Development and application of MaMBA and misHCR.

a, Schematics of the DNA oligo-conjugated nanobody production process. b, Coomassie stained SDS-PAGE gel of purified nanobody (Nb-NGL-His6), OaAEP1 reaction product (Nb-NGV-N3), and click reaction products (Nb-DNA oligo). c, MaMBA workflow. IgG antibodies are separately incubated with DNA oligo-conjugated nanobodies bearing unique DNA barcodes, forming antibody-Nb-DNA oligo complexes (Ab-Nb 1 to Ab-Nb 3). These complexes are subsequently purified using ultra centrifugal filters or IgG-conjugated beads, pooled, and applied to various assays. d, Fluorescence confocal images of human psoriatic skin section stained for DAPI and six antigens by three rounds of misHCR. IgG types of the employed antibodies are indicated in parentheses. Scale bar, 300 μm. Rb, rabbit; Mus, mouse; Th, tyrosine hydroxylase; PDGFRα, platelet-derived growth factor receptor A; KRT14, keratin 14; DCT, dopachrome tautomerase; αSMA, alpha-smooth muscle actin.

Multi-round immunostaining using misHCRn.

a, Schematics of the production process for cleavable DNA oligo-conjugated nanobodies. b, Workflow of misHCRn staining and imaging. The process begins with the first round of staining (1st round of misHCR), where a pool of primary antibodies barcoded with orthogonal cleavable HCR initiators (Ab-SS-i1 to Ab-SS-in) is applied to the sample simultaneously. This is followed by sequential rounds of HCR imaging (image rounds). Upon completion of the image rounds, HCR initiators are removed using TCEP to cleave the disulfide bonds, enabling the start of the second round of staining (2nd round of misHCR). Antibodies used in the 2nd round of misHCR can be equipped with the same set of HCR initiators as in the 1st round of misHCR (Ab-SS-i1’ to Ab-SS-in’). This process can be iteratively performed for multi-round misHCR. c, Fluorescence confocal images of the DRN in a mouse brain section, stained for 12 antigens by misHCR2 and counterstained with DAPI. The imaging location is highlighted in the red boxed region of the mouse brain atlas shown at the bottom left. IgG types of the employed antibodies are indicated in parentheses. Scale bar, 100 μm. d, Zoomed-in views of the region marked by a dotted-box in panel c. Arrows indicate three cell types (1, Ddc+Th+Tph2- cells; 2, Ddc+Th-Tph2+ cells; 3, Ddc+Th-Tph2-5-HT- cells). Scale bar, 50 μm. nNOS, neuronal nitric oxide synthase; GFAP, glial fibrillary acidic protein; NF-H, neurofilament-H; Tph2, tryptophan hydroxylase 2; GABA, gamma-aminobutyric acid; Tmem119, transmembrane protein 119; Ddc, dopa decarboxylase; 5-HT, 5-hydroxytryptamine; Iba1, ionized calcium binding adaptor molecule.

Barcode-linked immunosorbent assay (BLISA).

a, Workflow of direct antigen detection using BLISA. b, Schematics of multiplexed detection for GFP, mCherry, and α-tubulin using BLISA. c, Quantification of spiked-in purified GFP, spiked-in purified mCherry, and endogenous α-tubulin in cell lysates using qPCR-based BLISA. Negative control (NC) represents cell lysates without spiked-in proteins. ΔCt values are normalized to the NC group, where ΔCt (NC=1) = CtNC-CtTarget+1. The dose-response curves for purified GFP (green) and mCherry (red) are fitted using a five-parameter logistic (5PL) function. n = 3 replicates; mean±s.e.m.. d, Schematics of two drug-inducible gene expression systems for expressing GFP and mCherry in HEK293T cells. e, Western blot of cell samples treated with varying concentrations of ATc and FSK. f, Pearson’s correlation between BLISA and Western blot results for measuring relative expression levels (REL, normalized to α-tubulin) of GFP (left) and mCherry (right). REL=2-(CtGFP/mCherry-Ctα-tubulin). ATc, anhydrotetracycline hydrochloride; FSK, forskolin.

Magnetic bead-based BLISA for multiplexed detection.

a, Workflow of the magnetic bead-based BLISA. This panel was creatured using BioRender.com. d, Comparison of 7-plex BLISA (left) and ELISA (middle) for detecting the phosphorylation of seven endogenous proteins in U87 and U937 cells under various culture conditions. The right panel shows Pearson’s correlation coefficients between BLISA and ELISA results for each target protein. n = 3 replicates; mean±s.e.m.. e, Heatmap showing average Z-scores of the phosphorylation levels of seven proteins in U87 cells following different drug treatments. n = 2 biological replicates. d, ELISA quantification of phosphorylation levels for seven proteins in U87 cells. n = 3 biological replicates; two-sided t-tests. p-p38α, phospho-p38α (T180/Y182); p-ERK1/2, phospho-ERK1 (T202/Y204)/ERK2 (T185/Y187); p-JNK, phospho-JNK (Pan Specific); p-AMPKα1, phospho-AMPKα1 (T183); p-CREB, phospho-CREB (S133); p-Src, phospho-Src (Y419); p-Akt, phospho-Akt (S473); NT, non-treated; FBS, fetal bovine serum; Aniso, anisomycin.

HTS-based BLISA for high-throughput detection of IgGs in human serum samples.

a, Workflow of sandwich immunoassay-based IgG detection using BLISA. b, Pearson’s correlation between results from two technical duplicate. Red dots represent negative samples. c– d, Quantification of anti-RBD IgG levels in twelve samples from vaccinated donors using HTS-based BLISA (c) and ELISA (d). f, Pearson’s correlation between BLISA and ELISA results shown in (c) and (d). n = 2 replicates; mean±s.e.m.. f, Distribution of normalized IgG abundance in negative and test samples (n = 493). Red dots represent negative samples. The solid grey line indicates the mean normalized abundance of negative samples. The dotted grey line represents the mean plus three standard deviations (mean+3SD).

HTS-based BLISA for high-throughput simultaneous detection of two HBV antigens in human serum samples.

a, Workflow of sandwich immunoassay-based HBV antigen detection using BLISA. b–c, Sensitivity and specificity of two-plex BLISA for detecting HBsAg. n=462 ELISA-negative samples and 119 ELISA-positive samples from cohort 2. d–e, Sensitivity and specificity of two-plex BLISA on detecting HBeAg. n=505 ELISA-negative samples and 90 ELISA-positive samples from cohort 2. Solid grey lines indicate the mean normalized abundance of ELISA-confirmed negative samples. Dotted grey lines represent the mean+3 standard deviations (mean+3SD). f, Pearson’s correlation between technical duplicates for HBsAg (upper) and HBeAg (bottom) measurements in serum samples from cohort 2 (n= 600).

Efficiency assessment of nanobody-DNA oligo conjugation.

a, Coomassie-stained reducing SDS-PAGE gel of protein products at each step of OaAEP1mediated azide functional groups addition to nanobodies. The purified nanobody (i.e., input) was incubated with OaAPE1 and azide-bearing substrate peptides (i.e., Nb reaction mixture). After reaction, the products (i.e., Nb-NGV-N3) were purified by Ni-NTA agarose beads to remove unreacted His6-tagged nanobody (i.e., After Ni-NTA). The excess substrate peptides were further depleted using desalting columns (i.e., After desalting). b, Quantification of OaAPE1-mediated reaction yields, determined by densitometric measurement of Nb-NGV-N3 band intensities. c, Coomassie-stained reducing SDS-PAGE gel of Nb-NGV-N3 and nanobodies conjugated with DNA oligo of different lengths (i.e., Nb-oligo (62bp) and Nb-oligo (42bp)). d, Quantification of DNA conjugation yields, determined by densitometric measurement of band intensities corresponding to nanobody-DNA conjugates.

Comparison of antibody-DNA oligo labeling methods.

a, Schematics of non-site-specific labeling methods for generating DNA oligo-conjugated antibodies with increased labeling degree. This panel was creatured using BioRender.com. b, Silver-stained reducing SDS-PAGE gel of antibodies conjugated with 62bp DNA oligos using the amine-based method (Thunder-link antibody labeling kit) at various antibody:DNA ratios. c, Silver-stained reducing SDS-PAGE gel of antibodies conjugated with 62bp DNA oligos using the thiol-based method (thiol-maleimide reaction) at various antibody:DNA ratios. d, Western Blot of DNA oligo-labeled antibodies generated by MaMBA strategy using different antibody:nanobody-DNA oligo (Nb-oligo) ratios. Samples were separated by native PAGE and transferred to PVDF membrane for Western Blot. DNA oligos were 3’-biotin modified and detected using streptavidin. e, Western Blot of HEK 293T cell lysates with or without EGFP expression using different DNA oligo-labeled antibodies. The concentrations of anti-GFP antibodies were identical for each group. DNA oligos were 3’-biotin modified and detected using streptavidin. All groups were chemiluminescently detected simultaneously. Endogenous biotinylated carboxylases in cell lysates served as loading controls.

HTS-based BLISA for high-throughput detection of two HBV antigens in serum samples from cohort 1.

a, Scatter plot of normalized HBsAg and HBeAg abundances of individual serum samples (n= 529) quantified using HTS-based 2-plex BLISA. b, Frequency distributions of normalized HBsAg (blue) and HBeAg (red) abundances of test samples. c, Pearson’s correlation between technical duplicates for HBsAg (left) and HBeAg (right) measurements. d, HBsAg and HBeAg levels of control samples in cohort 1 measured by HTS-based BLISA. Histogram of normalized HBsAg and HBeAg abundances in negative (HBsAg-HBeAg-) and positive (HBsAg+HBeAg+) control groups were quantified using HTS-based BLISA. n = 25 for HBsAg-HBeAg- and 6 for HBsAg+HBeAg+; mean±s.e.m.; Two-sided unpaired t-tests. e–f, Sensitivity and specificity of two-plex BLISA for detecting HBsAg. n=87 ELISA-negative samples and 28 ELISA-positive samples from cohort 1. g–h, Sensitivity and specificity of two-plex BLISA for detecting HBeAg. n=108 ELISA-negative samples and 7 ELISA-positive samples from cohort 1. Solid grey lines indicate the mean normalized abundance of ELISA-confirmed negative samples. Dotted grey lines represent the mean+3 standard deviations (mean+3SD).

Performance evaluation of two-plex BLISA for detecting HBV antigens using ELISA-independent thresholds.

a–b, Sensitivity and specificity of two-plex BLISA for detecting HBsAg (a) and HBeAg (b) in serum samples from cohort 1. Solid grey lines indicate the mean normalized abundance of test samples (mean.t; n=529). Dotted grey lines represent the mean+standard deviations (mean.t+SD.t). c–d, Sensitivity and specificity of two-plex BLISA for detecting HBsAg (c) and HBeAg (d) in serum samples from cohort 2. Solid grey lines indicate the mean normalized abundance of test samples (mean.t; n=500). Dotted grey lines represent the mean+standard deviations (mean.t+SD.t).

Comparison of the amine-based, thiol-based, and MaMBA-based antibody labeling methods. This figures was creatured using BioRender.com.

Characterization of misHCR.

a, Schematics of isHCR and misHCR procedures. b–d, Fluorescence confocal images of mouse brain sections stained using misHCR or conventional fluorescent immunohistochemistry with fluorophore-conjugated nanobodies. n = 3 brain sections for each group; mean±s.e.m.; two-sided unpaired t-test. Scale bar, 100 μm. e, Fluorescence confocal images of HeLa cells stained against α-tubulin using a conventional method with a fluorophore-conjugated secondary antibody (2nd Ab) (left) or misHCR (right). Insets show zoomed-in views of the boxed region. Scale bar, 10 μm. Bottom panel shows representative intensity profiles along straight lines drawn perpendicular to microtubule structures. f, Quantification of full width at half maximum (FWHM) derived from intensity profiles along a straight line drawn perpendicular to microtubule structure stained using traditional immunohistochemistry (2nd Ab) or misHCR. n = 26 for 2nd Ab and 25 for misHCR; mean±s.e.m.; two-sided unpaired t-test.

Efficient and reproducible HCR signal removal by formamide treatment.

a, Workflow of misHCR imaging. Pooled antibodies (Ab-i1 to Ab-in), each conjugated with distinct HCR initiators, are applied simultaneously for antigen (Ag. 1 to Ag. n) detection. Subsequently, sequential rounds of imaging are performed via hybridization and dehybridization of orthogonal HCR amplifier pairs (Amp.). b, Fluorescence confocal images of mouse brain sections before and after HCR amplifiers removal using formamide. Scale bar, 500 μm. c, Fluorescence confocal images of the DRN in mouse brain sections stained against NeuN and NF-H through five consecutive cycles of HCR amplification/formamide treatment. Scale bar, 100 μm. d, Quantification of mean fluorescence intensity for NeuN (left) and NF-H (right) after each round of HCR amplification. Images of each group were acquired using identical microscopy settings. n = 3 brain sections for each group; mean±s.e.m.; Dunn’s multiple comparison tests.

MisHCR imaging of human psoriatic and healthy skin sections.

Fluorescence confocal images of skin sections from a patient with psoriasis (up) and a healthy donor (bottom) stained against six antigens. The psoriatic skin sample corresponds to that shown in Fig. 1d. Dotted lines and arrows indicate pathological features characteristic of psoriatic skins. Scale bar, 300 μm.

Efficient misHCR signal removal by TECP treatment.

Fluorescence confocal images of mouse brain sections before and after TCEP treatment to remove HCR amplifiers and initiators. The target molecules were MAP2, Ddc, and nNOS for the first round; NeuN, Iba1, and GFAP for the second round; Th, and NF-H for the third round. Scale bar, 300 μm. MAP2, microtubule-associated protein 2.

misHCRn for immunostaining nine target antigens using three orthogonal HCR initiators.

a, Fluorescence confocal images of mouse brain sections stained against nine targets by three rounds of misHCR. The red-boxed region of the mouse brain atlas shown in the bottom right indicates the imaging location. Scale bar, 300 μm. b, Zoomed-in views of the dotted-box region marked in a. Scale bar, 25 μm.

Validation of the specificity of misHCR.

Mouse brain sections were immunostained with primary antibodies against different antigens, followed by and simultaneous incubation with equal amounts of fluorophore-conjugated secondary antibodies (2nd Ab) and Nb-HCR initiators (misHCR). Scale bar, 100 μm.

BLISA for detecting proteins endogenously expressed in cells.

a, Schematics of the tetracycline-regulated gene expression system for driving GFP expression. b, Western blot of cell samples treated with varying concentrations of ATc. c, Pearson’s correlations between qPCR-based BLISA and Western blot results for measuring relative expression level (REL, normalized to α-tubulin) of GFP. REL=2-(CtGFP-Ctα-tubulin). d, Schematics of the cAMP response element-mediated gene expression system for driving mCherry expression. e, Western blot of cell samples treated with varying concentrations of FSK. f, Pearson’s correlations between qPCR-based BLISA and Western blot results for measuring REL (normalized to α-tubulin) of mCherry. REL=2-(CtmCherry-Ctα-tubulin).

Comparative analysis of standard curves for magnetic bead-based BLISA and ELISA.

a, Dose-response curves for seven purified phosphorylated proteins quantified using BLISA (qPCR, ΔCt) and ELISA (Absorbance). Curves were fitted using 4PL or 5PL functions. Limit of detection (LOD) was defined as the mean signal of the blank group plus three standard deviations. b, R2 for the standard curves shown in panel a. c, Multiplexed detection of seven purified phosphorylated proteins using magnetic bead-based BLISA. n = 3 replicates; mean±s.e.m..

Validation of HTS-based BLISA.

a, Schematics of DNA barcode design and amplicon generation for BLISA sequencing. Released DNA barcodes undergo a two-step PCR protocol for well barcode (well-bc) addition and Illumina sequencing adaptors tagging. b, UMI counts for three target proteins (GFP, mCherry, and α-tubulin) in cell lysates spiked in with varying concentrations of GFP and mCherry. n = 3 replicates; mean±s.e.m.. c, Dose-response curves for the GFP (upper) and mCherry (bottom) fitted using a 5PL function. n=3 replicates; mean±s.e.m..

Comparative analysis of the sensitivity of BLISA and ELISA for human IgG detection.

a, Schematics of ELISA (left) and BLISA (right) for human IgG detection. b, Dose-response curves for human IgG detection by BLISA and ELISA fitted using a 5PL function. The dynamic ranges of ELISA (blue) and BLISA (red) are indicated by dashed lines, spanning from the limit of quantification (LOQ) to the lower limit of 95% confidence interval. n = 3 replicates; mean±s.e.m.. c, Pearson’s correlation between qPCR-based BLISA and ELISA results for anti-RBD IgG detection. The red dot represents the negative sample.

Sequencing library preparation for BLISA.

a, Schematics of sequencing amplicon generation procedure. b, Annotated amplicon sequence of BLISA for Illumina sequencing.

Quantification of anti-RBD IgG in human serum samples using HTS-based BLISA.

a, Comparison of coefficient of variation (CV) between duplicates for raw (grey) and normalized (red) data. Histogram shows relative distributions of CV between duplicates. Box plots show median, interquartile range, and whiskers extending to the minimum and maximum. Two-sided paired t-test. b, Frequency distribution of normalized anti-RBD IgG abundances of test samples across assay plates. c, Scatter plot of normalized anti-RBD IgG abundances of individual serum samples (n=517) quantified using HTS-based BLISA. Each plate includes a negative control sample in duplicates.