Programmable in situ amplification for multiplexed imaging of mRNA expression

Abstract

In situ hybridization methods enable the mapping of mRNA expression within intact biological samples. With current approaches, it is challenging to simultaneously map multiple target mRNAs within whole-mount vertebrate embryos, representing a significant limitation in attempting to study interacting regulatory elements in systems most relevant to human development and disease. Here, we report a multiplexed fluorescent in situ hybridization method based on orthogonal amplification with hybridization chain reactions (HCR). With this approach, RNA probes complementary to mRNA targets trigger chain reactions in which fluorophore-labeled RNA hairpins self-assemble into tethered fluorescent amplification polymers. The programmability and sequence specificity of these amplification cascades enable multiple HCR amplifiers to operate orthogonally at the same time in the same sample. Robust performance is achieved when imaging five target mRNAs simultaneously in fixed whole-mount and sectioned zebrafish embryos. HCR amplifiers exhibit deep sample penetration, high signal-to-background ratios and sharp signal localization.

Figure 1. Multiplexed in situ hybridization using fluorescent HCR in situ amplification

(a) HCR Mechanism. Metastable fluorescent RNA hairpins self-assemble into fluorescent amplification polymers upon detection of a specific RNA initiator. Initiator I nucleates with hairpin H1 via base-pairing to single-stranded toehold ‘a’, mediating a branch migration that opens the hairpin to form complex I•H1 containing single-stranded segment ‘c*-b*’. This complex nucleates with hairpin H2 via base-pairing to toehold ‘c’, mediating a branch migration that opens the hairpin to form complex I•H1•H2 containing single-stranded segment ‘b*-a*’. Thus, the initiator sequence is regenerated, providing the basis for a chain reaction of alternating H1 and H2 polymerization steps. (b) Validation in a test tube. Agarose gel demonstrating orthogonal amplification in a reaction volume containing four HCR amplifiers and zebrafish total RNA. Minimal leakage from metastable states is observed in the absence of initiators. (c–e) Protocol summary. (c) Detection stage. Probe sets are hybridized to mRNA targets and then unused probes are washed from the sample. (d) Amplification stage. Initiators trigger self-assembly of tethered fluorescent amplification polymers and then unused hairpins are washed from the sample. (e) Experimental timeline. The same two-stage protocol is used independent of the number of target mRNAs. For multiplexed experiments (3-color example depicted), probe sets for different target mRNAs carry orthogonal initiators that trigger orthogonal HCR amplification cascades labeled by spectrally distinct fluorophores.

Figure 2. Validation of fluorescent HCR in situ amplification in fixed whole-mount zebrafish embryos

(a–i) The target is the transgenic transcript Tg(flk1:egfp), expressed below the notochord and between the somites (see the expression atlas of Fig. 3a). Embryo morphology is depicted by autofluorescence in the gray channel. Probe set: 1 RNA probe. Fluorescent staining (green channel) using in situ HCR in Target+ (a) and Target- (b) embryos compared to (green channel) autofluorescence in the absence of probes and hairpins (c). No amplification in the absence of probes (d) or of one hairpin species (e,f). Modification of hairpin stem sequences (H1’,H2’) disrupts (g,i) and restores (h) toehold-mediated branch migration, confirming that staining arises from triggered polymerization rather than from random aggregation of hairpins. Typical for zebrafish, the yolk sack (bottom left of each panel) often exhibits autofluorescence. (j–m) Characterizing signal-to-background for fluorescent HCR in situ amplification. The target is a muscle gene transcript (desm) expressed in the somites. Embryo morphology is depicted by autofluorescence in the gray channel. Pixel intensity histograms are calculated using the green channel. WT embryos. Probe set: 3 RNA probes, except panel m. (j) Sample penetration with in situ HCR: probes and hairpins penetrate the sample prior to executing triggered self-assembly of tethered amplification polymers in situ. (k) Sample penetration with ex situ HCR: probes trigger self-assembly of amplification polymers prior to penetrating the sample. (l) Background and signal contributions. Histograms of pixel intensity are plotted for a rectangle partially within the expression region and partially outside the expression region (e.g., see panels j and k). Background arises from three sources: autofluorescence (AF; buffer only), non-specific amplification (NSA; hairpins only); non-specific detection (NSD; in situ HCR amplification following detection of absent target Tg(flk1:egfp)). NSD studies employ a probe set of three RNA probes targeting transgenic transcript Tg(flk1:egfp), which is absent from the WT embryo. (m) Multiple probes per mRNA target. Comparison of autofluorescence and in situ HCR using probe sets with 1, 3, or 9 RNA probes (compare curves of the same color). The microscope PMT gain was decreased as the size of the probe set increased to avoid saturating pixels in the images employing in situ HCR amplification (this accounts for the reduction in AF intensity as the size of the probe set increases). Embryos fixed at 25 hpf. Scale bar: 50 μm.

Figure 3. Multiplexed imaging in fixed whole-mount and sectioned zebrafish embryos

(a) Expression atlas for five target mRNAs (lateral view: Tg(flk1:egfp), tpm3, elavl3, ntla, sox10). (b) mRNA expression imaged using confocal microscopy at four planes within an embryo. This multiplexed experiment is performed using the same two-stage protocol that is employed for single-color experiments (summarized in Figs 1c–e). Detection is performed using five probe sets carrying orthogonal initiators. The probe sets have different numbers of RNA probes (10,7,18,30,20) based on the strength of expression of each mRNA target and the strength of the autofluorescence in each channel. Amplification is performed using five orthogonal HCR amplifiers carrying spectrally distinct fluorophores. (c) Expression atlas for five target mRNAs (anterior view). (d) mRNA expression imaged within a 200-μm zebrafish section using confocal microscopy. Vibratome sectioning was performed after HCR in situ amplification and postfixation. See also the image stacks of Supplementary Movies 1 and 2. Embryos fixed at 27 hpf. Scale bars: 50 μm.

Figure 4. Sharp signal localization and co-localization in fixed whole-mount zebrafish embryos

Redundant two-color mapping of one target mRNA expressed predominantly in the somites (desm; two probe sets, two HCR amplifiers, channels 1 and 2) simultaneous with redundant two-color mapping of a second target mRNA expressed predominantly in the interstices between somites (Tg(flk1:egfp): two probe sets, two HCR amplifiers, channels 3 and 4). (a) Sharp co-localization of desm signal (Pearson correlation coefficient, r = 0.93). (b) Sharp co-localization of Tg(flk1:egfp) signal (Pearson correlation coefficient, r = 0.97). (c) Sharp signal localization within the two interleaved expression regions. The interstice between somites is only the width of a single stretched cell. Embryos fixed at 27 hpf. Scale bars: 10 μm.