Multiphoton flow cytometry to assess intrinsic and extrinsic fluorescence in cellular aggregates: applications to stem cells

Abstract

Detection and tracking of stem cell state are difficult due to insufficient means for rapidly screening cell state in a noninvasive manner. This challenge is compounded when stem cells are cultured in aggregates or three-dimensional (3D) constructs because living cells in this form are difficult to analyze without disrupting cellular contacts. Multiphoton laser scanning microscopy is uniquely suited to analyze 3D structures due to the broad tunability of excitation sources, deep sectioning capacity, and minimal phototoxicity but is throughput limited. A novel multiphoton fluorescence excitation flow cytometry (MPFC) instrument could be used to accurately probe cells in the interior of multicell aggregates or tissue constructs in an enhanced-throughput manner and measure corresponding fluorescent properties. By exciting endogenous fluorophores as intrinsic biomarkers or exciting extrinsic reporter molecules, the properties of cells in aggregates can be understood while the viable cellular aggregates are maintained. Here we introduce a first generation MPFC system and show appropriate speed and accuracy of image capture and measured fluorescence intensity, including intrinsic fluorescence intensity. Thus, this novel instrument enables rapid characterization of stem cells and corresponding aggregates in a noninvasive manner and could dramatically transform how stem cells are studied in the laboratory and utilized in the clinic.

Figure 1

Schematic diagram of the MPFC system. A microfabricated flow cell is placed on a multiphoton laser scanning microscope equipped with a Ti:Sapphire laser for excitation, photomultipliers, and scanning optics and electronics for signal collection. Sample fluid (i.e., biological specimens or beads) and sheath fluid are introduced into the microchannels using syringe pumps and associated tubing. See the Materials and Methods and the Results sections for additional details.

Figure 2

A microfabricated flow cell to accommodate large particles. Sheath and sample volumes were introduced into a PDMS microfabricated flow cell and flow rates were controlled via syringe pumps. A modular stage insert was fabricated to protect the flow cell from damage and to allow for turn-key application of the flow cell to other microscopy systems. A: Top-down view of the microfabricated flow cell. Features of the flow cell include waste output/reservoir drop, sample input and sheath input, and an interrogation point where optical analysis occurs. Scale bar, 5 mm. B: Top-down photograph of the flow cell housed in the modular stage insert. Features of the stage insert include a series of tubing adaptors that allow the flow cell to be easily added to and removed from the MPLSM system. C: Hydrodynamic control of the sample path. Sample fluid was labeled with fluorescein and sheath and sample flow rates were controlled via syringe pumping. Images captured using the WiscScan software illustrate how varying the sheath to sample flow velocity ratios (α) alter the width of the sample stream. D: Comparison of the predicted focused width with experimental data at different ratios of sheath volumetric flow rate to sample volumetric flow rate (100–1,000 and 50–500 µL/min, respectively). Model predictions were based on those previously published (Lee et al., 2006).

Figure 3

Size and intensity of fluorescent, polystyrene beads using the MPFC system. A: Accuracy and precision, using bright-field optics, of bead size using MPFC compared to bead size discerned using the MPLSM system without flow (i.e., static). Static measurements of bead diameter were made prior to introducing the beads into the MPFC system. Polystyrene beads were introduced into the MPFC at a maximum volumetric flow rate of 300 µL/min. Bead size measured under static conditions did not vary from bead size when measured on the MPFC system (P= 0.69, P= 0.91, P= 0.08). Standard deviation for each bead size measured on the MPFC was less than 8% of the total. B: Bright-field image of a 400 µm bead under static conditions. C: Bright-field image of a 400 µm bead using MPFC. Lack of warping and elongation indicate data acquisition rates were similar to the fluidic sample speed per unit time and that the microspheres experienced minimal mechanical disturbance. D: Relationship between bead diameter and normalized fluorescence intensity. Static measurements of mean bead intensity were made prior to measuring mean bead intensity using the MPFC system. Mean fluorescence intensity was plotted as a function of bead diameter. Best fit regression analysis was applied to the datasets and both static and MPFC conditions yielded second-order exponential relationships; R² values of 0.99 and 0.99, respectively. E: Fluorescence intensity image of a 400 µm bead (same bead as B) under static conditions. Static images were captured at the location corresponding to the maximum total intensity, thus the “center” of the bead. F: Fluorescence intensity image of a 400 µm bead using MPFC (same bead as C). Mean fluorescence intensity did not vary between static and MPFC acquisition modes for each bead size (P= 0.58, P= 0.72, P= 0.74). a.i.u, arbitrary intensity units. Scale bar, 200 µm.

Figure 4

Size of EBs using the MPFC system. A: Accuracy and precision of EB size using MPFC compared to EB size discerned using the MPLSM system without flow (i.e., static). Measurements of EB size, using bright-field optics, were made prior, during, and following introduction of the EBs into the MPFC system. Standard deviation for EB size was less than 13% of the total size. EB size discerned under static conditions after MPFC did not vary from EB size discerned using the MPFC system [P= 0.05 (pre versus MPFC), P= 0.29 (pre versus post), P= 0.13 (flow versus post)]. EBs were introduced into the MPFC at a volumetric flow rate of 200 µL/ min, which remained consistent between experiments. B: Bright-field images of pre-MPFC static, MPFC, and post-MPFC static EBs. Multicellular aggregates maintained original morphology and defined peripheral border during and after analysis with MPFC. Scale bar, 200 µm.

Figure 5

(Color online) Extrinsic fluorescence intensity of EBs using the MPFC system. The Ti-Sapphire laser was tuned to 890 nm to excite GFP and a 520/35 band-pass emission filter was used to exclude autofluorescence. A nontransfected mESC cell line (D3) and an α-MHC::GFP transfected mESC cell line were used to generate EBs. Histogram analysis of GFP expression of EBs (A) pre-MPFC, (B) MPFC, and (C) post-MPFC. Two separate experiments were conducted and approximately 50 EBs were analyzed per condition (i.e., before, during, and after MPFC). The maximum background intensity was defined such that 95% of the nontransfected EBs expressed mean fluorescence intensity levels below this intensity level (line in histogram). The percentage shown, indicating the fraction of EBs derived from transfected cells expressing GFP, was determined based on the background level. Representative images of D3 and α-MHC::GFP mESC-derived EBs, located directly adjacent to the corresponding plot, are provided for each condition. Color bar represents quantified intensity levels from 0 (black) to 255 (green). a.i.u., arbitrary intensity units. Scale bar, 200 µm.

Figure 6

Intrinsic fluorescence intensity of EBs using the MPFC system. The Ti:Sapphire laser was tuned to 780 nm to excite the intrinsic fluorescent metabolite, NADH. A: Mean fluorescence intensity of EBs before, during, and after analysis with the MPFC system. Mean fluorescence intensity of EBs did not vary between acquisition conditions [P = 0.55 (pre versus MPFC), P= 0.46 (pre versus post), P= 0.89 (flow versus post)]. B: NADH fluorescence intensity distributions before, during, and after analysis. Standard deviations within each condition described in panel A were large, and so the distribution of intensity between samples was compared. Though the distribution was large, the distribution profile did not vary substantially between conditions. C: Representative images depict NADH intensity acquired pre-MPFC, MPFC, and post-MPFC. Localized differences in intrinsic fluorescence were detected in all three conditions. The color bar represents quantified intensity levels from 0 (black) to 255 (white). a.i.u., arbitrary intensity units. Scale bar, 200 µm.

Figure 7

Particle recovery following MPFC acquisition. Polystyrene beads or EBs were introduced into the MPFC system and the relative fraction recovered post-MPFC was determined. Input (concentration introduced into the system, black bars) was compared to output (concentration acquired in the waste reservoir after MPFC analysis, gray bars) for beads and EBs. Particle recovery exceeding 100% likely reflects an increase in concentration of particles in the syringe prior to introduction into the system and thus the subsequent measurement of a concentrated sample. Minimum particle recovery was 81%.