Immunosignal hybridization chain reaction (isHCR) combines antibody–antigen interactions with hybridization chain reaction (HCR) technology, which results in amplification of immunofluorescence signals by up to two to three orders of magnitude with low background. isHCR's highly modular and easily adaptable design enables the technique to be applied broadly, and we further optimized its use in multiplexed imaging and in state-of-the-art tissue expansion and clearing techniques.
Subscribe to Journal
Get full journal access for 1 year
only $21.58 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Bobrow, M.N., Harris, T.D., Shaughnessy, K.J. & Litt, G.J. J. Immunol. Methods 125, 279–285 (1989).
Carvajal-Hausdorf, D.E., Schalper, K.A., Neumeister, V.M. & Rimm, D.L. Lab. Invest. 95, 385–396 (2015).
Sylwestrak, E.L., Rajasethupathy, P., Wright, M.A., Jaffe, A. & Deisseroth, K. Cell 164, 792–804 (2016).
Chen, F., Tillberg, P.W. & Boyden, E.S. Science 347, 543–548 (2015).
Pan, C. et al. Nat. Methods 13, 859–867 (2016).
Choi, H.M.T., Beck, V.A. & Pierce, N.A. ACS Nano 8, 4284–4294 (2014).
Lu, C.-H., Yang, H.-H., Zhu, C.-L., Chen, X. & Chen, G.-N. Angew. Chem. Int. Ed. Engl. 48, 4785–4787 (2009).
Varghese, N. et al. ChemPhysChem 10, 206–210 (2009).
Zakeri, B. et al. Proc. Natl. Acad. Sci. USA 109, 4347–4348 (2012).
Keppler, A. et al. Nat. Biotechnol. 21, 86–89 (2003).
Tanenbaum, M.E., Gilbert, L.A., Qi, L.S., Weissman, J.S. & Vale, R.D. Cell 159, 635–646 (2014).
Fridy, P.C. et al. Nat. Methods 11, 1253–1260 (2014).
Sutcliffe, B. et al. Genetics 205, 1399–1408 (2017).
Richardson, D.S. & Lichtman, J.W. Cell 162, 246–257 (2015).
Treweek, J.B. & Gradinaru, V. Curr. Opin. Biotechnol. 40, 193–207 (2016).
Chang, J.-B. et al. Nat. Methods 14, 593–599 (2017).
Renier, N. et al. Cell 159, 896–910 (2014).
Agasti, S.S. et al. Chem. Sci. 8, 3080–3091 (2017).
Shah, S., Lubeck, E., Zhou, W. & Cai, L. Neuron 92, 342–357 (2016).
Lin, R. et al. Protoc. Exch. https://dx.doi.org/10.1038/protex.2018.014 (2018).
Juillerat, A. et al. Chem. Biol. 10, 313–317 (2003).
Gibson, D.G. et al. Nat. Methods 6, 343–345 (2009).
Viswanathan, S. et al. Nat. Methods 12, 568–576 (2015).
Wang, D. et al. Front. Neuroanat. 9, 40 (2015).
Shi, J. et al. Nature 514, 187–192 (2014).
Liu, Z. et al. Neuron 81, 1360–1374 (2014).
Renier, N. et al. Cell 165, 1789–1802 (2016).
Chozinski, T.J. et al. Nat. Methods 13, 485–488 (2016).
Ramanathan, S.P. et al. Nat. Cell Biol. 17, 148–159 (2015).
We thank R. Vale (UCSF, San Francisco, California, USA) for the pHR-scFv-GCN4-sfGFP-GB1-NLS-dWPRE plasmid, K. Deisseroth (Stanford University, Stanford, California, USA) for the pAAV-EF1α-DIO-hChR2(H134R)-mCherry plasmid, and J.H. Snyder for manuscript editing. M.L. is supported by China MOST (grants 2012YQ03026005, 2013ZX0950910, and 2015BAI08B02), NNSFC (grants 91432114 and 91632302), and the Beijing Municipal Government.
The authors declare no competing financial interests.
Integrated supplementary information
(a) Validation with western blotting of purified scFv-GCN4-HA-GB, which was recognized by a biotinylated anti-HA primary antibody and visualized with Alexa Fluor 546-conjugated streptavidin (SA-546) or with Alexa Fluor 546-conjugated HCR amplifiers (isHCR-546). (b, c) To determine the dynamic range for Western blotting using isHCR, purified scFv-GCN4-HA-GB1 protein was serially diluted. The analyte proteins were recognized by a biotinylated anti-HA primary antibody and visualized with isHCR-546 (b) or a traditional HRP-ECL method (c). The peak intensity area of each lane was calculated and plotted using the signals acquired with short (up panels of b, c) and long (bottom panels of b, c) exposure time. The lanes that contained the lowest amount of analyte protein detected are highlighted in red color. The linear dynamic range using either detection method under each exposure time is plotted using a log2 scale and highlighted in blue color (b, c). Note that the isHCR amplification method covers a dynamic range of 7 units (b right panel), whereas the HRP-ECL method covers a range of 4 or 5 units depending on the exposure time (c right panel).
Supplementary Figure 2 isHCR dramatically increases the fluorescent intensity while maintains the spatial resolution of membrane-bound GFP (mGFP) immunosignals in cultured cells.
(a) HEK293T cells expressing mGFP were immunostained for GFP using a biotinylated anti-GFP antibody with SA-546 or isHCR-546. (b) mGFP proteins in HEK293T cells were immunostained and visualized using an antibody mixture that contained equal amounts of Alexa Fluor 488-conjugated secondary antibodies (GAR-488, green) and biotinylated secondary antibodies. The signals of biotinylated secondary antibodies were visualized using isHCR-546 (red). For comparison purpose, images were collected with different parameters to achieve similar intensity profiles for the two channels. (c) A series of straight lines perpendicular to the membrane were drawn as illustrated for (b). The corresponding intensity profiles were plotted. The average intensity of each channel was calculated. The mean intensity of the distance from the cell edge (between -3 μm to -2 μm) was calculated as the baseline (denoted m). The peak value of the curve was determined (denoted p). Half-peak intensity (I) was defined (p+m)/2. The full width at half maximum (FWHM) was quantified as the width of the average intensity curve at I. (d) No significant difference was observed between the FWHM calculated using the data from unamplified and isHCR-amplified samples (n = 12; P = 0.9163; two-sided paired t-test). Bars represent the mean FWHM and error bars indicate s.e.m. in (d). Scale bars, 100 μm (a), 10 μm (b).
Supplementary Figure 3 TH immunostaining using isHCR reveals much richer catecholaminergic innervations in the brain.
(a) Images of mouse brain sections immunostained against TH using SA-488 (upper panels) or isHCR-488 (lower panels). Right panels show zoomed-in views of the boxed regions. (b) Confocal images of the TH expression pattern in the VTA, the interpeduncular nucleus (IPN; upper panels), and the superior colliculus (lower panels) in mouse brain sections immunostained against TH using isHCR-488 or SA-488. (c) No immunopositive signals were observed when we omitted the primary antibody. (d) Images of mouse brain sections immunostained against TH using standard IHC and isHCR. A mixture containing equal amounts of Cy3-conjugated secondary antibodies (GAR-Cy3, unamplified) and biotinylated secondary antibodies was used to detect the anti-TH primary antibodies. The signals of the biotinylated secondary antibodies were amplified using isHCR-488. The signals of unamplified (red) and amplified (green) channels were colocalized. Note that we used different imaging parameters to achieve similar intensity profiles for the two channels. (e) A mixture containing equal amounts of Cy3-conjugated secondary antibodies (GAR-Cy3, unamplified) and biotinylated secondary antibodies was used to detect the anti-TH primary antibodies. The signals of the biotinylated secondary antibodies were amplified using isHCR-488. The laser powers were adjusted to produce similar intensity profiles for the two channels. The width of TH-positive neuronal processes was measured using signals from either unamplified or isHCR-amplified channels. An example of two measured fibers are shown. No significant difference of the mean FWHM was observed in either the DS or the PVN (n = 196, P = 0.1069 for DS; n = 256, P = 0.8327 for PVN; two-sided paired t-test). Bars represent the mean FWHM and error bars indicate s.e.m. in (e). Scale bars, 1 mm (a left panel), 100 μm (a right panel, b, c), 50 μm (d), 10 μm (e).
(a) The effect of unassembled HCR amplifiers and the addition of graphene oxide (GO) on the background fluorescence level after washing. (b) A schematic showing that GO adsorbs the single-stranded overhang of HCR amplifiers and quenches the fluorescence of HCR amplifiers that are not assembled into double-stranded polymers. Arrows indicate the 5’ to 3’ direction. (c) In microplate wells, GO quenched the fluorescence of unassembled but not polymerized HCR amplifiers. (P values as listed; One-way ANOVA). (d) Images of the anterior cingulate cortex (ACC) in mouse brain sections immunostained against NeuN using isHCR-546 without GO or isHCR-546 with GO. The addition of GO significantly reduced the background signals and enhanced the signal-to-noise ratio (P values as listed; two-sided t-test). (e) Images of the anterior cingulate cortex (ACC) in mouse brain sections immunostained against NeuN using isHCR-546 without GO or isHCR-546 with GO at different concentrations. (f) The quantification of the signal intensity, background intensity, and the S/N ratio of the images shown in (e). The addition of GO at 10 and 20 μg mL−1 significantly suppressed the background signals and therefore enhanced the S/N ratio. Note that the addition of GO at a high concentration (50 μg mL−1) lowered the signal intensity, which was possibly a result of the competition between GO and isHCR initiators for the DNA-fluorophore HCR amplifiers (P values as listed; One-way ANOVA). Bars represent the mean value, and error bars indicate s.e.m.. Scale bars, 50 μm (a, d), 20 μm (e).
(a) Mouse brain sections were immunostained against NeuN. Parallel experiments were performed using anti-NeuN primary antibodies with different dilution ratios. The signals were visualized using isHCR-546 (top) or SA-546 (middle and bottom). The images of isHCR-546 and SA-546 samples from the ACC area were acquired using identical microscopy settings. Bottom panels show the higher contrast images of SA-546 samples. (b) The quantification of the signal intensity in three regions of the samples shown in (a). The signal intensity of isHCR-amplified samples was significantly higher than that of unamplified SA-546 samples (P values as listed; t-test corrected for multiple comparisons using the Holm-Sidak method). (c) The amplification factors of isHCR-546 vs. SA-546 at different antibody dilutions. Bars represent the mean signal intensity and error bars indicate s.e.m.. Scale bar, 100 μm.
(a) Confocal images of the dorsal raphe (DRN) and median raphe nucleus (MRN) in mouse brain sections immunostained against Vglut3 using isHCR-546. Graphene oxide (GO) was mixed with HCR amplifiers to quench the background fluorescence. (b) No Vglut3-immunopositive signals were observed in brain sections with the primary antibody omitted (upper panel) or in the brain sections of Vglut3-/- mice immunostained using isHCR-546 (lower panel). (c) The quantification of data from brain section samples immunostained against Vglut3 using SA-546, isHCR546 without GO, or isHCR with GO (n=3 brain sections for each group; P < 0.0001; Kolmogorov-Smirnov test). (d) Comparison between isHCR and standard IHC. A mixture containing equal amounts of Alexa Fluor 488-conjugated secondary antibodies (GAR-488) and biotinylated secondary antibodies was used to detect the anti-Vglut3 primary antibodies. The signals of the biotinylated secondary antibodies were amplified using isHCR-546. The signals of unamplified (green) and amplified (red) channels were colocalized. (e) The mean diameter of Vglut3-positive puncta was measured using the data from unamplified and isHCR-amplified samples. A representative measured punctum is shown. A straight line across the punctum was drawn and rotated using the punctum as the center of rotation. For every 6 degrees, the intensity profile along the line of both the unamplified and isHCR-amplified channel was plotted. The average intensity of each channel was calculated, and baseline correction was then applied. The corrected average intensity of each channel was fit to a Gaussian distribution with a non-linear least square method. FWHM was calculated with the equation: , where σ is the standard deviation of the fitted Gaussian curve. The results of FWHM for the two channels were compared. No difference of the mean diameter of Vglut3-positive puncta was observed (n = 20; P = 0.1846; two-sided t-test). Bars represent the mean FWHM and error bars indicate s.e.m. in (e). Scale bars, 50 μm (a, d), 100 μm (b), 10 μm (e).
Mouse brain sections were immunostained for (a) aromatic L-amino acid decarboxylase (AADC; an enzyme in axon terminals), and (b) neuronal nitric oxide synthase (nNOS; a cytosolic enzyme). The signals were detected using isHCR-546 (upper) or SA-546 (middle and bottom). The images of upper and middle panels were collected from SA-546 and isHCR-546 samples using identical microscopy settings. The bottom panels show images of the same regions in middle panels collected using higher settings. Scale bars, 20 μm.
Supplementary Figure 8 isHCR dramatically enhances the detection of GAD67 with a biotinylated monoclonal antibody.
(a) Wide-field images of mouse brain sections immunostained against GAD67 using a biotinylated monoclonal primary antibody. The signals were visualized with isHCR-546 (top) or SA-546 (bottom). (b) High-power confocal images showing GAD67 immunoreactivity in boxed areas in (a). The images of SA-546 and isHCR-546 samples were collected using identical microscopy settings. The right panels show images of the same regions in middle panels collected using higher settings. (c) Images of the lateral habenula (LHb) in mouse brain sections immunostained against GAD67 using SA-546 and isHCR-546. The right panel shows the quantification of data from the images shown in the left panel (n=3 brain sections for each group; P < 0.0001; Kolmogorov-Smirnov test). (d) Zoom-in images of brain sections immunostained against GAD67. Scale bars, 1 mm (a), 100 μm (b), 50 μm (c), 20 μm (d).
(a) Images of mouse brain sections immunostained against c-Fos using a monoclonal antibody. The signals were visualized using isHCR-546 or SA-546. The left panel shows the images of the mice that explored the enriched environment. The right panel shows the images of the control mice that were not allowed to explore the enriched environment. (b) Images of HeLa cells infected with SopD2-FLAG S. Typhimurium mutant strain. The detection of FLAG-tagged SopD2 was achieved using isHCR-488 (up) but not the standard IHC using Alexa Fluor 488-labeled secondary antibody (unamplified, Donkey anti-mouse, DAM-488, bottom). (c) No specific signal was detected in HeLa cells infected with wild-type S. Typhimurium (up) or in HeLa cells that were not infected by S. Typhimurium (bottom). Scale bar, 50 μm (a), 10 μm (b) and applied to (c).
(a) Schematic overview of isHCRn. The initial and branching rounds use DNA-biotin HCR amplifiers for additional attachment of HCR initiators. The final round uses DNA-fluorophore HCR amplifiers to visualize the signals. (b) Western blots of purified scFv-GCN4-HA-GB1 protein after one or three rounds of isHCR (isHCR1 vs. isHCR3). (c) In order to evaluate the accessibility of biotin groups on the amplification polymers, we performed immunostaining against TH on brain sections using isHCRn. We tested the performance of two pairs of amplifiers with biotin attached at either an internal or the terminal site. After each round of amplification, SA-546 was applied to detect the biotin groups. Tagging biotin at the internal position led to substantially higher amplification efficiency than at the terminal site. (d) Mouse brain sections were immunostained against TH. The signals were first amplified by isHCR using biotin and Alexa-fluor-488 dual-labeled HCR amplifiers, and then further amplified by an additional round of isHCR using Alexa-fluor-546 labeled HCR amplifiers. The signals of the first and second round of amplification were colocalized. We used different imaging parameters to achieve similar intensity profiles for the two channels. (e) The width of TH-positive neuronal processes was measured using signals from either isHCR1-488 or isHCR2-546 channels. An example of a measured fiber is shown. No significant difference of the mean FWHM was observed (n = 50; P = 0.7119; two-sided paired t-test). (f) Images of mouse brain slices immunostained against substance P without amplification (SA-546) or with 1-3 rounds of amplification (isHCR1-546, isHCR2-546, and isHCR3-546). (g) Mouse brain sections were immunostained against substance P using isHCR3. Multi-round amplification by isHCR3 revealed a patchy distribution of substance P in the striatum. The right panel shows the zoomed-in image of the boxed region at the left. Arrows indicate the 5’ to 3’ direction in (a) and (c). Bars represent the mean FWHM and error bars indicate s.e.m. in (e). Scale bars, 500 μm (c, f), 100 μm (g left), 50 μm (d, g right), 10 μm (e).
(a) Images of the dorsal striatum in mouse brain sections double immunostained for TH (green) and choline acetyltransferase (ChAT, red). The signals of each antigen were visualized sequentially using corresponding biotinylated secondary antibodies and isHCR. (b) Two orthogonal HCR initiators can be conjugated directly onto secondary antibodies using either SMCC or Click Chemistry groups as linkers. (c) Images of HEK 293T cells immunostained against the nuclear protein Ki67 (red) and a mitochondria protein Tom20 (green) using two HCR initiator-conjugated secondary antibodies. The signals were then simultaneously amplified using isHCR-546 for Ki67 and isHCR-488 for Tom20. (d) Western blot of a protein mixture containing purified hGBP1 and purified GFP-hGBP1 proteins. An anti-GFP primary antibody and an anti-hGBP1 primary antibody were applied. The two primary antibodies were detected using two secondary antibodies that were conjugated with different HCR initiators. (e) Images of the dorsal striatum in mouse brain sections double immunostained against dopamine transporter (DAT, red) and neuronal nitric oxide synthase (nNOS, green) using two HCR initiator-conjugated secondary antibodies. The signals were then amplified simultaneously using isHCR-546 for DAT and isHCR-488 for nNOS. Arrows indicate the 5’ to 3’ direction in (a) and (e). Scale bar, 50 μm (a, e), 10 μm (c).
(a) HCR initiators were conjugated to GFP-nanobodies (LaG-16-2) using SMCC as the linker. mGFP proteins were expressed in the orbitofrontal cortex neurons in CaMKII-Cre transgenic mice using adeno-associated virus (AAV) vectors. The GFP signals in the superior colliculus were amplified using HCR initiator-conjugated GFP-nanobody and isHCR-546. (b) Schematic of rapid labeling using genetically encoded protein tags. Functionalized HCR initiators bind directly to protein tags and initiate the amplification process. AAV vectors that bear the Cre-dependent double-floxed inverse (DIO) open reading frame cassette containing genes encoding the SNAPf tag were constructed and packaged into AAV particles. The AAV vectors were injected into Cre-transgenic mice to achieve cell-type specific expression. (c) Confocal images of HEK293T cells and mouse brain labeled by SNAP-tag. The upper panel shows cells transiently expressing the mSNAPf-tag. The middle panel shows the DRN from SERT-Cre mice injected with AAV-DIO-mSNAPf. The bottom panel shows the medial septum from SERT-Cre mice injected with AAV-DIO-4xSNAPf. The tag-positive cells were labeled with benzylguanine-conjugated Alexa Fluor 546 (BG-546) or isHCR-546. Arrows indicate the 5’ to 3’ direction in (a) and (b). Scale bars, 200 μm (a), 100 μm (c).
(a) A modified isHCR strategy for ExM. HCR initiators are dual-labeled with biotin at the 3’-end and acrydite at the 5’-end. The initiators bind to antibodies through the biotin group. During the gelation process, initiators are incorporated into gels through the acrydite group, and therefore survive the subsequent protein digestion. HCR amplifiers are applied after digestion. Arrows indicate the 5’ to 3’ direction. (b) Stack images of Vglut3-positive signals in VTA in the mouse brain sections immunostained against Vglut3 using isHCR and ExM. (c) Images of the ventral tegmental area (VTA) in expanded brain sections immunostained for Vglut3 using SA-546 or isHCR-546. isHCR-amplified samples were incubated with DNA-Alexa Fluor546 HCR amplifiers for various times. The diameter of Vglut3-positive puncta was measured. No significant difference of the mean FWHM was observed (n = 60 for each group; P = 0.1355, One-way ANOVA). (d) isHCR dramatically increased the fluorescent signals (366 puncta from SA-546 samples and 424 puncta from isHCR-546 samples were analyzed; P < 0.0001; two-sided t-test). No difference of the mean diameter of Vglut3-positive puncta was observed (45 puncta from SA-546 samples and 100 puncta from isHCR-546 samples were analyzed; P = 0.0507; two-sided t-test). (e) Images of uDISCO-cleared thick brain sections (500 μm) immunostained for TH using SA-546 or isHCR546. Crosslinking HCR amplification polymers with adjacent proteins by formaldehyde fixation preserved the amplified fluorescence signals after uDISCO clearing. (f) The lung of an adult mouse (left lobe) was immunostained against Prosurfactant Protein C (proSP-C) and cleared using uDISCO. The middle panel shows the cross-section images of the two sections indicated in the left panel. The right panel shows a zoomed-in image of the boxed region in the middle panel. Bars represent the mean values and error bars indicate s.e.m.. Scale bars are adjusted according to an expansion factor of 4 for ExM samples and a shrinkage factor of 2 for uDISCO samples. Scale bars, 10 μm (b), 2.5 μm (c), 50 μm (e, f right panel), 1 mm (f left panel), 500 μm (f middle panel).
Samples are first incubated with a given primary antibody and then its corresponding biotinylated secondary antibody. After washing, samples are incubated sequentially in streptavidin and DNA-biotin HCR initiators at room temperature (RT). DNA-fluorophore HCR amplifiers are applied to visualize the signals. The amplifiers can be mixed with graphene oxide (GO) to reduce background fluorescence. If multiple rounds of amplification are desired, biotin-labeled HCR amplifiers are used, except for the final (visualization) round. For ExM samples, biotin and acrydite dual-labeled HCR initiators are used. HCR amplifiers are applied after gelation and digestion. For uDISCO-cleared samples, an additional formaldehyde fixation step is performed after HCR amplifier incubation. Fixed samples are then cleared according to the standard uDISCO protocol. If HCR initiator-conjugated secondary antibodies are used, the HCR amplifiers can be directly applied after wash. For samples that are labeled by genetically encoded tags, we apply initiator-conjugated binding partners and then HCR amplifiers after wash.
Supplementary Figure 15 Schematic of isHCR initiators and amplifiers designed for various applications.
The sequences labeled with same color are complementary to each other. (a) The designs and modifications of DNA HCR initiators. Initiators can be modified with different functional groups at their 5’-ends, which can then be attached to target proteins through corresponding linkers. (b) The designs and modifications of DNA HCR amplifiers. Amplifiers can be terminally labeled with fluorescent dyes. Optionally, additional biotin groups can be added at internal sites; these biotin groups function as the interface for multiple-round isHCR amplification.
About this article
Cite this article
Lin, R., Feng, Q., Li, P. et al. A hybridization-chain-reaction-based method for amplifying immunosignals. Nat Methods 15, 275–278 (2018). https://doi.org/10.1038/nmeth.4611
A molecular recognition-activatable DNA nanofirecracker enables signal-enhanced imaging in living cells
Chemical Communications (2020)
Nanobody-based electrochemical competitive immunosensor for the detection of AFB1 through AFB1-HCR as signal amplifier
Microchimica Acta (2020)
The Analyst (2020)
16S rRNA‐functionalized multi‐HCR concatemers in a signal amplification nanostructure for visual detection of Salmonella
Biotechnology and Applied Biochemistry (2020)
Sensitive and specific detection of tumour cells based on a multivalent DNA nanocreeper and a multiplexed fluorescence supersandwich
Chemical Communications (2020)