, we will generate a high-speed hyperspectral mid-infrared photothermal chemical imaging platform that is able to map the live cell metabolism at sub-micron spatial resolution. Metabolic imaging of live drug-resistant cancer cells by this platform opens new opportunities of unveiling hidden signatures that can potentially lead to adaptive therapies that inhibit the development of drug resistance in cancers.
