Zebrafish
Zebrafish (Danio rerio) was raised and maintained at 28.5 °C in water system under Zebrafish technology platform of CAS Center for Excellence in Molecular Cell Science. This study was approved by the Ethical Review Committee of CAS Center for Excellence in Molecular Cell Science, Chinese Academy of Sciences (CAS), China. Zebrafish were staged as previously described [50]. The wild type AB strain was used in this study. For each set of experiment, about twenty couples of males and females (AB, 5–18 months old) were randomly selected and crossed to generate embryos. The embryos collected for microinjections came from random parents mating, and 100–1000 embryos were injected for each set of experiment.
Plasmid construction
To construct the plasmid expressing GCN4 RNA element for CRISPR-dCas13 system targeting in zebrafish, the zebrafish β-actin promoter (cloned from pTol2-zbactin-E2A-mCherry plasmid donated from Weijun Pan Lab, Shanghai Institute of Nutrition and Health, CAS) and the 48× GCN4 sequence were inserted into pcDNA3+ vector, using the T4 DNA ligase (NEB, Cat. No. M0202S). Other plasmids were constructed using Hieff Clone One Step Cloning Kit (Yeasen, Cat. No. 10905ES25) according to the manufacturer’s protocol.
To construct plasmids for 6× His-tagged dCas13 proteins expression from E. coli, dCas13 ORFs were individually cloned into pET28a+ vector, fused with fluorescent proteins (FP) (EGFP, mScarlet or sfGFP), and a SV40 nuclear localization signal (NLS) in the N-terminus as well as a Nucleoplasmin NLS in the C-terminus was also included in the vector (abbreviated as dCas13-FP). Of note, dPspCas13b-3× sfGFP was fused with SV40 NLS in the N-terminus and NLSs from SV40 and Nucleoplasmin in the C-terminus.
To construct plasmids expressing mRNAs for in vitro transcription (IVT), the coding sequences of pom121-mScarlet and SV40NLS-dPspCas13b/-dRfxCas13d-EGFP-Nucleoplasmin NLS, as well as cDNA of ddx19a, gle1, nxf1, alyref, ddx39b, thoc2, eny2 and pcid2 genes were cloned into the pCS2+ vector containing the SP6 promoter and the SV40 polyadenylation signal, respectively.
To construct plasmid of Fibrillarin (Fbl)-mCherry, the coding sequence was cloned into pEGFP-C1vector.
All oligos used for plasmid constructions are listed in Additional file 2: Table S1.
dCas13-FP (fluorescent protein) expression and purification
A number of dCas13-FP plasmids, including dPspCas13b-EGFP, dPspCas13b-3× sfGFP, dBba2Cas13b-EGFP, dPba3Cas13b-EGFP, dHgm4Cas13b-EGFP, dHgm6Cas13b-EGFP, dMisCas13b-EGFP, dPguCas13b-EGFP, dRfxCas13d-EGFP, and dRfxCas13d-mScarlet, were individually transformed into the E. coli expression strain, the Transetta (DE3) chemically competent cells (Transgen Biotech, Cat. No. CD801), according to the manufacturer’s protocol. After transformation, the cells were cultivated at 37°C, 250 rpm for 2 h, followed by transferring into 1 L of LB culture media for further growing at the same condition. Once the absorbance of the culture media reached the OD600 around 0.6–0.8, isopropyl β-D-1-thiogalactopyranoside (IPTG) was added to the final concentration of 0.5 mM (GoldBio, Cat. No. I2481C50) to induce protein expression, and cultured at 16°C, 180 rpm for another 18 h.
The next day, cell pellets were collected by centrifugation (5000g, 10 min, 4°C), and resuspended in 25 mL lysis buffer (50 mM HEPES, 500 mM NaCl, 2 mM MgCl2, 50 mM Imidazole, 1 mM DTT, 0.5 mM Phenylmethylsulfonyl fluoride (PMSF)). Then, the resuspension was sonicated at 4°C by high-pressure homogenizer (Ultrahigh pressure cell crusher UH-06; Union-biotech) followed by centrifugation at 10,000 rpm for 45 min at 4°C. After that, the supernatant cell lysates were collected and sterile-filtered through a 0.22-μm polyvinylidene difluoride membrane (Millipore, Cat. No. GSWP04700). The supernatant was then incubated for 10 min with 1 mL Ni-NTA beads in the column (referred to as 1 column volume, Ni Sepharose 6 Fast Flow, GE healthcare, Cat. No. 17-5318-01) and then flowed through. Next, the Ni-NTA beads were washed twice with 10 column volumes of the lysis buffer and the bound proteins were eluted with 10 column volumes of the elution buffer (50 mM HEPES, 500 mM NaCl, 300 mM Imidazole, 0.01% v/v Triton X-100, 10% glycerol, 1 mM DTT, 0.5 mM PMSF). Proteins were then concentrated using a Amicon® Ultra-15 Centrifugal Filter (50K, Millipore, Cat. No. UFC905008) by centrifugation at 4000g at 4°C and were sterile-filtered before purification by Akta Pure FPLC (GE healthcare). The proteins were further purified through a 5-mL HiLoad Superdex 200 PG of gel filtration chromatography column, which was first equilibrated with storage buffer (50 mM Tris-HCl, 500 mM NaCl, 10% glycerol, 2 mM DTT, pH 7.5). Protein-containing fractions were collected (Additional file 1: Fig. S1c, d) and concentrated, followed by quantification with serial dilutions of standard BSA (2, 1, 0.5, 0.25, and 0.125 μg) by SDS-PAGE gel using Coomassie Brilliant blue staining (Additional file 1: Fig. S1e). Finally, proteins were snap-frozen in liquid nitrogen and stored in aliquots at −80°C.
RNA synthesis and purification
For the chemically modified gRNAs, including 5′ end 2′-O-methyl 3′ phosphorothioate (MS) modified and/or cyanine 3 (Cy3) labeled, and 3′ end Cy3 labeled gRNAs for dPspCas13b targeting, as well as the 3′ end MS-modified gRNAs for dRfxCas13d targeting, were synthesized at GenScript Company.
For gRNAs produced by IVT, the DNA templates for gRNAs were amplified by primers, followed by agarose gel purification. The gRNA sequences containing the T7 promoter (GAAATTAATACGACTCACTATA) are listed in Additional file 2: Table S1. The IVTs were done with T7 polymerase (Promega, Cat. No. P1300) to produce gRNAs. After that, the gRNAs were purified from denatured PAGE gel.
For mRNAs produced by IVT, DNA templates were digested from constructed pCS2-coding sequence plasmids with NEB restriction endonuclease, either NotI or KpnI, respectively, followed by purification with StarPrep Gel Extraction Kit StarPrep (GenStar, Cat. No. D205-04). mRNAs were then transcribed and purified in vitro by mMESSAGE mMACHINE™ SP6 kit (Thermo Scientific, Cat. No. AM1340) according to the manufacturer’s protocol.
Construction of dCas13-EGFP/gRNA complexes in vitro
To make the in vitro assembled dCas13-EGFP/gRNA complex, we first diluted the IVT or chemically modified gRNAs into 2 μL annealing buffer (10 mM Na-HEPES pH 7.4, 30 mM KCl, and 1.5 mM MgCl2) and annealed the gRNAs by heating at 75°C for 5 min, and slowly cooling down to room temperature at a rate of −0.1 °C/s. Then, each dCas13-EGFP protein was mixed with the annealed gRNAs at the molar ratio of 1:1.5 or 1:3 in the assembly buffer (20 mM Na-HEPES, pH 7.0, 200 mM KCl, 1 mM TCEP) to the final volume of 4 μL, respectively, and incubated at 37°C for 20 min for assembling. The assembled dCas13-EGFP/gRNA complexes were then checked by electrophoretic mobility shift assay (EMSA). Briefly, the 1 μL assembled complexes were loaded onto a native PAGE gel, consisting of 6% acrylamide at the top half and 12% acrylamide at the bottom half, and run the gel in 0.5× TBE buffer at 20 mA for 50 min at 4°C. Then, the proteins of dCas13-EGFP/gRNA complexes were detected by Coomassie Brilliant blue staining and imaging, and the gRNAs were detected by ethidium bromide (EB) staining and imaging (Additional file 1: Fig. S1f, g). The assembled complexes were made to appropriate concentrations for zebrafish zygotic microinjection in different experiments.
To make the dCas13-EGFP/gRNA mixture in vitro without assembly, the modified gRNAs were firstly added into 2 μL annealing buffer, and then mixed with the dCas13-EGFP protein in the assembly buffer to the final volume of 4 μL at room temperature. The final concentrations of the gRNA and dCas13-EGFP in the mixtures for different experiments were shown below in the next section.
The assembled complexes or mixtures were placed on ice prior to microinjection.
Microinjection of zebrafish embryos
We prepared each corresponding sample detailed below and injected ~1 nL sample into the 1-cell of each embryo at the 1-cell stage. Of note, the concentrations of the CRISPR-dCas13 systems, plasmids, and mRNAs used for microinjection were optimized and had no obvious effect for embryo development.
To screen a panel of dCas13-EGFP proteins for 48× GCN4 RNA labeling (Additional file 1: Fig. S1e), the pre-assembled 5.6 μM dCas13-EGFP/8.4 μM gRNA complexes were used. We either co-microinjected the β-actin-48× GCN4 plasmid together with the dCas13-EGFP/gRNA complexes or the corresponding dCas13-EGFP protein alone to target 48× GCN4 RNA. For example, 50 pg β-actin-48× GCN4 plasmid together with 0.9 ng dPspCas13b-EGFP/160 pg gRNA complex or 0.9 ng dPspCas13b-EGFP protein was injected into each embryo, respectively.
For dCas13-EGFP proteins, including dBba2Cas13b, dPba3Cas13b, dHgm4Cas13b, dHgm6Cas13b, and dRfxCas13d, we could hardly detect the EGFP signal in the nucleus post the microinjection of either the complex or the protein alone at the indicated concentration above after 6 h post fertilization (hpf) by widefield microscopy imaging. Tested high concentrations of these pre-assembled dCas13-EGFP/gRNA complexes or the individual dCas13-EGFP proteins alone (i.e., 22.0 μM dPba3Cas13b-EGFP) in the microinjection experiments still yielded poor signals. Of note, for dHgm6Cas13b-EGFP, which was difficult to dissolve and obtain high concentration, 4.5 μM protein/8.4 μM gRNA pre-assembled complex was used in this study.
To label endogenous mRNAs, we microinjected 0.9–1.5 ng dPspCas13b protein/53-160 pg modified gRNA for the CRISPR-dPspCas13b-mediated RNA labeling, and 4–7 ng dRfxCas13d protein/55–160 pg modified gRNA for CRISPR-dRfxCas13d-mediated RNA labeling, either the pre-assembled complex or the mixture. Of note, for the CRISPR-dRfxCas13d system, we found that the low concentration of the protein needed to assemble with high concentration of the modified gRNA to yield the better visualization signal. In our hands, the assembled 0.9 ng dPspCas13b-EGFP/160 pg modified gRNA, or 7 ng dRfxCas13d-EGFP/55 pg modified gRNA were injected into the one-cell stage embryo that yielded reliable SNR after 15 cell cycles in developing embryos.
To label endogenous muc5.1 and 100537515 mRNAs, respectively, 0.9 ng dPspCas13b-EGFP/160 pg MS-gRNA mixture was injected into each embryo. To achieve two different endogenous RNAs with dual-color, 0.9 ng dPspCas13b-EGFP/160 pg MS-gRNA and 7 ng dRfxCas13d-mScarlet/160 pg gRNA-MS mixture were co-microinjected into each embryo.
To track endogenous eppk1 and 100537515 transcriptions, as well as 100537515 mRNP motions, we microinjected 1.5 ng dPspCas13b-FPs/160 pg modified gRNA mixture to each embryo.
To express transport factors, 50 pg alyref, 200 pg ddx19a, 200 pg ddx39b, 200 pg eny2, 200 pg gle1, 200 pg pcid2, 50 pg nxf1 or 200 pg thoc2 mRNAs produced by IVT were individually injected into each embryo together with the CRISPR-dPspCas13b system, respectively. Of notes, embryos were developed normally by microinjecting these mRNAs under these tested concentrations. To visualize NPCs, 100 pg pom121-mScarlet mRNA together with CRISPR-dPspCas13b system were injected into each embryo. For dCas13-EGFP mRNA injection, 250 pg dPspCas13b-EGFP or 250 pg dRfxCas13d-EGFP mRNA was injected into each embryo. Detailed concentrations of other samples were indicated in the figure legends and methods.
Widefield microscopy imaging
We checked and collected embryos with relatively uniform fluorescence intensity in examined embryos under stereomicroscope (Nikon SMZ18) before imaging on DeltaVision. Then, live embryos were dechorionated with 1-mL syringe at corresponding developmental stages or fixed embryos after performing smFISH (detailly described below) were mounted on the bottom of the dish (Cellvis, Cat. No. 35-10-1.5-N) with 1% low melting agarose (Thermo Scientific, Cat. No. 16520050). Imaging of embryos was done with DeltaVision Elite imaging system (GE Healthcare) equipped with a 60× /1.42 NA Plan Apo oil-immersion objective. After that, the raw images were deconvolution treated. The deconvolution parameters included enhanced ratio (aggressive) deconvolution, 10 number of cycles, applied correction, normalized intensity, used photosensor, 50% camera intensity offset, and Olympus_60X_142_10612.otf files.
Screening endogenous RNAs containing repeated sequences
To find endogenous targets of CRISPR-dCas13 system, marker genes that are specifically expressed in each cell of clusters ranging from 4 to 24 hpf were screened from the published single-cell RNA-seq datasets [48]. The pipeline used for identifying the transcripts containing repeated sequences was detailed as below (also referred to Fig. 3a). Firstly, we selected 1383 transcripts, with the longest isoform of each marker gene, as candidates to search short repeated sequences that were used for CRISPR-dCas13 system targeting. Secondly, we searched 20 nt repeated motifs in these transcripts, and compared those fragment sequences to determine mismatches between any two fragments. We then collected the fragment clusters of repeated sequences, in which all fragments matched exactly or had only one mismatch in one of the fragments. Meanwhile, the number and position of the cluster-containing fragments in transcripts were recorded. Thirdly, the cluster-containing fragments presented in more than one gene were removed to achieve unique repeated sequences for CRISPR-dCas13 system targeting. Finally, we selected the fragments having the maximum one mismatch compared to each other within a cluster and used as candidates for labeling.
This computational pipeline generated 134 transcripts containing at least two repeats without overlapping in position Additional file 1: Fig. S4a; Additional file 6: Table S5), among which 15 transcripts contained at least eight repeated sequences (Additional file 1: Fig. S4b). The single-cell count matrices (accession number: GSE112294) were downloaded from Gene Expression Omnibus. The expression of transcripts containing repeated sequences in different cell types as reported in Wagner et al. [48] were calculated as the average of normalized UMI (Unique Molecular Identifiers) counts in each cell (Additional file 1: Fig. S4a, b).
The 134 transcripts with repeated sequences and their expression levels are listed in Additional files 4,5,6: Tables S3,4,5.
Whole-mount single-molecule RNA fluorescent in situ hybridization (smFISH)
All probes used for smFISH were designed via Stellaris Probe Designer with default parameters (https://www.biosearchtech.com/stellaris-designer) and synthesized in Tsingke biotechnology company.
Probes were labeled with Cyanine 3 (Cy3, 100537515 and muc5.1 probe) or Red 650 (GCN4 and eppk1 probe) at the 3′ ends by Terminal Transferase (NEB, Cat. No. M0315L). In brief, reaction mixture for a 20 μL system includes the following: 1 μL probe mix (stock 100 μM), 2 μL Cy3 or Red 650 (stock 0.2 mM), 2 μL CoCl2, 2 μL enzyme buffer, 0.5 μL Terminal Transferase, and 12.5 μL ddH2O to make a 20 μL final volume, according to the manufacturer’s protocol. Reaction was carried out at 37°C for 4 h and then purified with sodium acetate precipitation.
The procedure of whole-mount smFISH for zebrafish embryos was referred to the previous study with modifications [83]. Fish embryos were fixed in 4% paraformaldehyde (PFA) at 4 °C overnight. The next day, completely removed the fixation solution and washed the embryos twice with 1× PBST (1× PBS and 0.1% Tween-20) for 5 min each, followed by dechorionizing the embryos with 1-mL syringe. The embryos were then dehydrated with 50% methanol/50% 1× PBST once for 5 min and 100% methanol once for 5 min and kept in 100% methanol at −20°C for at least 4 h. After rehydration with 75% methanol/25% 1× PBST, 50% methanol/50% 1× PBST, and 25% methanol/75% 1× PBST one by one (each step took 5 min once), the fixed embryos were then washed twice with 1× PBST for 5 min each and were incubated in 2× SSCT (2× SSC and 0.1% Tween-20) once for 5 min. After that, the embryos were transferred to the prehybridization buffer (10% formamide, 2× SSC, 0.1% Triton X-100) at 30°C and kept for 10 min. Meanwhile, diluted the probe stock solution (5 μM) by the hybridization buffer (10% formamide, 2× SSC, 0.1% Triton X-100, 0.02% BSA, 2 mM Ribonucleoside Vanadyl Complex and 10% dextran sulfate) at the ratio of 1:20. The hybridization was done by incubating the embryos with 100-μL probe at 30°C for overnight (14–16 h). Then, the embryos were washed twice with the prehybridization buffer at 30°C for 30 min each, once with 2× SSCT at 30°C for 30 min, and once with 1× PBST for 5min at room temperature.
For nucleus staining, embryos were incubated with DAPI solution (1:1000, Thermo Scientific, Cat. No. D1306) for 2–5 min and washed twice with 1× PBST for 20 min each. Embryos post smFISH were imaged under the widefield microscopy.
Probe sequences are listed in Additional file 2: Table S1.
Total RNA isolation, cDNA synthesis and RT-qPCR
Zebrafish embryos at corresponding developmental stages were collected, and the total RNAs from equal number of embryos were extracted with Trizol Reagent (Invitrogen, Cat. No. 15596026) according to the manufacturer’s protocol. The cDNA synthesis was carried out using 5× PrimeScript RT Master Mix (TaKaRa, Cat. No. RR036A) according to the manufacturer’s protocol. Quantitative (q)PCR was performed using SYBR Green Realtime PCR Master Mix (TOYOBO, Cat. No. QPK-201) and with StepOnePlus real-time PCR system (Applied Biosystems).
Primer sequences for RT-qPCR are listed in Additional file 2: Table S1.
Whole mount in situ hybridization (WISH)
The DNA template for the probe targeting eppk1 contained a T7 promoter sequence (GAAATTAATACGACTCACTATAGGG) and was amplificated from cDNA of 24 hpf zebrafish embryos by primers (Additional file 2: Table S1). After purifying with the agarose gel, the eppk1 probe was transcribed in vitro by T7 polymerase (Thermo Scientific, Cat. No. EP0111) with 10× Digoxigenin RNA Labelling Mix (Roche, Cat. No. 11277073910) according to the manufacturer’s protocol and further purified with MEGAclear Kit (Thermo Scientific, Cat. No. AM1908). WISH was performed as described previously [84]. In brief, the embryos were fixed in 4% PFA at room temperature for 4 h. After dehydration and rehydration, which were the same as smFISH described above, embryos then were hybridized with eppk1 probe (1 μg/mL) at 65°C for 16–18 h, followed by incubation with anti-Dig-AP antibody (Roche, Cat. No. 11093274910) at 4°C overnight. The WISH signals were developed in 0.5 mL NBT/BCIP solution (one NBT/BCIP tablet dissolved in 10 mL ddH2O containing 0.1% Tween 20) (Roche, Cat. No. 11697471001), and the embryos were observed and captured with a stereomicroscope (Nikon SMZ18).
Time-lapse imaging to track transcription and mRNP motion
Before imaging, live embryos at corresponding developmental stages were dechorionated with 1-mL syringe and were mounted on the bottom of the dish with 1% low melting agarose. Embryos were maintained on the equipped live cell imaging chamber at the experimental temperature of 28.5°C.
To track RNA transcription, a serial 3D stack imaging (0.4 μm z-step) in time series was carried out using Olympus SpinSR confocal microscopy with the 60× /1.42 NA UPLXAPO oil-immersion objective and achieved at the 2048 × 2048 pixels field. During tracking eppk1 transcription in cell cycle 13, due to its relatively short cell cycle period, the image stack was recorded every 2 min. During tracking eppk1 transcription in cell cycle 14 as well as 100537515 transcription in cell cycles 14 and 15, the image stack was recorded every 5 min. The total tracking time lasted approximately 4 h. Twenty percent 488 nm laser power and 100 ms exposure time were used. Then, a clear image stack in time series was produced by the maximum intensity projection, and the signals at the transcription sites were tracked and analyzed over time manually.
mRNP motions were tracked with different time series. To track mRNP motion at 10-ms interval, serial 2D image stacks of each cell were acquired at the 512 × 512 pixels widefield by Multi-SIM, developed by Dong Li lab, Institute of Biophysics, CAS). The image stack in time series was collected with 100× /1.49 NA oil objective (Nikon CFI SR HP Apo) and detected by a sCMOS camera (ORCA-Fusion, Hamamatsu) with 80% 488 nm laser power, or with 100% 561 nm laser power (for simultaneous imaging of NPCs and mRNPs), at 2 ms exposure time and 10-ms interval for 30 s (3000 fames). Image stacks in time series were denoised as described in the section “Single-particle tracking” below.
mRNP export events were tracked in different time series. To track mRNP export at 50-ms, 200-ms, and 2-s intervals, different 2D image stacks in time series were acquired by Olympus SpinSR confocal microscopy. One or multiple cells were recorded at different position of the 2048 × 2048 pixels field using a 100× /1.50 NA UPLXAPO oil-immersion objective with 100% 488 nm laser power. Forty nine-millisecond exposure and 50-ms interval for 100 s (2000 fames), 100-ms exposure and 200-ms interval for 100 s (1000 fames), and 100-ms exposure and 2-s interval for 15 min (450 frames) were applied, respectively.
Imaging processing and analysis
Images of fixed and live embryos were analyzed by Fiji (ImageJ, https://imagej.net/Welcome). Representative images from widefield imaging stacks were performed with maximum intensity projection.
Colocalization analysis
The signals were selected using straight line and were analyzed by plot profile. We recorded each channel of data and quantified the relative intensity over the distance performed with GraphPad Prism 8.
Quantification of signal-to-noise ratio (SNR)
SNR was defined as the ratio of the intensity of a fluorescent signal to the power of the background noise. The puncta at the transcription sites were selected with a circle (the diameter of which was 2–3 μm), and the puncta signal was measured with the max intensity. The center of the puncta (exclude the puncta) as background was measured with mean intensity of background. Calculating the SNR with the formula below:
$$\textrm{SNR}=\left(\max\ \textrm{intensity}\ \textrm{of}\ \textrm{puncta}\ \textrm{signal}-\textrm{mean}\ \textrm{intensity}\ \textrm{of}\ \textrm{background}\ \textrm{EGFP}\ \textrm{signal}\right)/\textrm{std}.\textrm{dev}.\textrm{of}\ \textrm{background}\ \textrm{EGFP}\ \textrm{signal}.$$
Detection of mRNP export events
We generated all the mRNPs’ trajectories of each cell using maximum time projection from acquired time-lapse movies. The trajectories of export events were detected between the nuclear edge and cytoplasm. We confirmed the exporting events in real time and calculated the time of mRNP export manually. In brief, the nucleocytoplasmic export of mRNPs was firstly observed along the nuclear boundary shown by maximum time projection of dPspCas13b-3× sfGFP or along NPCs labeled by Pom121-mScarlet. Before releasing into cytoplasm, many mRNPs would dwell on NPC region for varying time from milliseconds to minutes. After releasing into cytoplasm, these mRNPs would begin quick diffusion to leave the nuclear edge within two continuous frames, from which we could determine the start and the end of the exporting process. Then we calculated the time of mRNPs moving from the nuclear boundary or NPCs to the cytoplasm. In the cytoplasm, nuclear exported mRNPs could be tracked for about 1 μm or even longer distance, for example shown in Fig. 5b and Additional file 7: Movie 14. To calculate the exporting time of directed export events, we first estimated the length of NPCs, which was about 200 nm (nuclear basket ~75 nm, central framework ~70 nm, and cytoplasmic filaments ~50 nm) [85, 86]. Then, we calculated the velocity of the directed export mRNPs from the single-particle tracking data and estimated the time of directed export events as 139 ms ± 66 ms.
Quantification of normalized transcription activity
The complicated background, including the movement of EVL cells, other cell type interference, and non-specific aggregation in a fraction of EVL cells (Additional file 7: Movies 1, 2, 5 and 6), made it challenging to extract fluorescent traces of each allele over time with the available algorithm. The maximum intensity projection was performed to produce clear image stack in time series. The signals were identified at the transcriptional locus manually. During cell mitosis, we determined this process by using nuclear morphology as the EGFP fluorescence shown. After re-establishing the clear nucleus, we began to record the signal to measure the re-initial transcriptional activity and observed that most of the eppk1 and 100537515 allelic re-initial expression occur within 10 min post-mitosis. A puncta region of signals was detected by circle region of interest (ROI-1) to measure the maximum intensity as the puncta transcriptional activity (ROI-1max). A larger circle region (ROI-2, three times the radius than ROI-1) was used surrounding the puncta to measure the mean intensity out of the puncta region as the local background (ROI-2mean). In general, for each data set, the ROI with fixed size was chosen to include the local signal and background throughout all time points. If the puncta signal was too low to be identified, the “puncta signal” was then assigned as the local background signal. Thus, the normalized transcription activity at each punctum and each time point was calculated by the following formula:
Normalized signal intensity = (ROI-1max − ROI-2mean) / ROI-2mean, which would correct the photobleaching over time. All normalized transcription activities were then plotted against the time/duration.
Quantification of smFISH signals
To quantify the 100537515 smFISH signals in the nucleus and cytoplasm, we used 3D Object Counter plugin in Fiji (https://imagej.nih.gov/ij/plugins/track/ objects.html). Briefly, we cropped the image of single EVL cell from raw data with deconvolution. The smFISH signals were identified by the manual bandpass threshold using objects counter 3D, and the number of particles was measured as the total number of 10053751 mRNA signals. Then, the nucleus was chosen based on DAPI signals, and the chosen region was added into ROI manager. After that, we obtained the 10053751 mRNA signals in nucleus and cytoplasm by calculating the smFISH particles.
The correlation of inter-allelic transcription output and inter-allelic difference expression
The total output of alleles was the integrated area under the normalized fluorescence trajectory. In details, the normalized signal intensity of puncta1 or puncta2 was cumulated from each observed cell. Then the cumulated fluorescent signals of puncta1 and puncta2 were used for drawing correlation plot in each cell cycle by ggplot (v3.3.4). Pearson correlation coefficient, Spearman’s rank correlation coefficient, and slope were calculated by R (v4.1.3, http://www.R-project.org/). For comparison, the cumulated fluorescent signals of puncta1 and puncta2 from randomized cells were paired and used for correlation analysis as described above.
For measurement of the allelic difference expression within a cell, the difference expression was calculated between normalized signal intensity of puncta1 and puncta2 for each observed cell, then the significance levels of difference expression comparison at each time point were calculated from paired two-tailed Student’s t test. For inter-allelic correlation at each time point, Pearson correlation coefficient was calculated with normalized signals intensity of inter-alleles, and the significance levels of the correlation were calculated from unpaired two-tailed Student’s t test. To compare inter-allelic correlation of gene’s two cell cycles, we collected Pearson correlation coefficient at all time points for each cell cycle; then, p-values of the comparison of Pearson correlation coefficient between different cycles were calculated from unpaired two-tailed Student’s t test. No statistical method was used to pre-determine sample size. No data were excluded from the analyses. All statistical analyses were performed with R package 4.1.1.
Single-particle tracking
Live cell images were acquired using the single-molecule tracking mode integrated into the Multi-Modality Structured Illumination Microscope (Multi-SIM) [87]. For each cell, 3000 frames of wide-field images were acquired at the speed of 10 ms per frame. Before single-molecule tracking analysis, the time-lapse images from a total of 17 cells were denoised with the optimization function:
$$\mathit{\arg}\ \underset{g}{\mathit{\min}}{\left\Vert f-H\otimes g\right\Vert}_2^2+\lambda {\left\Vert g\right\Vert}_1+\mu {\left\Vert \nabla g\right\Vert}_2^2$$
(1)
where f denoted the raw image, g denoted the denoised image, H denoted the point spread function of the system, ∇g denoted the 1st-order derivative of denoised image, ‖·‖n denoted the nst matrix norm, and λ and μ denoted the weight of each corresponding term.
After denoising, the image stacks were analyzed with the Fiji plugin of TrackMate developed for automated single-particle tracking [88], which generated the position and time information for all tracks. After filtering out the tracks of less than 20 frames duration, more than 10,000 tracks were collected for motion classification. Next, the tracked data were analyzed with MATLAB (https://www.mathworks.com). First, we computed the mean squared displacement (MSD) for each track, and then fitted with different model functions to classify them into four types of directed, diffusive, corralled, and stationary according to the following protocol.
-
(1)
If the ratio of the smaller to the larger principal radius of gyration was less than 0.001 and the max displacement was longer than 2 μm, then the track was mostly linear, which was classified as directed motion [6, 89]. And its MSD-t curve was fitted with:
$${\displaystyle \begin{array}{c} MSD(t)={(vt)}^2\end{array}}$$
(2)
where v was the linear velocity.
-
(2)
If a trajectory was not directed motion, we fitted its MSD-t curve with the following two equations, respectively [90, 91].
$${\displaystyle \begin{array}{c} MSD(t)=2 mD\ast {t}^{\alpha}\end{array}}$$
(3)
where t was lag time, m was the dimensionality of image, D was diffusion coefficient, α was the anomalous exponent.
$${\displaystyle \begin{array}{c} MSD(t)=\frac{L^2}{3}\left(1-{e}^{-\frac{t}{\tau }}\right)\end{array}}$$
(4)
where L was the size of confined microdomains, τ represented equilibration time, and its diffusion coefficient D was given by D = L2/(12τ).
Then we checked which fitting result was better [92]. If Eq. (3) more precisely described the trajectory, the molecule underwent diffusive motion and had few interactions with surrounding components. Otherwise, the molecule was confined into a limited area that was classified as corralled motion. For directed motion, the diffusion coefficients were fitted with Eq. (3), where alpha approached 2.
-
(3)
We found some molecules were classified as diffusive or corralled motion, but their max displacement and diffusion coefficient were quite small. Therefore, we defined such motions as stationary particles if the max displacement and diffusion coefficient of a trajectory were smaller than 400 nm and 0.03 μm2/s, respectively.
Other quantification analysis
Significant difference was calculated with unpaired two-tailed Student’s t test, paired two-tailed Student’s t test, or Mann-Whitney test, and histogram and line chart were plotted with GraphPad Prism 8. The data was presented as mean ± standard deviation (SD) or mean ± standard error of the mean (SEM) in triplicate experiments, unless otherwise stated; see also figure legends and methods for details. At least two independent experiments were done to gain representative images for microscopy imaging. For the statistical significance and sample size of all graphs, please see figure legends for details.
