Supplementary Figure 1 Schematic diagram of Kőhler illumination. The green beam path represents the excitation path and the red represents the emission path.
Supplementary Figure 2 Microscope base components and assembly process. Numbers correspond to assembly flow as indicated in Steps 1 3.
Supplementary Figure 3 Camera mount components and assembly process. Numbers correspond to assembly flow as indicated in Steps 4 7.
Supplementary Figure 4 Spirit level alignment. Example of using (a) bullseye spirit level [Step 8] and (b) tubular spirit level [Step 11] to align components.
Supplementary Figure 5 Tube lens mount components and assembly process. Numbers correspond to assembly flow as indicated in Step 9A.
Supplementary Figure 6 Dichroic cube mount components and assembly process for epifluorescence imaging feature. Numbers correspond to assembly flow as indicated in Step 9B.
Supplementary Figure 7 Motorized microscope objective stage components and assembly process. Numbers correspond to assembly flow as indicated in Steps 10 16.
Supplementary Figure 8 Microscope objective mounted on translation stage adapter. The base of the objective lens (above the treaded section) should be resting on the tube lens mount. Once the objective lens is properly secured to the stage mount, shift the lens mount up and down to ensure that the lens threading does not rub against the mount.
Supplementary Figure 9 Microscope specimen stage components and assembly process. Numbers correspond to assembly flow as indicated in Steps 17 26.
Supplementary Figure 10 Fabrication steps for microscope stage. (a) Two pairs of 1 x 3 glass slides are bonded to each other by applying superglue on the areas shown by blue arrows. Apply enough superglue so that the entire area of the base is covered. (b) Bond two additional glass slides on top and bottom of the two slide pairs and (c) wait until the glue has cured. (d) Mount the completed glass stage on the XY translation adapter, and tightly position it by sliding and locking the slider. The glass stage can then be secured by tightening four screws on the top of the XY translation adapter.
Supplementary Figure 11 Transillumination source components and assembly process. Numbers correspond to assembly flow as indicated in Step 27A.
Supplementary Figure 12 Epifluorescence illumination module components and assembly process. Numbers correspond to assembly flow as indicated in Step 27B.
Supplementary Figure 13 Schematic diagram showing lens and diaphragm location for Kőhler illumination. Beam path showing the approximate location of lenses and iris diaphragms, where f is the focal length and f b is the back focal length of the lens. The actual positions may vary during the setup; hence calibration must be performed to achieve good Kőhler illumination. For example, the first iris diaphragm (condenser diaphragm) must be located where the LED source is clearly visible after the condenser lens.
Supplementary Figure 14 Calibration of microscope magnifying power. To achieve the desired magnifying power, (a) acquire an image of a calibration grid and measure the grid length (in pixels) with ImageJ. The effective magnification power is given by M LLM =detector pixel size [μm/px] /(grid size [μm] /measured number of pixels [px]). (b) If needed, adjust camera position. Repeat steps if necessary.
Supplementary Figure 15 Imaging area comparison with and without epifluorescence illumination add-on. (a) Micrographs of 100 μm spaced target grid captured with and epi-fluorescence illumination module. Significant vignetting can be observed. (b) In comparison, the imaging area is larger if the epi-fluorescence module is omitted. Scale bar, 100 μm.
Supplementary Figure 16 Comparison between analog and digital scintillation images. Analog scintillation images captured with (a) no binning, 1200 EM gain, and 1 min exposure time and (b) 4 binning, ~700 EM gain, and 2 min exposure time. (c) Digitial scintillation image captured with 4 binning, 1200 EM gain, 30 ms exposure time per frame, and 20,000 frames. Hot spots (red arrows) are caused by stray ionizing radiation directly hitting the EMCCD detector, frequently appears in analog images, whereas such signals are rejected in digitally reconstructed images. Scale bar, 200 μm.
Supplementary Figure 17 Multimodal micrographs of confluent MDA-MB-231 cells. Brightfield, fluorescence, bioluminescence, and overlay images from MDA-MB-231/Luc cells. Top row. Nikon 20x objective lens (effective magnification of 4x). Scale bar, 200 μm. Bottom row. Olympus 40x oil-immersion objective lens (effective magnification of ~11x). Scale bar, 100 μm.
Supplementary Figure 18 Example of bioluminescence imaging captured under short exposure times for real-time imaging. (a) Brightfield image of MDA-MB-231/Luc cells and its corresponding bioluminescence images captured with different imaging parameters : (b) no binning, ~1060 EM gain, and 20 s exposure time; (c) no binning, 1200 EM gain, and 3 s exposure time; (d) 2 binning, 1200 EM gain, and 0.75 s exposure time. Scale bar, 200 μm.
Supplementary Table 1, Theoretical numerical aperture (NA) and brightness (B) for different microscope objectives. Make Model M LLM Eq.(2) NA LLM (NA orig ) 1 B LLM,trans B LLM,epi Nikon Olympus Zeiss CFI Plan Apo Lambda 2 0.5 Yes 0.1 400.0 4 CFI LU Plan Fluor EPI P 5 1.25 Yes 0.15 144.0 3.24 CFI S Fluor 10 2.5 No 0.45 (0.5) 330.6 68.3 CFI Plan Apo Lambda 20 5 No 0.71 (0.75) 204.1 104.1 CFI S Plan Fluor ELWD ADM 40 10 Yes 0.6 36 13.0 CFI Plan Fluor (oil) 40 10 Yes 1.3 169 285.6 CFI Plan Apo Lambda (oil) 60 15 Yes 1.4 87.1 170.7 CFI Plan Apo Lambda 100 25 Yes 1.45 33.6 70.7 MPLN 5 1.4 Yes 0.1 51.8 0.5 UPLSAPO 10 2.8 Yes 0.4 207.4 33.2 UPLSAPO 20 5.6 Yes 0.75 182.3 102.5 UPLSAPO 40 11.1 Yes 0.9 65.6 53.1 UPLFLN 40 11.1 Yes 1.3 136.9 231.3 UPLSAPO 60 16.7 Yes 1.35 65.6 119.6 UPLSAPO 100 27.8 Yes 1.4 25.4 49.8 C Epiplan-Apochromat 5 1.5 Yes 0.2 172.2 7.0 Epiplan-Apochromat 10 3.0 Yes 0.4 172.2 27.9 Plan-Apochromat 20 6.1 No 0.78 (0.8) 164.7 99.6 Plan-Apochromat 40 12.1 No 0.93 (0.95) 58.5 50.3 Plan-Apochromat 40 12.1 Yes 1.4 133.4 261.5 Plan-Apochromat 63 19.1 Yes 1.4 53.8 105.4 Alpha Plan-Apochromat 100 30.3 Yes 1.46 23.2 49.5 Table comparing theoretical performance of different microscope objective manufacturers. The theoretical transillumination brightness of standard microscopes ranges between 4-16 and 16-25 for 5 and 10 magnification, respectively. References 1. Kim, T. J., Tuerkcan, S., Ceballos, A. & Pratx, G. Modular platform for low-light microscopy. Biomed. Opt. Express 6, 4585-4598 (2015).
Supplementary Note 1, Basics on radioactivity. In a radioluminescence microscopy (RLM) experiment, the rate of radioactive decay must be considered because the amount of radioactivity cannot be regarded as constant over the duration of the experiment. The radioactivity A by definition is, A 1/2 dn dt ln 2, t N, (1a) where N is the time-decaying number of radioactive atoms, λ is the decay rate constant, and t 1/2 is the radioactive half-life. The user can refer to a table of nuclides for information about individual radionuclides (e.g. http://www.nndc.bnl.gov/chart/). Solving differential equation (1) yields, (1b) T N N0e, (2) where N 0 is the number of radioactive atoms at the beginning of the radioluminescence exposure and T is the total duration of the exposure. To quantify the molecular uptake of individual cells using the radioluminescence microscopy, two parameters should be considered, sensitivity and radioactive yield. The sensitivity accounts for the probability that a radioactive emission be detected by the microscope. It is a function of the isotope, scintillator, and light collection efficiency by the microscope. For quantitative estimation of radionuclide concentration, calibration must be performed, as explained in Steps 71-79. Also, the radioactive yield must be taken into consideration since the radioluminescence microscope only detects particulate radiation (α, β - or β + ). This is due to the fact that the main detectable signals will originate from particles that undergo collision cascades in the scintillator rather than from photons (γ-rays). Other modes of radioactive decay such as electron capture and internal transition cannot be detected. Accounting for the aforementioned parameters, the number of measured events D is related to the number of actual radioactive decays N B, D NB N0 N, (3) SY where S is the sensitivity calibration factor and Y is the radioactive yield for particulate radiation. Solving for equation (3), the number of molecules initially present per pixel is, N 0 D S Y 1 1 e T. (4) It should be noted that equation (4) can also be expressed in terms of activity ( A N).