Hybridization Chain Reaction for Quantitative and Multiplex Imaging of Gene Expression in Amphioxus Embryos and Adult Tissues

Abstract

In situ hybridization (ISH) methods remain the most popular approach for profiling the expression of a gene at high spatial resolution and have been broadly used to address many biological questions. One compelling application is in the field of evo-devo, where comparing gene expression patterns has offered insight into how vertebrate development has evolved. Gene expression profiling in the invertebrate chordate amphioxus (cephalochordate) has been particularly instrumental in this context: its key phylogenetic position as sister group to all other chordates makes it an ideal model system to compare with vertebrates and for reconstructing the ancestral condition of our phylum. However, while ISH methods have been developed extensively in vertebrate model systems to fluorescently detect the expression of multiple genes simultaneously at a cellular and subcellular resolution, amphioxus gene expression profiling is still based on single-gene nonfluorescent chromogenic methods, whose spatial resolution is often compromised by diffusion of the chromogenic product. This represents a serious limitation for reconciling gene expression dynamics between amphioxus and vertebrates and for molecularly identifying cell types, defined by their combinatorial code of gene expression, that may have played pivotal roles in evolutionary innovation. Herein we overcome these problems by describing a new protocol for application of the third-generation hybridization chain reaction (HCR) to the amphioxus, which permits fluorescent, multiplex, and quantitative detection of gene expression in situ, within the changing morphology of the developing embryo, and in adult tissues. A detailed protocol is herein provided for whole-mount preparations of embryos and vibratome sections of adult tissues.

Keywords: Amphioxus; Fluorophore-labeled; Gene expression profiling; HCR; In situ hybridization; Multiplex; Single-cell resolution.

Fig. 1

Embedding and vibratome sectioning of adult amphioxus brain tissues. Once the specimen has been anesthetized (a) the head is separated from the rest of the body and fixed for 24 h as indicated in Subheading 3.3, step 3. The separated head is then washed and glued into the self-adhesive foam either vertically (b) or horizontally (c), depending on the orientation needed. For coronal sections further dissection of the ventral side is recommended to properly align brain and neural tube in a same section. The foam with the attached head is then transferred into a peel-a-way square embedding mold and filled up with liquid low-melting agarose (d). Once the agarose has solidified the block is extracted from the mold and trimmed as a pyramid to provide better grip while sectioning (e). The wider part of the pyramid is then glued to the vibratome holder and the head is sectioned with the dorsal fin facing the blade (f)

Fig. 2

Triple HCR in whole-mount embryos of B. lanceolatum. HCR staining was performed on whole amphioxus embryos reared in our facility [24] to 24 hpf at 21 °C with varying concentrations of probe, and in the presence and absence of proteinase K digestion. For this combination, 20-pair probe sets were designed against Soxb1c for broad labeling of the neural tube, Mnx for labeling of motor neuron progenitors, and Elav for postmitotic neurons. Exposure to hairpins in the absence of probes generates channel-specific background fluorescence profiles (a). This is conspicuous in all channels, but most severe following excitation at a 488 nm wavelength. A 4 nM probe concentration, as suggested for use on zebrafish embryos, generates specific signal but with a poor signal:noise ratio, even after extensive washing (b). In the 488 channel, signal can be difficult to distinguish from background. Signal:noise is greatly improved with expose to 20 nM (c) and 40 nM (d) probe concentrations for common incubation times. At 40 nM, signal can readily be resolved from background. Proteinase K (PK) treatment is common in amphioxus ISH protocols to enhance embryo permeability. However, treatment with PK did not enhance HCR signal beyond that achieved with only permeabilization in TritonX-100 and DMSO (e). (f) Magnified view of parasagittal section through anterior neural tube of embryo in (d), revealing regional differences in expression profiles for each gene, and a single triple-positive cell. Scale bars measure 100 μm (a–e) and 20 μm (f)

Fig. 3

Triple HCR in B. lanceolatum brain vibratome sections. HCR staining was performed on floating vibratome sections of adult amphioxus maintained in our amphioxus facility [24]. Sections were stained with 20-pair probe sets designed against Pax4/6, used here as a neuronal marker, Nkx2.1, used here as a specifier of GABAergic fate and GAD, used here as a marker of GABAergic neurons. All probes were used at a concentration of 40 nM, as it was found to show the best signal:noise ratio for most of the genes in whole-mount preparations of embryos (see Fig. 2). Sections that had been preincubated in methanol after fixation (b) (see Note 1) showed a better signal:noise ratio than those that were not exposed to methanol (a). Low-expressed genes such as Nkx2.1 are for example hardly visible when sections are not preincubated in methanol (compare a and b). Additional pretreatment with proteinase K (c) improves the sharpness of the signal for all genes, including the low-level-expressing Nkx2.1, the mid-level-expressing Pax4/6, and the highly expressed GAD. Magnified views of merged images in d, e, and f show co-localization of the transcripts at a single-cell resolution