Enhanced Multiplexing of Immunofluorescence Microscopy Using a Long Stokes Shift Fluorophore

Abstract

Immunofluorescence labeling and microscopy offers a highly specific means to visualize proteins or other molecular species in a sample by labeling target antigens with fluorescent probes. These fluorescent probes can then be visualized using a fluorescence microscope, allowing their relative spatial relationships to be determined. Due to spectral overlap of common fluorophores, however, it can be challenging to analyze more than three antigens in a single sample with standard imaging approaches. This protocol describes multiplexed labeling and imaging of four target antigens through the use of a long Stokes shift fluorophore—a fluorophore with an unusually large gap between its excitation and emission maxima—in tandem with three conventional fluorophores. This combination allows for multiplexed imaging of four antigens in a single sample with excellent spectral discrimination suitable for sensitive analyses using standard imaging hardware. Particular advantages of this approach are its flexibility in terms of target antigens, and the lack of any specialized procedures, reagents, or equipment beyond the commercially-available long Stokes shift fluorophore-coupled labeling reagent.

INTRODUCTION:

Immunofluorescence (IF) is an invaluable tool to localize target antigens in biological tissues. This technique permits visualization of biomolecules via the direct or indirect binding of a fluorescent probe, followed by imaging with a fluorescence microscope. Since its first demonstration almost 80 years ago(Coons et al., 1942), the development of a deep catalogue of fluorescently-labeled probes, as well as the ubiquitous deployment of fluorescence microscopes, have made IF an extremely powerful and highly-accessible technique.

One of the most common and useful applications of IF involves determining the spatial relationship of antigens in situ, a technique that requires the multiplexing of fluorescent probes. A key limiting factor in the power of these characterizations, however, is the number of fluorescent labels that can be both imaged efficiently and well-separated from one another (spectrally distinguished).

The total number of such multiplexed probes in a given sample is often practically limited to three, roughly occupying the blue-green, red, and infrared bands of the electromagnetic spectrum. Additional multiplexing is generally limited to architectural contrast (e.g., cytoskeletal, DNA), as opposed to targeting a specific biomolecule of experimental importance, but moderate-to-low abundance or antigenicity. That is, additional multiplexing generally relies on biomolecules for which uniquely efficient probes exist (e.g., nuclear contrast) or of uncommonly high abundance (e.g., cytoskeletal antigens), or both, that overcome inherent imaging inefficiencies related to light sources, probes, or tissue autofluorescence.

A number of approaches exist to address this limitation. From the standpoint of imaging and analysis, spectral imaging with linear unmixing and related approaches make use of specialized but now widely available hardware and algorithmic image processing to determine the underlying components of overlapping emission patterns from biosamples(Zimmermann, 2005). These approaches, however, generally rely on assumptions regarding fluorophore reference emission spectra that may be sample- or instrument-specific, or may present multiple valid solutions to convolved spectra(McRae et al., 2019;Neher et al., 2009;Tsurui et al., 2000). From the standpoint of labels, probes such as quantum dots with narrower emission spectra vs. standard organic fluorophores (among other advantages) enhance spectral discrimination, but present challenges related to their physical and biological properties (large size and potential toxicity, among other application-specific issues) that have limited their application(Jaiswal and Simon, 2004;Resch-Genger et al., 2008;Wegner and Hildebrandt, 2015). Existing approaches, therefore, leave room for an accessible, low-friction method to enhance multiplexing beyond the standard three probes.

This protocol demonstrates such a method: application of a long Stokes shift organic fluorophore. The Stokes shift in this context is the difference between the maxima of the absorption and emission spectra for a specific fluorophore (Fig. 1). This difference is substantially greater for long Stokes shift vs. conventional fluorophores. As a result, standard emission filters can adequately distinguish signal arising from a conventional fluorophore from that arising from a fluorophore with completely overlapping excitation properties, but a longer Stokes shifted emission maximum (Fig. 1), without resorting to unmixing algorithms. In tandem with two other conventional probes, such an approach permits the labeling and analysis of four target antigens in situ (Basic Protocol 1; Fig. 2), all via standard IF histological and imaging methods. The below protocol furthermore describes approaches to rigorously avoid both spectral bleed-through and reagent cross-talk, as well as correction of chromatic aberration. This allows the spatial relationships of four biomolecule species to be precisely determined in a single sample. In this protocol, two fluorophores excited by a 488 nm laser line are utilized, one of which exhibits a long Stokes shift of 165 nm (ATTO 490LS(ATTO-TEC GmbH - ATTO 490LS); Fig. 1). Following IF labeling, we describe general aspects of image acquisition including laser and filter selection on a confocal microscope (Basic Protocol 2).

Figure 1.

Excitation and emission spectra of Alexa Fluor 488 (light green) and ATTO 490LS (dark green). The 488 nm laser line is indicated (blue line). Note relative lack of overlap of emission spectra despite overlapping excitation spectra, due to long Stokes shift of ATTO 490LS (Fluorescence SpectraViewer - US; ATTO-TEC GmbH - ATTO 490LS).

Figure 2:

Image of mouse cortex co-labeled against CD68 (blue), synapsin 1/2 (green), Iba1 (magenta), and GFAP (white). Individual fluorophore label images are shown to left. Synapsin 1/2 is labeled with the long Stokes shift fluorophore.

STRATEGIC PLANNING

Appropriate controls (Table 1) are paramount for ensuring that both primary and secondary reagents (antibodies or streptavidin) bind solely to the expected antigen, to ensure the absence of spectral cross-talk or problematic background, and to correct for chromatic aberrations so that accurate spatial relationships can be precisely determined. The relative importance of these confounds will depend upon the specific application. The below protocol describes strategies designed to rigorously avoid—or at minimum, reveal the contribution of—such confounds. Simpler alternatives, appropriate for some applications, are also briefly introduced below.

Table 1:

Controls.

Control		Purpose	Ckn-anti-GFAP	Gp-anti-SYN	Rb-anti-Iba1	Rt-anti-CD68	Gt-anti-Ckn 647	Gt-atnti-Gp Biotinylated	Streptavidin 490LS	Gt-anti-Rb 594	Gt-anti-Rt 488
No-primaries	No-GFAP	Ensure non-specific binding and spectral cross talk are measured and/or eliminated	−	+	+	+	+	+	+	+	+
	No-synapsin		+	−	+	+	+	+	+	+	+
	No-Iba1		+	+	−	+	+	+	+	+	+
	No-CD68		+	+	+	−	+	+	+	+	+
No-label		Measure autofluorescence and background	−	−	−	−	−	−	−	−	−
Channel alignment		Measure chromatic aberration	−	+	−	−	*	+	+	*	*

^*Replace fluorophores targeting other host species with fluorophores targeting the host species of the primary antibody used in the channel alignment control. For above example, one would use Gt-α-Gp 647, Gt-α-Gp 594, and Gt-α-Gp 488 in lieu of the other fluorophores.

Abbreviations: GFAP, Glial Fibrillary Acidic Protein; Ckn, chicken; SYN, synapsin; Gp, guinea pig; Rb, rabbit; Rt, rat; Gt, goat.

This protocol employs no-primary, no-label, and channel alignment controls. No-primary controls, a panel of controls that each lack only one of the four primary antibodies, ensure that non-specific binding and spectral cross-talk are avoided, or at least measured. Single color controls, in which a single primary-secondary-fluorophore combination is used in isolation, are simpler to perform and sufficient for many applications. We have found, however, that commercial secondary antibodies frequently cross-bind at low levels with non-targeted primary antibodies. Though spectral cross talk can be assessed with single color controls, such cross-binding cannot be measured.

No-label controls consist of tissue not treated with any primary or secondary labeling reagents. These controls give a means to quantify autofluorescence and other sources of background. The channel alignment control, a tissue sample in which a single primary antibody is labeled with a cocktail of secondary reagents containing all four fluorophores, offers a means to quantify and correct chromatic aberrations specific to both the imaging system and the sample. A more straightforward alternative entails the imaging of fluorescent microspheres that emit in all needed wavelength ranges (e.g., ThermoFisher cat. No. T7279). This approach cannot assess aberrations inherent to the sample itself, and the wavelengths of microsphere excitation/emission will not precisely match those of the experimental fluorophores, but they should nonetheless be adequate for many applications.

In this example, we targeted the following four antigens: glial fibrillary acidic protein (GFAP; a marker of astroglia), synapsin 1/2 (a family of presynaptic proteins), Iba1 (a marker of microglia), and CD68 (a marker of microglial lysosomes) in 50 μm-thick, paraformaldehyde-fixed mouse brain sections. As the only unique reagent is the long Stokes shift-coupled labeling reagent, the protocol should be easily adapted to other biomolecular target combinations, fixation methods, or non-nervous tissues, though alternate or additional application-specific processing steps may be required.

Table 1 shows the primary and secondary antibodies that were utilized for each control sample in this example. These must be tailored to a given experiment’s target antigens.

BASIC PROTOCOL 1

FOUR-PROBE IMMUNOFLUORESCENCE LABELING

Introductory paragraph:

This protocol describes immunolabeling brain tissue with primary antibodies against four target antigens followed by labeling of each primary with a spectrally separable fluorophore, including the long Stokes shift fluorophore (Fig. 1). While immunolabeling in brain tissue is described below, this protocol could be applied to other tissue samples with appropriate antibody and section preparation modifications. The below protocol utilizes the following primary antibodies: Chicken-anti-GFAP, Guinea pig-anti-Synapsin, Rabbit-anti-Iba1, and Rat-anti-CD68. This general approach, however, should be applicable to any combination of primary antibodies that can be reliably distinguished by secondary reagents. Additionally, the below protocol describes labeling free-floating, microtome-cut 50 um-thick paraformaldehyde-fixed sections, but is applicable to cryostat-cut, slide-mounted tissue with minor adjustments.

Note we describe sequential labeling with two of the four antibody combinations (targeting GFAP and synapsin). This was performed to eliminate traces of cross-binding we observed between antibody species for SEQUIN analysis, a sensitive quantitative application (Sauerbeck et al., 2020; Reitz et al., 2020). Depending on experimental goals and antibody characteristics, it may be reasonable to perform all primary followed by all secondary antibody incubations simultaneously.

Materials:

50 µm-thick sections of paraformaldehyde-fixed brain tissue

Phosphate buffered saline (PBS)

6, 12, and 24 well plates

Net wells

Blocking Buffer

Antibody Buffer

Four selected primary antibodies (antibodies listed are those used below)

• Chicken-anti-GFAP (Aves Labs, cat. no. F-1005, RRID: AB_2313547)

• Guinea pig-anti-Synapsin (Synaptic Systems, cat. no. 106004, RRID: AB_1106784)

• Rabbit-anti-Iba1 (Wako, cat. no. 019–19741, RRID: AB_839504)

• Rat-anti-CD68 (Invitrogen, cat. no. MA5–16674, RRID: AB_2538168)

Three secondary antibodies conjugated with Alexa Fluor 488, 594, and 647 fluorophores (see note below under Critical Parameters, Antibody Selection)

• Goat-anti-Chicken 647 (Invitrogen, cat. no. A-21449, RRID: AB_2535866)

• Goat-anti-rabbit 594 (Invitrogen, cat. no. A11037, RRID: AB_2534095)

• Goat-anti-rat 488 (Invitrogen, cat. no. A-11006, RRID: AB_2534074)

Three secondary antibodies conjugated with Alexa Fluor 488, 594, and 647 fluorophores targeting same species as biotinylated antibody (used for the channel alignment control; see Strategic Planning)

• Goat-anti-Guinea pig 488 (Invitrogen, cat. no. A11073, RRID: AB_2534117)

• Goat-anti-Guinea pig 594 (Invitrogen, cat. no. A11076, RRID: AB_2534120)

• Goat-anti-Guinea pig 647 (Invitrogen, cat. no. A-21450, RRID: AB_2735091)

Biotinylated secondary antibody targeting one of the primary antibodies (in this case Goat-anti-Guinea pig; Invitrogen, cat. no. A18779, RRID: AB_2535556)

Streptavidin conjugated ATTO 490LS (ATTO-TEC, cat. no. AD 490LS)

Tris-MWL 4–88 (Electron Microscopy Sciences, cat. no. 17977–150)

AF 300 (Electron Microscopy Sciences, cat. no. 17977–25)

Shaker

Centrifuge capable of 14,000 g

1. Move tissue sections from storage plate containing cryoprotectant to a 6 well plate containing 10–15 mL 1x PBS/well using a clean paintbrush.

   Ensure sections are fully submerged and free to move independently. Netwells can be used in the plate to allow for easy transfer of sections between 6 well plates following the first and second rinse (step 3).

2. Rinse sections on shaker at slow speed for 5 min at room temperature.

   Thorough rinsing of tissue removes any remaining cryoprotectant, preparing tissue for blocking and antibody staining.

3. Move sections to new 6 well plate with 1x PBS and rinse for 5 min.

4. Repeat steps 2 and 3 once more for a total of 3 rinses.

5. Prepare blocking buffer (see Support Protocol 1).

   Blocking buffer can be prepared while rinsing tissue.

6. Remove one section from wash and place in 1x PBS at 4° C.

   This section serves as no-label control (see Strategic Planning).

7. Add blocking buffer to a 12 well plate (750 µL/well) and move sections using a clean paintbrush.

   Blocking buffer should be prepared with the host species of the secondary antibody (in this case goat) to minimize non-specific binding of the secondary antibody.

8. Block for 1 hr on shaker at slow speed at room temperature.

9. Prepare antibody buffer (see Support Protocol 2).

   Antibody buffer contains Triton 100-X, a detergent that aids in antibody penetration.

10. Add Chicken-anti-GFAP antibody to buffer at a concentration of 1:1000 and mix well.

11. Add antibody from step 10 to new plates, 750 µL per well in a 12 well plate for up to 8 sections per well and 500 µL per well in a 24 well plate for up to four sections per well, and move tissue using a clean paintbrush.

    The channel alignment control and no-GFAP control should not receive Chicken-anti-GFAP, but should instead be plated in antibody buffer free of antibody.

    Alternate: omit channel alignment and no-primary control steps in Basic Protocol 1. Prepare a slide for channel alignment by plating appropriate multiply-fluorescent microspheres (e.g., ThermoFisher cat. No. T7279) on a slide in the experimental mountant. Prepare a series of single-label controls by incubating tissue sections with single primary-secondary antibody-fluorophore combinations (one for each fluorophore used for experimental labeling).

12. Fill remaining wells with water and cover with lid.

13. Incubate 16–24 hours on shaker at room temperature.

14. Rinse 3x for 5 min each in 1x PBS (see steps 2–4).

15. Prepare antibody buffer (see Support Protocol 2).

    Antibody buffer can be stored for up to 24 hours at 4° C for use on subsequent day. Antibody buffer should be at room temperature before use.

16. Add Goat-anti-Chicken 647 secondary antibody to antibody buffer at a concentration of 1:200 and mix well.

    While modern fluorophores are generally photostable in ambient light, we protect both antibodies in solution and on tissue from light from this step forward.

17. Add antibody solution from step 16 as necessary in 12 (750 µL/well) or 24 well plates (500 µL/well).

18. Transfer sections from wash to appropriate well containing secondary antibody using a clean paintbrush.

    All sections but the channel alignment control should receive Goat-anti-Chicken 647. The channel alignment control should be placed in antibody buffer free of antibody.

19. Incubate 4 hours on shaker at room temperature.

20. Transfer tissue to 6 well plate containing 1x PBS using a clean paintbrush and rinse 3x for 5 min (see steps 2–4).

21. Prepare second primary by adding Guinea pig-anti-Synapsin at a concentration of 1:500 to antibody buffer prepared in step 15 and mixing well.

22. Plate antibody from step 21 as necessary in 12 (750 µL/well) and 24 well plates (500 µL/well) and transfer sections from wash to antibody with a clean paintbrush.

    No-synapsin control should be placed in antibody buffer free of antibody.

23. Fill remaining wells with water and cover with lid.

24. Incubate 16–24 hours on shaker at room temperature.

25. Transfer tissue to 6 well plate containing 1x PBS using a clean paintbrush and rinse 3x for 5 min (see steps 2–4).

26. Prepare fresh antibody buffer (see Support Protocol 2).

27. Prepare antibody solution by adding biotinylated Goat-anti-Guinea pig to antibody buffer at a concentration of 1:200 and mix well.

28. Prepare antibody for the channel alignment control by removing 500 µL from the antibody solution prepared with biotinylated Goat-anti-Guinea pig and adding Goat-anti-Guinea pig 594, Goat-anti-Guinea pig 647, and Goat-anti-Guinea pig 488, all at a concentration of 1:200. Mix well.

    All fluorophores are added to the channel alignment section at the same time to allow simultaneous labeling.

29. Add antibody solution from steps 27 and 28 as necessary in 12 (750 µL) or 24 well plates (500 µL) and transfer sections from wash to antibody with a clean paintbrush.

    The channel alignment control receives a different solution, prepared in step 28, than other experimental sections and controls.

30. Cover plates with lids and incubate on shaker at a slow speed in the dark for 4 hours at room temperature.

31. Transfer tissue to 6 well plate containing 1x PBS using a clean paintbrush and rinse 3x for 5 min (see steps 2–4).

32. Prepare labeling solution by adding Streptavidin 490 LS to antibody buffer prepared in step 26 at a concentration of 1:200 and mix well.

33. Add labeling solution from step 32 as necessary in 12 (750 µL) or 24 well plates (500 µL) and transfer sections from wash with a clean paintbrush.

    All experimental and control sections receive the same treatment in this step.

34. Cover plates with lids and incubate on shaker at a slow speed in the dark for 4 hours at room temperature.

35. Transfer tissue to 6 well plate containing 1x PBS using a clean paintbrush and rinse 3x for 5 min (see steps 2–4).

36. Prepare master-mix antibody solution for Rabbit-anti-Iba1 (working concentration of 1:500) by adding antibody to antibody buffer prepared in step 26 at double the working concentration of antibody (1:250). Mix well.

    Master-mix antibody solutions are used when two or more primary antibodies are applied simultaneously to help minimize pipetting errors.

37. Prepare master-mix antibody solution for Rat-anti-CD68 (working concentration of 1:400) by adding antibody to antibody buffer prepared in step 26 at double the working concentration of antibody (1:200). Mix well.

38. Combine master-mixes with each other (for experimental sections) and with antibody buffer (for no-primary controls) such that all solutions are diluted to appropriate working concentrations and mix well.

39. Place the channel alignment control in fresh 1x PBS, protect from light, and refrigerate at 4° C.

    The channel alignment labeling is complete and will remain in refrigerator until all sections are ready to be mounted.

40. Add antibody prepared in step 38 as necessary in 12 (750 µL) or 24 well plates (500 µL) and transfer sections from wash to antibody with a clean paintbrush.

41. Fill remaining wells with water and cover with lid.

42. Incubate 16–24 hours on shaker at room temperature, protected from light.

43. Transfer tissue to 6 well plate containing 1x PBS using a clean paintbrush and rinse 3x for 5 min (see steps 2–4).

44. Prepare fresh antibody buffer (see Support Protocol 2).

45. Add Goat-anti-Rabbit 594 and Goat-anti-Rat 488 secondary antibodies to antibody buffer at a concentration of 1:200 and mix well.

    As all tissue receives the same combination of fluorophores; master-mix creation is not necessary when combining multiple secondary antibodies.

46. Plate antibody from step 45 as necessary in 12 (750 µL) or 24 well plates (500 µL).

47. Transfer sections from wash to appropriate well containing secondary antibody using a clean paintbrush.

48. Cover plates with lids and incubate on shaker at a slow speed in the dark for 4 hours at room temperature.

49. Prepare Mowiol mounting media by adding AF 300 to MWL 4–88 at a ratio of 1:9 and vortexing well. Spin in a benchtop centrifuge until bubbles are no longer present.

    Mounting media should be prepared fresh day of use. Following preparation, store in dark at room temperature until ready to be used.

    Calculate amount to prepare based on 150–200 µL/slide, preparing a minimum volume of 250 µL.

    MWL 4–88 is a hardening mounting media whose refractive index increases as it cures, which reduces spherical aberration, a fluorescent aberration caused by light scatter. AF 300 is an anti-fade agent which reduces fluorophore fade over time.

    Depending on the sensitivity of the planned experiment to spherical aberrations, other mountants may be used.

50. Prepare clean coverslips.

51. Retrieve channel alignment and no antibody control tissue.

52. Transfer both tissue finishing secondary incubation and controls from step 51 to 6 well plate containing 1x PBS using a clean paintbrush and rinse 3x for 5 min (see steps 2–4).

53. Mount free-floating tissue in 1x PBS on a charged slide and allow to dry at room temperature in a flat, dark environment (~5–10 minutes).

54. Dunk slide in distilled water 5 times very quickly to remove salt residue and allow to dry again at room temperature in a flat, dark environment (~2–10 minutes).

55. Apply 150–200 µL mounting media/slide, dotting the mountant on tissue sections. Care should be taken to not touch sections with pipette tip.

56. Slowly apply coverslip and allow excess mountant to be expelled from under the coverslip.

57. Store slides in dark at room temperature.

    We recommend allowing Mowiol to cure for at least 3 days before imaging, as refractive index of mounting media takes several days to increase and plateau.

BASIC PROTOCOL 2

FOUR-PROBE IMMUNOFLUORESCENCE IMAGING

Introductory paragraph:

This protocol describes imaging tissue labeled with four immunofluorescent probes, including the long Stokes shift fluorophore. While the presented data were acquired using a Zeiss LSM 880 confocal microscope, any fluorescence microscope, confocal or otherwise, with appropriate light source(s) for exciting the necessary fluorophores, and with the filters necessary to adequately separate their emissions, can be used.

Materials:

Labeled, mounted tissue from Basic Protocol 1

Fluorescence microscope

Immersion oil (optional, for use with oil objective)

Lens paper

Lens cleaner

Lab wipes

1. Turn on microscope and all lasers/light sources that are needed.

   The minimum laser lines needed for the current example include 488, 561, and 633. Typical examples include the argon-ion laser, the helium-neon laser, the Ti:Sapphire laser, or various epifluorescent light sources. The precise laser line may differ from those above depending on the configuration of the microscope system.

   Allow time for light sources to reach optimum performance before imaging.

2. Clean objective with lens cleaner and lens paper, as needed.

3. Clean slide well with lens cleaner and lab wipes, as needed.

4. Configure the light sources, light path, filter sets, and detectors to optimally image each fluorophore.

   The steps needed to accomplish this step are unique to each microscope system, but certain principles apply. See Critical Parameters: Microscope Configuration (Light sources and Light path) and figure 3 for additional guidance.

5. Tune the acquisition parameters and laser power to optimize signal detection while avoiding photobleaching/toxicity.

   Specifics will depend on the given experiment, system, and goal. This is best accomplished using either a non-critical sample or region. See Critical Parameters: Microscope Configuration (Acquisition parameters) for additional guidance.

6. Set any remaining parameters, including stack thickness and Z-step size, if necessary, and acquire the image or image stack.

7. Save data and perform any post-acquisition processing steps specific to the system.

8. Quantify and correct chromatic aberration by measuring the shift in puncta position (or similar structure for other biomolecular targets) in X, Y, and Z between fluorophores targeted to the same structure in the channel alignment control sample. Alternatively, measure the shift in position for individual, multispectral microspheres imaged along with experimental tissue. Apply an appropriate correction using image processing software such as that supplied with the microscope, ImageJ/FIJI, or similar.

Figure 3:

Laser lines and choice of emission filters for imaging each fluorophore without spectral cross-talk. The depicted filter band (grey box) corresponds to the wavelength range passed to the detector. Note that this illustrates one successful configuration; other configurations are also possible.

REAGENTS AND SOLUTIONS:

Antibody buffer (1 mL)

• 900 μL 1x PBS

• 100 μL Normal Goat Serum (NGS; Vector Labs, cat. no. S-1000, RRID: AB_2336615)

• 3 μL Triton 100-X

• Note: After thawing, centrifuge NGS at 14,000–17,000 g for 5 minutes and when pipetting, avoid pellet

• Note: Antibody buffer can be prepared by adding 1x PBS to blocking buffer in a 1:1 ratio and then adding Triton 100-X

Blocking buffer

• 800 μL 1x PBS

• 200 μL Normal Goat Serum

• Note: After thawing, centrifuge NGS at 14,000–17,000 g for 5 min and when pipetting, avoid pellet

COMMENTARY

BACKGROUND INFORMATION:

While most lasers produce excitation light at one or a small number of discreet wavelengths, fluorophores are excited by and emit photons over a relatively broad range (e.g., Fig. 1). This leads to challenges in multiplexing probes due to the spectral overlap of emitted light. Well-equipped imaging systems can, in theory, distinguish a large number of multiplexed fluorophores in a single sample, allowing spatial relationships between fluorophores (and by extension their antigenic targets of interest) to be determined. In practice, however, a combination of system inefficiencies, fluorophore properties, autofluorescence/background signal, and optimizations common in commercial systems frequently limit the number of multiplexed probes analyzable in a single sample to three (in approximately the blue-green, red, and infrared regions of the electromagnetic spectrum).

Exceptionally abundant target biomolecules or bright probes can, in some cases, overcome these limitations allowing greater multiplexing. For example, 4′,6-diamidino-2-phenylindole (DAPI) interacting with DNA, an extremely abundant biomolecule in tissues, results in exceptionally bright ultraviolet (UV) fluorescence that can overcome the high level of UV autofluorescence in nervous tissues(Bottiroli et al., 1998; Pascu et al., 2009). Specialized probes (e.g., quantum dots, see above), optics, lasers, and detectors can in other instances enhance multiplexing capabilities, but these may not be available, or may require compromises in efficiency, probe choice, and interpretability that similarly restrict their application.

The use of a long Stokes shift fluorophore offers a relatively low-friction, accessible alternative, allowing biomolecules of average to low abundance to be specifically targeted, imaged and analyzed. This approach has several important advantages, the most important of which is its accessibility: The method relies on standard immunolabeling and is thus highly adaptable to experimental needs. Moreover it requires no additional hardware/optics or analysis approaches beyond those used in standard three-label IF applications, thus most microscopes equipped to analyze fluorescence in standard wavelengths will be equipped to image the long Stokes shifted fluorophore described above (ATTO 490LS). Limitations include the need to identify a series of primary and secondary antibodies that do not exhibit cross-binding (utilization of biotin-avidin binding is advantageous in this regard, and the above protocol includes steps to identify cross-binding). Where problematic, such cross-binding can often be overcome by sequential labeling or direct primary antibody-fluorophore conjugation. Several kit-based labeling systems for conventional or long Stokes shift fluorophores are available to accomplish this.

The above approach provides a low-friction method to enhance the information that can be gleaned from precious biospecimens. This should assist in particular with the structural characterization of complex microscopic apparatuses such as synapses that are defined by the functional apposition of numerous and diverse molecular components(Sauerbeck et al., 2020; O’Rourke et al., 2012).

CRITICAL PARAMETERS:

Antibody Selection

Proper antibody selection is critical to ensuring minimal antibody cross-binding and nonspecific staining. When choosing primary and secondary antibodies, be sure to choose antibodies raised in species that can be adequately distinguished by secondary reagents. One can further mitigate species cross-binding by using highly cross-absorbed secondary antibodies, such as those listed in Basic Protocol 1.

Microscope Configuration (refer to Figure 3)

Light sources: The 488 (or similar) laser line (or the Ti:Sapphire laser with appropriate tuning, or an epifluorescent light source with appropriate emission filter) will be needed to image the Alexa Fluor 488- and ATTO 490LS-labeled biomolecules. The 561 laser line (same alternates as above) will be needed to image the Alexa Fluor 594-labeled biomolecule. The 633 laser line (same alternates) will be needed to image the Alexa Fluor 633-labeled biomolecule.

Light path: Choosing filter sets and/or beam splitters that maximize signal from a given fluorophore while minimizing signal from others is an integral part of multiplexed immunofluorescence imaging. While scanning, many systems change laser lines and filter sets between frames. Especially when two fluorophores are excited by the same laser line (in the above example, Alexa Fluor 488 and ATTO 490LS), it is imperative that signal from only one fluorophore is able to reach the detector. This is achieved through the use of appropriate emission filters (band pass, long pass, short pass, variable, or a combination) that only permit the desired wavelengths to reach the detector(s). Fig. 3 demonstrates a successful strategy to achieve this on a confocal microscope with the fluorophores used in this example. For other implementations, it is often helpful to refer to an online fluorescence spectra viewer (e.g., https://www.thermofisher.com/us/en/home/life-science/cell-analysis/labeling-chemistry/fluorescence-spectraviewer.html#!/) to assist with choosing the most appropriate filter sets and/or beam splitter combinations for a given fluorophore. Such viewers are helpful in setting up all aspects of an IF experiment, but may not include all reagents or equipment available (e.g., ATTO 490LS is generally not included in online spectra viewers). If inappropriate light path configurations are selected, spectral cross talk is likely (see Troubleshooting for methods to determine whether this is occurring).

Acquisition parameters: Adjust the laser power, detector gain, scan speed, matrix size, pinhole size, and other system or application-specific parameters to achieve adequate signal to collect the image. If photobleaching/toxicity is noted after collecting a test image, move to a neighboring non-critical region/sample and reduce the laser power with concomitant changes to detector gain, pinhole size, or other parameters to compensate. The ideal parameters will represent a balance of speed, sensitivity, and resolution appropriate to the available signal, and must be experimentally determined for each system, sample, and experimental goal. For 3-dimensional acquisitions (Z-stacks) using a confocal microscope, it is important to remember that photobleaching/toxicity will occur at all Z-positions regardless of which optical section is currently being acquired. Thus acquisition parameters must be tested using the full planned experimental settings, not only for a single optical section.

Fluorophore Imaging Order

As the 488 laser is used to excite both Alexa Fluor 488 and ATTO 490LS, but only one fluorophore is imaged at a time, photobleaching is a concern. While negligible in our experience, photobleaching can be minimized by imaging the fluorophore with the lower laser power requirement (generally the least abundant or antigenic biomolecular target) first.

TROUBLESHOOTING:

The primary issue that may arise with this approach is spectral or antibody cross-talk/cross-binding (see Understanding Results). Assuming there was no opportunity for antibody contamination, signal from the no-primary controls (appearing to come from a primary antibody that was specifically omitted) can result from three potential sources: antibody cross-binding, spectral cross-talk, or background sources such as autofluorescence. The common signatures of these sources, as well as steps to mitigate them, are described below.

Antibody cross-binding occurs when a secondary antibody binds to an antibody other than its target and manifests as an identical (often much dimmer) image of an off-target antigen appearing in the no-primary image. That is, the secondary antibody-fluorophore targeted to the omitted primary can be detected at sites where off-target antibodies are present. To mitigate antibody cross-binding, choose antibodies carefully (see Critical Parameters: Antibody Selection). To further address problematic antibody-cross binding, antibody labeling can be separated into sequential steps. For example, in the above protocol, the Goat-anti-Chicken 647 antibody demonstrated a small amount of non-specific binding. To mitigate this, we applied the Chicken-anti-GFAP antibody followed by the secondary Goat-anti-Chicken 647 antibody before other primary or secondary antibodies were applied, and thoroughly rinsed the tissue to ensure any unbound antibody was removed.

Spectral cross-talk occurs when filters inadequately separate fluorescence from two or more fluorophores. This can appear similar to antibody cross-binding. Imaging parameters, informed with online spectra viewers, should be adjusted to eliminate this cross-talk. If this fails, alternate fluorophore combinations may be required. See Critical Parameters: Microscope configuration for details. If it remains unclear whether spectral cross-talk or antibody cross-binding is responsible for artifactual signal, single color controls (see Strategic Planning) can help resolve the issue as antibody cross-talk is eliminated as a potential source.

Autofluorescence is light emission by endogenous tissue components (collagen, lipofuscin, etc.) or processing-related chemicals (fixative). General background can also arise from numerous other sources, including endogenous antibody sources or receptors. To evaluate the contribution of autofluorescence or other background signals, compare images from the no-label control to the problematic image. Autofluorescence can be mitigated by avoiding problematic structures, adjusting tissue processing steps, applying certain pre-treatments, or in some cases by computational processing post-acquisition.

UNDERSTANDING RESULTS:

No-primary controls, a series of controls that receive all primary antibodies except one (varies, see Table 1) and all secondary antibodies, elucidate the contribution of spectral cross-talk and non-specific antibody binding. While some degree of autofluorescence is common in fixed brain tissue, images from no-primary controls should exhibit very low levels of signal for the missing primary. Other targeted antigens should not be contaminated by signal from the omitted primary (Fig. 4).

Figure 4:

No-primary controls revealing no spectral cross-talk or non-specific labeling. Columns represent individual experimental or no-primary control samples (the same field is represented down the column). Fields labeled with * represent antigens for which a primary antibody was omitted during labeling. Note lack of signal in these images.

To properly assess no-primary images, it is essential that the tissue from each group (experimental samples and each no-primary control) is from the same experimental condition and labeling run, and imaged the same day using identical settings. Following image acquisition, all signals from each sample image should be intensity adjusted using identical parameters to those employed for experimental images.

TIME CONSIDERATIONS:

When using the four primary antibodies used in Basic Protocol 1, tissue labeling requires three overnight incubations, lasting four days in sum. Performing this labeling fully in parallel would reduce this to one overnight incubation (see Basic Protocol 1 and Troubleshooting for notes on choosing parallel or sequential antibody labeling). Following labeling, we recommend the mounting media be left to cure for at least three days to reach an optimal refractive index before imaging. Duration of image acquisition will vary between systems, imaging parameters, and the amount of imaging data required.

Data Availability Statement

Data available on request from the authors