Self-renewable pluripotent human embryonic stem cells (hESC) can readily differentiate into embryoid bodies (EBs) that contain virtually all cell types. A major technical hurdle in this field has been the lack of an effective yet non-invasive method for definitively discriminating among undifferentiated hESCs and their derivatives. This isolation is vital for subsequent cell-based therapies. Although several physical methods have been effective for purifying lineages that express cell-specific markers on their surfaces, those of many other highly specialized cells remain unknown. Conventional purification methods that rely on genetic manipulation or immunostaining are unsuitable for clinical use. Here we plan to use laser tweezers Raman spectroscopy of single cells to determine differences between individual human embryonic stem cells and their derivatives. This technique is based on the intrinsic biochemical signatures of cells obtained through non-destructive inelastic light scattering. It could become the basis for a novel, rapid and entirely non-destructive cell screening method that can accurately identify individual cells. Inelastic light scattering from specific molecular vibrations within a cell yields information on the biochemical composition of individual cells. Light from a laser is focused to a tight spot that covers the dimensions of a single cell. A small fraction of this laser light will interact with proteins, DNA, RNA, and other biomolecules in the cell, and lose parts of its energy to excite molecular vibrations. The outgoing (scattered) laser light is slightly red-shifted and contains distinct sharp peaks, that can be used to determine the relative distributions and concentrations of specific biochemicals within the cell. We found that by analyzing the strength of these peaks, different cells can be accurately distinguished. This technique is very fast, non-destructive, non-invasive, and does not require special sample preparation. It is important to note that the absence of external reagents (such as fluorescent labels) yields test cells that are neither destroyed/modified nor contaminated during the analysis. We expect our technique to be well suited for the future rapid analysis and isolation of hESCs, which is crucial for their use in clinical applications. It is also more accurate and specific than standard flow cytometry, the current gold standard for cellular diagnostics that is based on light scattering and fluorescence detection of a limited number of exogenous biomarkers. This grant will support important, necessary experiments that will define the spectroscopic signature of a wide range of hESC cells and their derivatives. These signatures will lay the foundation for commercially viable rapid cell sorting based on select peaks determined from these spectra.
Statement of Benefit to California:
Self-renewable pluripotent human embryonic stem cells (hESC) can readily differentiate into embryoid bodies (EBs) that contain virtually all cell types. A major technical hurdle in this field has been the lack of an effective yet non-invasive method for definitively discriminating among undifferentiated hESCs and their derivatives. This isolation is vital for subsequent cell-based therapies. Although physical methods such as magnetic bead sorting have been effective for purifying lineages that express cell-specific surface markers (e.g. CD34 for hematopoeitic cells), those of many other highly specialized cells remain unknown. For instance, all heart-restricted markers known are either cytoplasmic or nuclear. Conventional purification methods, which rely on genetic manipulation or immunostaining, are unsuitable for clinical use. With the research proposed here, we will demonstrate and establish an important new capability of rapid, non-invasive hESC identification. Rapid discrimination at the single cell level will permit the isolation of uncontaminated hESCs for use in cell-based therapies. This will remove a major clinical hurdle and rapidly benefit patients in need of hESC-based treatments. We anticipate to generate basic discoveries as well as patentable application-based technologies for licensing activities. Collectively, this will help establish the State of California as a pioneer in stem-cell based therapies and as a biotech hub, and provide its citizens with both academic and economic benefits. When the initial funding period ends and with the data generated from this proposal, we also expect to bring in federal monies via the NIH to California.
SYNOPSIS: The goal of the proposal is to use single cell Raman Spectroscopy to characterize biochemical signatures of hES cells and their derivatives. The goal will be classification into different cell types, for example identification of different cardiac cell types for cardiac repair. The Raman technique is nondestructive and could form the basis for rapid viable cell sorting based on spectral peaks rather than FACS.
INNOVATION AND SIGNIFICANCE: The project is reasonably innovative. The innovation here lies in use of different instrumentation to gain insights into biochemical states of individual cells and for cell type identification. There is only a small literature on Raman spectra of stem cells and none at all on stem cell of an alternative Raman technique, CARS (Coherant anti-Stokes Raman Scattering) microscopy useful for imaging in real time. It is known that IR and Raman spectra of bacteria can identify them at the strain level. Analogous findings are expected for stem cells – i.e. identification based on small differences.
Although it is well established that Raman spectroscopy can provide positive identifications of cells and changes in cells, it is not widely used by biologists, perhaps in part because the spectral signatures are not easily attributable to specific molecules within cells and perhaps because the spectroscopy is unfamiliar and the instrumentation is not widely available in life sciences laboratories. A successful demonstration of rapid and reliable hESC classification could catalyze its use in this field where techniques are still rapidly changing. New methods for cell purification would also be an important adjunct for the field.
STRENGTHS: The strength here lies in the new technology to be deployed that has the potential for real time monitoring and for identifying and classifying cell types. The capacity to interrogate cells on the single cell level is an asset. Collaboration with cardiologists who have isolated specific cell types from animals (eg guinea pigs) is an asset as well as it will ground the project to some biology.
The Raman spectroscopy/optical tweezers experiment is carefully explained. Using optical trapping to follow changes in stem cells is elegant. The PCA-based classification schemes is less clearly explained, but the methodology is familiar in the spectroscopy world and has been widely applied to related problems such as identification of strains of bacteria, discrimination between normal and cancerous cells. It would appear that simplest classification techniques are envisioned here. There is almost no discussion of alternatives to either the optical trapping method or the simple PCA classification scheme. That is acceptable because the techniques are proven and will almost certainly work.
WEAKNESSES: The experiments will employ in vitro differentiation of hES cells. This will generate a multitude of different cell types, many of which will be of fetal/embryonic character, rather than those of adult phenotypes. Making the assignments of the "types" of individual cells with specific spectra may be challenging. This may be the most challenging aspect of the project. One might argue that additional work in mESC(where differentiation may be manipulated more easily) might facilitate this work. Use of hESC, though suitable for this RFA, may be premature.
The CARS experiments are not as clearly thought out as the other experiments. Exactly what would be gained by CARS is never really explained. It is stated that CARS imaging on various bands would be tried, but why this would be better for following changes than a complete Raman spectrum and a simple video is not clearly explained. CARS has been successful largely in imaging lipid bilayers. That application works because lipid C-H stretches are intense and occur at very different frequencies from the aliphatic C-H stretches of, for example, proteins. Very little has been done with other bands in the spectrum of a cell, largely because they are weaker and interference between the vibrational band and the non-resonant background can make interpretation difficult. There is no acknowledgement of this problem, although it is familiar to everybody who worked with CARS or has even just read the literature. Attempts to circumvent the non-resonant background problem with time-resolved measurements, polarization control, modulation schemes and other variations on CARS is an active area of research and there is no consensus on the best method.
It would be better to limit this proposal to microspectroscopy and to develop alternatives to the simple cluster analysis proposed to classify cells. There has been much work in identification of bacteria by IR and Raman spectroscopy and the project could benefit by use of some of the more recent classification methods.
DISCUSSION: The comment was made that this is a high technology instrumentation application in search of cells to study. There is a great deal more classification potential here than people realize. Classification in bacteria is on the verge of going commercial thus there is possible downstream commercial potential for this type of work.