Skip to main content
NCBI home page
RNA logoLink to RNA
. 2023 Jun;29(6):836–846. doi: 10.1261/rna.079482.122

Single-molecule RNA in situ detection in clinical FFPE tissue sections by vsmCISH

Meng Jiang 1,2, Kaipeng Wei 3, Meiqing Li 4, Chen Lin 1,✉, Rongqin Ke 1,✉
PMCID: PMC10187679  PMID: 36813533

Abstract

Although RNA plays a vital role in gene expression, it is less used as an in situ biomarker for clinical diagnostics than DNA and protein. This is mainly due to technical challenges caused by the low expression level and easy degradation of RNA molecules. To tackle this issue, methods that are sensitive and specific are needed. Here, we present an RNA single-molecule chromogenic in situ hybridization assay based on DNA probe proximity ligation and rolling circle amplification. When the DNA probes hybridize into close proximity to the RNA molecules, they form a V-shape structure and mediate the circularization of circle probes. Thus, our method was termed vsmCISH. We successfully applied our method to assess HER2 mRNA expression status in invasive breast cancer tissue and investigated the utility of albumin mRNA ISH for differentiating primary from metastatic liver cancer. The promising results on clinical samples indicate that our method has great potential for application in diagnosing diseases using RNA biomarkers.

Keywords: RNA in situ hybridization, rolling circle amplification, single-molecule detection

INTRODUCTION

Accurate measurements of biomarkers such as DNA, RNA, or proteins play a critical role in classifying and diagnosing diseases for precision medicine. With technological advancements and novel findings in single-cell biology, RNA biomarkers are increasingly being applied to cancer diagnosis, prognosis, and therapy guidance (Kwa et al. 2017; Wu et al. 2020; Froeling et al. 2021). Currently, there are many existing methods for measuring RNA molecules, such as northern blotting, microarrays, quantitative reverse transcriptase polymerase chain reaction (qRT-PCR), reverse transcription digital PCR, and traditional RNA in situ hybridization. Among these technologies, RNA in situ hybridization (RNA ISH) has been a commonly used laboratory technique for analyzing gene expression and localizing particular RNA molecules in cells (Chu et al. 2019). However, the development of RNA molecules in situ detection methods for clinical applications still needs to catch up compared to DNA in situ hybridization (DNA ISH) (Warford 2016) and immunohistochemistry (IHC) (Magaki et al. 2019). The disadvantages of traditional RNA ISH, including low sensitivity, poor specificity, and low signal-to-noise ratio, can be effectively avoided by signal amplification technology. In recent years, a commercial RNA ISH method called RNAscope has become increasingly popular. RNAscope is based on the branched DNA technology for signal amplification and a pair of double “Z” probes to ensure the specificity (Player et al. 2001; Wang et al. 2012; Anderson et al. 2016). It has been widely used in basic science research such as neuroscience (Shiers et al. 2020; Kameneva et al. 2021), stem cell (Fernandez-Barral et al. 2020; Li et al. 2021), and developmental biology (Kersigo et al. 2018; He et al. 2020), as well as in retrospective studies of clinical samples such as cancer, HPV virus (Hendawi et al. 2020), and COVID-19, the virus that is currently gathering the most attention (Liu et al. 2020; Khan et al. 2021).

To meet the ever-increasing clinical research and diagnosis demands, novel RNA detection assays are needed. Here, we present a single-molecule RNA chromogenic in situ hybridization assay for highly specific and sensitive RNA in situ detection in clinical samples. Our method is based on the proximity ligation assay principle to generate a DNA circle and rolling circle amplification (RCA) for signal enhancement. Hybridization of a pair of DNA probes in close proximity forms a “V” shape ligation template, and chromogenic in situ hybridization was used for the readout of individual RCA products (RCPs). Thus, we term our method vsmCISH. We established a complete and rigorous workflow suitable for formalin-fixed, paraffin-embedded (FFPE) samples, including external positive and negative controls. Notably, the results of vsmCISH can be directly observed with bright-field microscopy with chromogenic signals, similar to the IHC experimental procedure and staining pattern used in clinicopathology. Compared to our previously developed single-molecule chromogenic in situ hybridization (smCISH) technique (Jiang et al. 2019), vsmCISH has higher sensitivity and a faster detection procedure. We successfully applied our method for assessing the human epidermal growth factor receptor 2 (HER2) RNA status in invasive breast cancer and to investigate the utility of albumin (ALB) in the diagnosis of hepatocellular carcinoma. These applications show that our method has great potential to become a clinical diagnostic tool.

RESULTS

Workflow and specificity of the vsmCISH assay

The workflow of vsmCISH is illustrated in Figure 1A. First, three to five pairs of DNA probes are hybridized to the target sequences on the RNA molecules, forming a “V” shape structure. The DNA probes are therefore called “V probes.” After adequate and specific hybridization, circle probe precursors are hybridized with V probes, followed by DNA ligation using T4 DNA ligase to form a DNA circle. The 3′ end of the V probe hybridized with the DNA circle can then be used as an RCA primer to trigger the rolling circle amplification reaction mediated by Phi29 DNA polymerase. RCA products (RCPs) are formed under the above reaction conditions. The RCPs are then hybridized with horseradish peroxidase (HRP) labeled detection probes, followed by enzyme-catalyzed color development using 3,3′-Diaminobenzidine (DAB) as substrate. RCPs are stained brown and seen as discrete dots under bright-field microscopy to visualize the corresponding individual RNA molecules.

FIGURE 1.

FIGURE 1.

Open in a new tab

Implementation of vsmCISH. (A) The workflow of vsmCISH. (B) Detection of HPV-18 E6/E7 expression in HeLa (positive) and A549 (negative) cells. Detection of PGR expression in T47D (positive) and MCF-7 (negative) cells. Scale bar, 50 µm. (C) Statistics plot of the relative frequency of cells without target gene expression.

To investigate the specificity of vsmCISH, the assay was first applied to detect the expression levels of PGR and HPV Type 18 (HPV-18) E6/E7 mRNA in different cell lines (Fig. 1B). We detected PGR mRNA in two breast cancer cell lines, T47D and MCF-7, respectively. Based on the human protein atlas (HPA) database (www.proteinatlas.org), we know that MCF-7 (nTPM = 0) is a PGR-negative cell line and T47D (nTPM = 170.3) is a PGR-positive cell line. Similarly, we detected HPV Type 18 (HPV-18) E6/E7 mRNAs in HeLa and A549. HeLa is an HPV-18 positive cervical cancer cell line, and the lung cancer cell line A549 was used as a negative control (Lin et al. 2019). Massive signals with diffused staining patterns were observed in T47D and HeLa cells. Meanwhile, almost no target RNA signal was detected in the negative-expressing cell lines. In MCF-7 cells, the number of RCPs per cell (mean ± SD) detected was 0.0912 ± 0.2936 (n = 625). In A549 cells, the number of RCPs per cell detected was 0.2226 ± 0.4563 (n = 575). In addition, we counted the number of cells without target gene expression (RCPs/cell = 0) and found that these cells were up to 80% or more in both sets of results (Fig. 1C).

For the length of base pairing between V probes and circle probe precursors, we designed three sets of oligonucleotide probes for ESR1 detection in MCF-7. In the first group, V probe-L or R has nine bases pairing with circle probe precursor-L and eight bases pairing with circle probe precursor-S (Supplemental Fig. S1A). In the other two groups, the length of base pairing between V probe-L or R and circle probe precursor-L was extended to 13 (Supplemental Fig. S1B) or 18 bases (Supplemental Fig. S1C), between V probe-L or R, and circle probe precursor-S was also extended to 13 or 18 bases. Our experimental results showed that eight to nine base pairings between V probes and circle probe precursors could make a sufficient and stable hybridization under the present reaction conditions. As the hybridization length increased, the detected number of RCPs per cell (mean ± SD) were 19.7039 ± 12.1506 (n = 635), 16.0096 ± 7.9605 (n = 523), 13.9119 ± 6.8199 (n = 545), respectively (Supplemental Fig. S1D).

Compared to smCISH using padlock probes, vsmCISH shows better detection efficiency

To compare the performance of the V probe–based vsmCISH with the padlock probe–based smCISH assay, we detected ESR1 in MCF-7 and A549. According to the HPA database, A549 cells were used as a negative control. We selected five hybridization sites on the ESR1 mRNA to design V probes and padlock probes, respectively. Our experimental results showed that the vsmCISH assay has higher detection efficiency and maintains specificity. First, we compared five pairs of V probes with five padlock probes on the same target hybridization sites (Fig. 2A,C). In MCF-7 cells, the detected number of RCPs per cell (mean ± SD) was 13.2991 ± 8.5009 (n = 468) when using five pairs of V probes for vsmCSH, while it was 9.5282 ± 6.6826 (n = 496) by padlock probe-based smCISH assay. For the negative control, there were no statistically significant differences between the two assays in A549 cells (P > 0.05).

FIGURE 2.

FIGURE 2.

Open in a new tab

Evaluation of the detection efficiency of vsmCISH. vsmCISH shows better detection efficiency than smCISH either by using five hybridization sites (A) or one hybridization site (B). Scale bar, 50 µm. (C) Statistics plot of the detected RCPs per cell for ESR1 in MCF-7 and A549 cells. (****) P < 0.0001.

It has been previously demonstrated that padlock probes containing C or G at the donor site showed reduced detection efficiency compared to A or T (Liu et al. 2021). Therefore, to avoid the effects of different ligation rates of SplintR DNA ligase, we compared the detection efficiency of two assays only using ESR1-padlock probe 1 containing A and T at the donor site and one pair of V probes with the same hybridization site (Fig. 2B,C). The results show that vsmCISH still has better detection efficiency, even compared to the padlock probe with higher ligation efficiency. In MCF-7 cells, the number of RCPs per cell (mean ± SD) detected using one pair of V probes was 4.0951 ± 3.4648 (n = 368) and 1.8616 ± 1.8371 (n = 318) by padlock probe assay. For negative control A549 cells, there were no statistically significant differences between the two assays (P > 0.05). Thus, these results demonstrated that vsmCISH shows better detection efficiency than smCISH.

Compared to smCISH using the padlock probe, vsmCISH generates more visible RCPs

The size of RCPs reflects the RCA efficiency. Thus, we analyzed the areas of individual RCPs generated by vsmCISH and smCISH assays to investigate their RCA efficiency. HER2 mRNA in MCF-7 cells was detected using these two methods by performing an RCA reaction for 2 h. As shown in Figure 3, the median area of RCPs generated by padlock probe-based smCISH assay was 0.6864 µm2 (n = 1100), while RCPs generated by vsmCISH were bigger with a median size of 0.8757 µm2 (n = 1123). The average size of RCPs in vsmCISH is also significantly larger than that of smCISH (0.7296 ± 0.4340 µm2 vs. 0.8241 ± 0.4024 µm2, P < 0.0001). When observed by the naked eye, vsmCISH RCP dots also seem clearer and more visible than the original smCISH (Fig. 3A,B). Under this premise, higher RCA reaction efficiency made it possible to shorten the RCA reaction time to 2 h for vsmCISH, resulting in a faster protocol than smCISH.

FIGURE 3.

FIGURE 3.

Open in a new tab

Evaluation of RCA efficiency of vsmCISH. Compared to smCISH using the padlock probe (A), vsmCISH (B) generates more visible RCPs. Scale bar, 10 µm. (C) Statistics plot of RCP sizes by area of HER2 mRNA signal in MCF-7 cells. (****) P < 0.0001.

External controls applicable to the clinical detection protocol

For IHC staining, positive and negative controls must be established, and different staining intensities of the tissue as an external control is optimal. This can also apply to RNA ISH assays. To establish a complete and rigorous workflow suitable for clinical FFPE samples, we selected bacterial gene dapB from Bacillus subtilis strain SMY as a negative control and three housekeeping genes with different expression levels as positive controls, including PPIB, TPT1, and GAPDH. We tested these four genes in MCF-7 cells and normal breast tissue sections to verify if the results were in line with the trends of expression level data from the HPA database (https://www.proteinatlas.org). There was almost no dapB signal detected in either MCF-7 cells or the breast cancer tissue by vsmCISH (Fig. 4), indicating that dapB can be used as a negative control gene for clinical detection. In MCF-7 cells (Fig. 4A), we observed increasing expression levels from PPIB to TPT1 and TPT1 to GAPDH (PPIB nTPM = 1102.6, TPT1 nTPM = 3449.9, GAPDH nTPM = 11849.8). In normal breast tissue (Fig. 4B), we observed increased expression levels from PPIB to GAPDH and GAPDH to TPT1 (PPIB nTPM = 361.9, TPT1 nTPM = 4788.5, GAPDH nTPM = 817). Thus, the expression levels of these three housekeeping genes detected by vsmCISH in MCF-7 cells and normal breast tissues were all in line with the trends of the expression level database data.

FIGURE 4.

FIGURE 4.

Open in a new tab

Establishment of positive and negative controls. Detection of dapB, PPIB, TPT1, and GAPDH mRNA expression in MCF-7 cells (A) and FFPE normal breast tissue sections (B). Scale bar, 50 µm for part A and 150 µm for part B.

The utility of ALB as an RNA biomarker in hepatocellular carcinoma diagnosis

Albumin is a specific marker of the hepatocytes (Varma and Cohen 2004). First, we detected ALB mRNA in HepG2 (Fig. 5A) and SK-BR-3 (Fig. 5B) cells. HepG2 is an ALB-positive cell line isolated from hepatocellular carcinoma, and human breast adenocarcinoma SK-BR-3 cells were used as a negative control. Massive positive signals were detected in HepG2 cells, and almost no target RNA was detected in SK-BR-3 cells. Quantification was not performed because the RCPs detected in HepG2 were strongly diffusely stained and overcrowded. To further verify the specificity of ALB mRNA in situ detection using vsmCISH, we performed the assay in clinical FFPE samples of hepatocellular carcinoma and colorectal cancer tissue sections, respectively. As shown in Figure 5C,D, the hepatocellular carcinoma sample had an ALB-positive result, and the colorectal cancer sample had an ALB-negative result. Next, we conducted ALB mRNA in situ detection in the FFPE sample of colon cancer liver metastases (Fig. 5E). It was apparent that ALB was positive in normal hepatocytes, and ALB was negative in colon cancer cells adjacent to liver tissue. Overall, ALB manifested as an RNA biomarker can play a cueing role in differentiating primary liver cancer from metastatic liver cancer.

FIGURE 5.

FIGURE 5.

Open in a new tab

Detection of albumin mRNA expression by vsmCISH. ALB detection in HepG2 cells (A), SK-BR-3 cells (B), hepatocellular carcinoma cancer FFPE tissue sections (C), colorectal cancer FFPE tissue sections (D), and colon cancer liver metastases FFPE tissue sections (E). (F) H&E staining of the same colon cancer liver metastases specimen. Scale bar, 50 µm for parts A and B, 200 µm for parts C–F.

Accessing of HER2 mRNA expression by vsmCISH in invasive breast cancer

According to the National Comprehensive Cancer Network (NCCN) guidelines for breast cancer from version 3.2022, HER2 testing should be performed on all new primary or newly metastatic breast cancers using the methodology outlined in the American Society of Clinic Oncology/College of American Pathologists Clinical Practice (ASCO/CAP) HER2 testing guideline. Breast cancer specimens can be started with an IHC detection to assess its expression at the protein level. IHC 3+ was classified as HER2 positive, while IHC 0 and 1+ were classified as HER2 negative. Patients with IHC 2+ status require further gene amplification status by DNA FISH.

To investigate the potential clinical implications of HER2 RNA expression levels in breast tumors and their relationship with IHC and FISH, we performed in situ HER2 mRNA detection by vsmCISH in 30 cases of invasive breast cancer FFPE tissue sections and quantified HER2 mRNA expression in single-cell levels (at least 50 infiltrating cancer cells) (Fig. 6J). First, the detection was successfully performed in IHC scores 0, 1+, and 3+ samples. It is apparent that IHC 3+ samples had levels of HER2 mRNA higher than levels in IHC 0 and 1+ samples (Fig. 6A–C,I). To further explore the HER2 mRNA expression levels of the IHC 2+ samples, we selected 10 HER2 DNA amplification-negative cases and five HER2 DNA amplification-positive cases among them. The vsmCISH results show a noticeable variation at the RNA expression level, even though these samples were all HER2 IHC 2+ at the protein level (Fig. 6D–G).

FIGURE 6.

FIGURE 6.

Open in a new tab

Detection of HER2 mRNA expression in invasive breast cancer FFPE tissue sections. (A) HER2 IHC 0, RNA score 0. (B) HER2 IHC 0, RNA score 1. (C) HER2 IHC 1+, RNA score 0. (D) HER2 IHC 2+, DNA FISH negative, RNA score 0. (E) HER2 IHC 2+, DNA FISH negative, RNA score 1. (F) HER2 IHC 2+, DNA FISH negative, RNA score 2. (G) HER2 IHC 2+, DNA FISH positive, RNA score 3. (H) HER2 IHC 2+, DNA FISH positive, RNA score 4. (I) HER2 IHC 3+, RNA score 4. (J) Quantification of HER2 mRNA detection in 25 cases of invasive breast cancer FFPE tissue sections, excluding five cases with strong and diffuse expression. (K) The relationship between IHC classification and RNA score. Scale bar, 100 µm.

We also developed a five-tier scoring system for HER2 mRNA expression semiquantitative evaluation: score 0 was registered if the mean number of RCPs/cell was 0 to 2; score 1 was registered if the mean number was 3 to 5; score 2 was registered if the mean number was 6 to 10; score 3 was registered if the mean number was higher than 11; score 4 was registered for diffused strong positive. Score 0 or 1 was classified as HER2 vsmCISH negative, and score 3 or 4 was classified as HER2 vsmCISH positive. Score 2 was classified as HER2 vsmCISH equivocal status. Semiquantitative scoring of vsmCISH data is summarized in Table 1. IHC 0 and 1+ samples scored 0 or 1, and IHC 3+ samples scored 3 or 4. It can be seen more clearly that vsmCISH results of the above 15 cases analyzed by our scoring system were in good agreement with IHC results. Further analysis (Fig. 6K) showed that 80% (8/10) of cases of IHC 2+ FISH– were vsmCISH negative, and 80% (4/5) cases of IHC 2+ FISH+ were vsmCISH positive. Only 20% (3/15) of IHC 2+ were classified as HER2 equivocal status by vsmCISH assay. Compared to IHC, vsmCISH can reduce the number of HER2 equivocal cases and provide a more definitive HER2 gene status consistent with DNA FISH results.

TABLE 1.

Semiquantitative scoring system

graphic file with name 836tb01.jpg

Open in a new tab

DISCUSSION

In this study, we presented a robust RNA chromogenic in situ hybridization assay called vsmCISH for in situ RNA detection. Similar to smCISH, vsmCISH is also based on RCA for signal enhancement and chromogenic readout for single-molecule detection. In this method, DNA probes that hybridize to the target RNA sequences are “V probes” and the DNA circles formed templated by the V probes after ligation are not constrained by the RNA templates, different from that of the padlock probe-based smCISH assay. The experiments compared with the padlock probe approach indicated that vsmCISH had better RCA reaction efficiency and detection efficiency. We speculate that the presence of the V probe provides a suitable space, more conducive to rolling circle amplification for the circle probes. At the same time, the RNA templates may be hindering RCA in the smCISH assay because the padlock probes are directly ligated on the RNA templates that formed the DNA–RNA hybrid helix. Higher RCA reaction efficiency may also be the key to better performance in detection sensitivity due to more RCPs with detectable sizes being generated. Moreover, the oligonucleotide length of the V probe and circle probe precursors are all shorter than the padlock probe. Shorter probes may penetrate the cells more efficiently, allowing more targets to be found and detected. Notably, because of the higher RCA efficiency, the reaction time of RCA in vsmCISH has been shortened considerably to 2 h, a dramatic improvement over the conventional overnight reaction for smCISH. This makes vsmCISH a faster protocol that is advantageous for clinical diagnostics.

In probe design, in addition to the advantages of using the two-step production of RCA templates to improve RCA efficiency, vsmCISH is also based on the concept of proximity ligation assay to ensure the specificity (Fredriksson et al. 2002; Greenwood et al. 2015). Therefore, only when the V probes pairs hybridize to the target region simultaneously could the circle probe precursors have the complete templates for follow-up ligation reaction. This is different from the RollFISH method, where only one hybridization and one ligation event are required to generate a signal per probe (Wu et al. 2018). RollFISH probe consists of a set of smFISH probes (at least 4–5 oligodeoxynucleotides containing 30 nt sequence complementary to the RNA target) and padlock probes docked to smFISH probes. Once an oligodeoxynucleotide hybridizes to the wrong target region, a nonspecific RCA signal is also generated. Wu et al. obtained a significant positive correlation (Pearson's R2 = 0.97) between RollFISH HER2 transcript counts with available RNA-seq data for six cell lines. RNA-seq data in this paper showed that MCF-7 cells expressed a higher level of HER2 than A549 cells. However, RollFISH detected more signals in A549 cells. The expression trends of HER2 mRNA in MCF-7 and A549 from vsmCISH detection results were consistent with the above RNA-seq data (Supplemental Fig. S2). The detected number of RCPs per cell (mean ± SD) was 24.2382 ± 16.3556 (n = 424) in MCF-7 and 12.1968 ± 5.9951 (n = 249) in A549 (Supplemental Fig. S2C). Therefore, we infer that vsmCISH is superior to RollFISH in specificity.

To evaluate the potential utility of our vsmCISH for clinical diagnostics, we applied vsmCISH to quantify and locate HER2 mRNA single molecule in invasive breast cancer and albumin mRNA in liver cancer FFPE tissue sections. Accurate detection and evaluation of Her2 protein expression and gene amplification status in breast cancer are crucial for clinical treatment and prognosis. Patients with HER2-positive breast cancer are highly sensitive to anti-HER2-targeted therapy, which has significantly improved clinical outcomes (Ryan et al. 2008; Baselga et al. 2012; Koleva-Kolarova et al. 2017; Loibl and Gianni 2017; Howie et al. 2019; von Minckwitz et al. 2019; Rugo et al. 2021). According to ASCO/CAP HER2 testing guidelines, HER2 IHC 3+ or IHC 2+ with DNA FISH-positive cases are defined as Her2-positive. In other words, the conventional assay determines the status of the HER2 oncogene at the protein and DNA levels. In addition, recent clinical trials have shown that a novel HER2-targeted antibody-drug conjugate (ADC), trastuzumab deruxtecan (T-DXd), is effective for HER2 low breast cancer patients (Modi et al. 2022). Consequently, accurate quantification and localization of HER2 oncogene expression at the RNA level may have a role in guiding HER2-targeted therapies for breast cancer, especially HER2-low breast cancer. Compared to IHC, vsmCISH can reduce the number of HER2 equivocal cases and provide a more definitive HER2 gene status consistent with DNA FISH results. Regarding HER2-low breast cancer, vsmCISH has demonstrated great potential in further stratification of the lower range of HER2 expression. However, further studies with a larger sample size must replicate our findings. A better cut-off threshold for each stratification could be found from more data. Accurate image analysis of larger areas is another critical issue to be addressed, including the identification and segmentation of target cells and segmentation of clustered signals for generating a convincing gene expression profile of single cells. Thus, HER2 heterogeneous expression data further analyzed from sing-cell data may provide more information for clinical diagnosis and treatments of breast cancer.

Albumin synthesized and secreted by hepatocytes is a more specific marker of the hepatocytes (Varma and Cohen 2004). However, the detection of ALB protein expression in FFPE samples by IHC has limitations, such as background overstaining and false-positive results (Murray et al. 1992). To overcome this problem, mRNA of albumin can be detected instead because they remain in the hepatocytes. We thus applied vsmCISH for ALB mRNA detection and found that they are specifically detected in hepatocytes. We also demonstrated that the ALB mRNA detected by vsmCISH can differentiate primary liver cancer from metastatic cancer. This result again showed the great value of vsmCISH applications in clinical diagnosis. Detecting secreted proteins at their RNA level can be used as an alternative way to evaluate their gene expression level, making it possible to use genes that produce secreted proteins as in situ biomarkers. It can also be expanded to detect noncoding RNAs, offering novel biomarker discovery and translations opportunities.

To sum up, the technical advantages and the successful use for HER2 and ALB detection as exemplified in clinical applications demonstrate that vsmCISH can become a useful molecular diagnostics tool for RNA biomarker in situ detection.

MATERIALS AND METHODS

Cell culture and sample preparation

MCF-7 (ATCC), T47D (Stem Cell Bank of Chinese Academy of Sciences), HepG2 (Stem Cell Bank of Chinese Academy of Sciences), and A549 (ATCC) cells were cultured in high-glucose DMEM (Shanghai Basalmedia Technologies) supplemented with 10% fetal bovine serum (FBS, cat. no. 10099-141C; Gibco). SK-BR-3 (ATCC) cells were cultured in RPMI 1640 (Shanghai Basalmedia Technologies) supplemented with 10% FBS. HeLa (Stem Cell Bank of Chinese Academy of Sciences) cells were cultured in MEM (containing NEAA, cat. no. PM150410; Procell) supplemented with 10% FBS. Cells were cultured in a humid incubator with 5% CO2 at 37°C. When the number of adherent cells was sufficient, cells were detached into cell suspension by treatment with 0.25% trypsin-EDTA solution (cat. no. S330JV; Shanghai Basalmedia Technologies). The cells were then seeded on adhesion microscope slides (cat. no. 188105; Citotest) and placed in the EasYDish (cat. no. 150468; Nunc, Thermo Scientific) containing culture medium. Cells were grown on the glass slides under the same culturing condition before the cell density reached ∼80%. Slides were washed twice for 3 min each with diethyl pyrocarbonate (cat. no. D5758; Sigma-Aldrich) treated 1× phosphate-buffered saline (1× DEPC-PBS) after removal of culture medium in the EasYDish. Next, fixation was performed by incubating slides with 4% paraformaldehyde (PFA) (cat. no. 16005; Sigma-Aldrich) in 1× DEPC-PBS for 30 min at room temperature. The slides were washed twice for 3 min each with 1× DEPC-PBS and dehydrated with an ethanol series of 70%, 85%, and absolute for 5 min each. After air drying at room temperature, slides can be stored at −80°C until further use.

Pretreatment steps of FFPE tissue sections

Formalin-fixed, paraffin-embedded (FFPE) tissue sections (4 µm thick) of human invasive breast cancer were obtained from the Department of Pathology, Xiamen Maternal and Child Health Hospital. Human hepatocellular carcinoma, colorectal cancer, and colon cancer liver metastases FFPE tissue sections were obtained from the Department of Pathology, The 910th Hospital. All the use of clinical samples was approved by the ethics committee of the School of Medicine, Huaqiao University (M2021005). FFPE tissue sections were first baked at 60°C for 30 min and dewaxed in xylene twice for 15 min and 10 min, respectively. The rehydration process was then completed by submerging in absolute ethanol twice, 95% ethanol, and 70% ethanol twice for 2 min each time, followed by dipping in DEPC treated H2O for 5 min and washing with 1×DEPC-PBS for 2 min. The tissue sections were fixed by 4% PFA in 1× DEPC-PBS for 10 min at room temperature and washed with 1× DEPC-PBS for 2 min. Next, a solution of 0.1 M HCl containing 0.1 mg/mL pepsin (cat. no. P7012; Sigma-Aldrich) was applied to the tissue sections for 30 min at 37°C. After washing with DEPC-H2O for 5 min and DEPC-PBS for 2 min, dehydration was performed with an ethanol series of 70%, 85%, and ethanol absolute for 1 min each.

V probe hybridization

ImmEdge Hydrophobic Barrier Pen (cat. no. H-4000; Vector) was first used to draw a circle around the sample area to better hold reaction liquid on the tissue area during subsequent steps. The sample was then washed three times with 0.1%(v/v) Tween-20 (cat. no. P9416; Sigma-Aldrich) in 1× DEPC-PBS (DEPC-PBST). Cell samples were permeabilized with 0.1 M HCl for 5 min before proceeding to the following steps. To block endogenous peroxidase, enhanced endogenous peroxidase blocking buffer (cat. no. P0100B; Beyotime) was applied to the sample for 10 min, followed by washing with DEPC-PBST three times for 3 min each. Next, 100 nM V probes (or padlock probes) in a hybridization buffer containing 15% formamide (cat. no. 47671; Sigma-Aldrich) and 6× SSC (cat. no. S6639; Sigma-Aldrich) were added to the sample. The sample was incubated at 37°C for two h, followed by three washes for 3 min each with DEPC-PBST and three washes for 5 min each with washing buffer containing 20% formamide and 2× SSC.

Circle probe precursors’ hybridization and ligation

The hybridization reaction solution for this step consisted of 100 nM of each circle probe precursors, 1× T4 DNA ligase buffer, 0.2% tween-20, 250 mM NaCl (cat. no. S3014; Sigma-Aldrich), 2 mg/mL BSA (cat. no. B600036; Sangon) and 1 U/µL RiboLock RNase inhibitor (cat. no. EO0381; Thermo Scientific). After three washes with DEPC-PBST, the sample was incubated with the above hybridization reaction solution at 37°C for 1 h, followed by three washes for 3 min each with DEPC-PBST. A ligation mix containing 1× T4 DNA ligase buffer, 0.2% tween-20, 250 mM NaCl, 2 mg/mL BSA, 1 U/µL RiboLock RNase inhibitor, 1 mM ATP (cat. no. R0441; Thermo Scientific) and 0.05 Weiss U/µL T4 DNA ligase (cat. no. EL0011; Thermo Scientific) was then added to the sample, incubating at 37°C for 0.5 h. After probe circularization, three washes for 3 min each were performed with DEPC-PBST.

RCA and RCA products’ detection

An RCA reaction mix consisted of 1× EquiPhi29 DNA Polymerase Reaction Buffer, 5% (v/v) glycerol (cat. no. G9012; Sigma-Aldrich), 2 mg/mL BSA, 1 mM dNTPs (cat. no. R0182; Thermo Scientific), 1 U/µL EquiPhi29 DNA Polymerase (cat. no. A39391; Thermo Scientific) and 1 mM DTT. The incubation was carried out at 42°C for 2 h, followed by three washes for 3 min each with DEPC-PBST. A detection mix containing 100 nM HRP labeled detection probes, 20% formamide, and 2× SSC was added to the sample and incubated at 37°C for 0.5 h.

Staining and mounting

Before DAB staining, the sample was washed three times for 3 min each with DEPC-PBST. RCPs staining was performed with DAB Immunohistochemistry Color Development Kit (cat. no. E670033; BBI Life Sciences) for 8 min according to the manufacturer's manual. Hematoxylin (cat. no. H3136; Sigma-Aldrich) was used for nuclear counterstaining. After air drying, the slides were mounted with neutral balsam mounting medium (cat. no. E675007; BBI Life Sciences).

Image acquisition and analysis

Images were taken using a Leica DM6B microscope equipped with a DFC7000T camera. We used ImageJ Fiji software to split the DAB staining channel for color deconvolution. Then, the image after color deconvolution was analyzed with the user-friendly interactive machine learning-based imaging analysis software Ilastik (https://www.ilastik.org) to segment clustered signals and identify signals. Next, CellProfiler (https://www.cellprofiler.org/) was used to identify the nucleus, segment cell regions, and identify the relation of signals with their host cells. After the above image processing steps, we finally got the single-cell data. In addition, the area of RCPs was measured by the Image Pro Plus software. Statistical analysis and plotting were performed using OriginPro software (https://www.originlab.com/).

Padlock probe assay

The experimental protocol of padlock probe ligation and RCA primer hybridization was carried out as described previously in the smCISH (Jiang et al. 2019, 2020). Padlock probe hybridization, RCA, RCA product detection, staining, and mounting remain the same as above.

SUPPLEMENTAL MATERIAL

Supplemental material is available for this article.

Supplementary Material

Supplemental Material
supp_29_6_836__DC1.html (741B, html)

ACKNOWLEDGMENTS

This study was supported by the Science Foundation of Fujian Province (2022J06022), the Quanzhou Science and Technology Plan Project (2021C040R), and the Scientific Research Funds of Huaqiao University.

Footnotes

REFERENCES

  1. Anderson CM, Zhang B, Miller M, Butko E, Wu X, Laver T, Kernag C, Kim J, Luo Y, Lamparski H, et al. 2016. Fully automated RNAscope in situ hybridization assays for formalin-fixed paraffin-embedded cells and tissues. J Cell Biochem 117: 2201–2208. 10.1002/jcb.25606 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Baselga J, Cortes J, Kim SB, Im SA, Hegg R, Im YH, Roman L, Pedrini JL, Pienkowski T, Knott A, et al. 2012. Pertuzumab plus trastuzumab plus docetaxel for metastatic breast cancer. N Engl J Med 366: 109–119. 10.1056/NEJMoa1113216 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chu YH, Hardin H, Zhang R, Guo Z, Lloyd RV. 2019. In situ hybridization: introduction to techniques, applications and pitfalls in the performance and interpretation of assays. Semin Diagn Pathol 36: 336–341. 10.1053/j.semdp.2019.06.004 [DOI] [PubMed] [Google Scholar]
  4. Fernandez-Barral A, Costales-Carrera A, Buira SP, Jung P, Ferrer-Mayorga G, Larriba MJ, Bustamante-Madrid P, Dominguez O, Real FX, Guerra-Pastrian L, et al. 2020. Vitamin D differentially regulates colon stem cells in patient-derived normal and tumor organoids. FEBS J 287: 53–72. 10.1111/febs.14998 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Fredriksson S, Gullberg M, Jarvius J, Olsson C, Pietras K, Gustafsdottir SM, Ostman A, Landegren U. 2002. Protein detection using proximity-dependent DNA ligation assays. Nat Biotechnol 20: 473–477. 10.1038/nbt0502-473 [DOI] [PubMed] [Google Scholar]
  6. Froeling FEM, Casolino R, Pea A, Biankin AV, Chang DK. 2021. Molecular subtyping and precision medicine for pancreatic cancer. J Clin Med 10: 149. 10.3390/jcm10010149 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Greenwood C, Ruff D, Kirvell S, Johnson G, Dhillon HS, Bustin SA. 2015. Proximity assays for sensitive quantification of proteins. Biomol Detect Quantif 4: 10–16. 10.1016/j.bdq.2015.04.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. He J, Mo D, Chen J, Luo L. 2020. Combined whole-mount fluorescence in situ hybridization and antibody staining in zebrafish embryos and larvae. Nat Protoc 15: 3361–3379. 10.1038/s41596-020-0376-7 [DOI] [PubMed] [Google Scholar]
  9. Hendawi N, Niklander S, Allsobrook O, Khurram SA, Bolt R, Doorbar J, Speight PM, Hunter KD. 2020. Human papillomavirus (HPV) can establish productive infection in dysplastic oral mucosa, but HPV status is poorly predicted by histological features and p16 expression. Histopathology 76: 592–602. 10.1111/his.14019 [DOI] [PubMed] [Google Scholar]
  10. Howie LJ, Scher NS, Amiri-Kordestani L, Zhang L, King-Kallimanis BL, Choudhry Y, Schroeder J, Goldberg KB, Kluetz PG, Ibrahim A, et al. 2019. FDA approval summary: pertuzumab for adjuvant treatment of HER2-positive early breast cancer. Clin Cancer Res 25: 2949–2955. 10.1158/1078-0432.CCR-18-3003 [DOI] [PubMed] [Google Scholar]
  11. Jiang M, Liu L, Hong C, Chen D, Yao X, Chen X, Lin C, Ke R. 2019. Single molecule chromogenic in situ hybridization assay for RNA visualization in fixed cells and tissues. RNA 25: 1038–1046. 10.1261/rna.070599.119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Jiang M, Lin C, Ke R. 2020. Detection of individual RNA in fixed cells and tissues by chromogenic ISH. Bio Protoc 10: e3510. 10.21769/BioProtoc.3510 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kameneva P, Artemov AV, Kastriti ME, Faure L, Olsen TK, Otte J, Erickson A, Semsch B, Andersson ER, Ratz M, et al. 2021. Single-cell transcriptomics of human embryos identifies multiple sympathoblast lineages with potential implications for neuroblastoma origin. Nat Genet 53: 694–706. 10.1038/s41588-021-00818-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kersigo J, Pan N, Lederman JD, Chatterjee S, Abel T, Pavlinkova G, Silos-Santiago I, Fritzsch B. 2018. A RNAscope whole mount approach that can be combined with immunofluorescence to quantify differential distribution of mRNA. Cell Tissue Res 374: 251–262. 10.1007/s00441-018-2864-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Khan M, Yoo SJ, Clijsters M, Backaert W, Vanstapel A, Speleman K, Lietaer C, Choi S, Hether TD, Marcelis L, et al. 2021. Visualizing in deceased COVID-19 patients how SARS-CoV-2 attacks the respiratory and olfactory mucosae but spares the olfactory bulb. Cell 184: 5932–5949.e5915. 10.1016/j.cell.2021.10.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Koleva-Kolarova RG, Oktora MP, Robijn AL, Greuter MJW, Reyners AKL, Buskens E, de Bock GH. 2017. Increased life expectancy as a result of non-hormonal targeted therapies for HER2 or hormone receptor positive metastatic breast cancer: a systematic review and meta-analysis. Cancer Treat Rev 55: 16–25. 10.1016/j.ctrv.2017.01.001 [DOI] [PubMed] [Google Scholar]
  17. Kwa M, Makris A, Esteva FJ. 2017. Clinical utility of gene-expression signatures in early stage breast cancer. Nat Rev Clin Oncol 14: 595–610. 10.1038/nrclinonc.2017.74 [DOI] [PubMed] [Google Scholar]
  18. Li DQ, Kim S, Li JM, Gao Q, Choi J, Bian F, Hu J, Zhang Y, Li J, Lu R, et al. 2021. Single-cell transcriptomics identifies limbal stem cell population and cell types mapping its differentiation trajectory in limbal basal epithelium of human cornea. Ocul Surf 20: 20–32. 10.1016/j.jtos.2020.12.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lin J, Chen L, Jiang W, Zhang H, Shi Y, Cai W. 2019. Rapid detection of low-level HeLa cell contamination in cell culture using nested PCR. J Cell Mol Med 23: 227–236. 10.1111/jcmm.13923 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Liu J, Babka AM, Kearney BJ, Radoshitzky SR, Kuhn JH, Zeng X. 2020. Molecular detection of SARS-CoV-2 in formalin-fixed, paraffin-embedded specimens. JCI Insight 5: e139042. 10.1172/jci.insight.139042 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Liu S, Punthambaker S, Iyer EPR, Ferrante T, Goodwin D, Furth D, Pawlowski AC, Jindal K, Tam JM, Mifflin L, et al. 2021. Barcoded oligonucleotides ligated on RNA amplified for multiplexed and parallel in situ analyses. Nucleic Acids Res 49: e58. 10.1093/nar/gkab120 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Loibl S, Gianni L. 2017. HER2-positive breast cancer. Lancet 389: 2415–2429. 10.1016/S0140-6736(16)32417-5 [DOI] [PubMed] [Google Scholar]
  23. Magaki S, Hojat SA, Wei B, So A, Yong WH. 2019. An introduction to the performance of immunohistochemistry. Methods Mol Biol 1897: 289–298. 10.1007/978-1-4939-8935-5_25 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Modi S, Jacot W, Yamashita T, Sohn J, Vidal M, Tokunaga E, Tsurutani J, Ueno NT, Prat A, Chae YS, et al. 2022. Trastuzumab deruxtecan in previously treated HER2-low advanced breast cancer. N Engl J Med 387: 9–20. 10.1056/NEJMoa2203690 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Murray GI, Paterson PJ, Ewen SW, Melvin WT. 1992. In situ hybridisation of albumin mRNA in normal liver and hepatocellular carcinoma with a digoxigenin labelled oligonucleotide probe. J Clin Pathol 45: 21–24. 10.1136/jcp.45.1.21 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Player AN, Shen LP, Kenny D, Antao VP, Kolberg JA. 2001. Single-copy gene detection using branched DNA (bDNA) in situ hybridization. J Histochem Cytochem 49: 603–612. 10.1177/002215540104900507 [DOI] [PubMed] [Google Scholar]
  27. Rugo HS, Im SA, Cardoso F, Cortes J, Curigliano G, Musolino A, Pegram MD, Wright GS, Saura C, Escriva-de-Romani S, et al. 2021. Efficacy of margetuximab vs trastuzumab in patients with pretreated ERBB2-positive advanced breast cancer: a phase 3 randomized clinical trial. JAMA Oncol 7: 573–584. 10.1001/jamaoncol.2020.7932 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ryan Q, Ibrahim A, Cohen MH, Johnson J, Ko CW, Sridhara R, Justice R, Pazdur R. 2008. FDA drug approval summary: lapatinib in combination with capecitabine for previously treated metastatic breast cancer that overexpresses HER-2. Oncologist 13: 1114–1119. 10.1634/theoncologist.2008-0816 [DOI] [PubMed] [Google Scholar]
  29. Shiers S, Klein RM, Price TJ. 2020. Quantitative differences in neuronal subpopulations between mouse and human dorsal root ganglia demonstrated with RNAscope in situ hybridization. Pain 161: 2410–2424. 10.1097/j.pain.0000000000001973 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Varma V, Cohen C. 2004. Immunohistochemical and molecular markers in the diagnosis of hepatocellular carcinoma. Adv Anat Pathol 11: 239–249. 10.1097/01.pap.0000131822.31576.9d [DOI] [PubMed] [Google Scholar]
  31. von Minckwitz G, Huang CS, Mano MS, Loibl S, Mamounas EP, Untch M, Wolmark N, Rastogi P, Schneeweiss A, Redondo A, et al. 2019. Trastuzumab emtansine for residual invasive HER2-positive breast cancer. N Engl J Med 380: 617–628. 10.1056/NEJMoa1814017 [DOI] [PubMed] [Google Scholar]
  32. Wang F, Flanagan J, Su N, Wang LC, Bui S, Nielson A, Wu X, Vo HT, Ma XJ, Luo Y. 2012. RNAscope: a novel in situ RNA analysis platform for formalin-fixed, paraffin-embedded tissues. J Mol Diagn 14: 22–29. 10.1016/j.jmoldx.2011.08.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Warford A. 2016. In situ hybridisation: technologies and their application to understanding disease. Prog Histochem Cytochem 50: 37–48. 10.1016/j.proghi.2015.12.001 [DOI] [PubMed] [Google Scholar]
  34. Wu C, Simonetti M, Rossell C, Mignardi M, Mirzazadeh R, Annaratone L, Marchio C, Sapino A, Bienko M, Crosetto N, et al. 2018. RollFISH achieves robust quantification of single-molecule RNA biomarkers in paraffin-embedded tumor tissue samples. Commun Biol 1:209. 10.1038/s42003-018-0218-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Wu Y, Liu Z, Xu X. 2020. Molecular subtyping of hepatocellular carcinoma: a step toward precision medicine. Cancer Commun (Lond) 40: 681–693. 10.1002/cac2.12115 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Material
supp_29_6_836__DC1.html (741B, html)
supp_079482.122_Supplemental_Material.pdf (506KB, pdf)

Articles from RNA are provided here courtesy of The RNA Society

RESOURCES