In vivo imaging of mitochondrial DNA mutations using an integrated nano Cas12a sensor

Mutations in mitochondrial DNA (mtDNA) play critical roles in many human diseases. In vivo visualization of cells bearing mtDNA mutations is important for resolving the complexity of these diseases, which remains challenging. Here we develop an integrated nano Cas12a sensor (InCasor) and show its utility for efficient imaging of mtDNA mutations in live cells and tumor-bearing mouse models. We co-deliver Cas12a/crRNA, fluorophore-quencher reporters and Mg2+ into mitochondria. This process enables the activation of Cas12a’s trans-cleavage by targeting mtDNA, which efficiently cleave reporters to generate fluorescent signals for robustly sensing and reporting single-nucleotide variations (SNVs) in cells. Since engineered crRNA significantly increase Cas12a’s sensitivity to mismatches in mtDNA, we can identify tumor tissue and metastases by visualizing cells with mutant mtDNAs in vivo using InCasor. This CRISPR imaging nanoprobe holds potential for applications in mtDNA mutation-related basic research, diagnostics and gene therapies.

PAM sequence of Cas12a is indicated by bold orange letter.Targe site in dsDNA of Cas12a is indicated by underscore.SNV is indicated by bold red letter.Each mismatched position is indicated by bold blue letter.Notes to Supplementary Fig. 2 c-g.

Supplementary Table 2. Fluorescent dyes used in this study.
We performed Bio-TEM experiments for HepG2 cells treated with InCasorCyt C apt+ or InCasorCyt C apt-.First of all, different spherical nucleic acids (SNA) were synthesized for labeling DNF and Cas12a RNP on InCasor respectively) for Bio-TEM experiments.
In brief, we designed a two-domain DNA comprising an DNF domain (for labeling DNF, the DNA domain is similar to apart of CLR; for labeling Cas12a RNP, the DNA domain is complementary to 5` of crRNA, blue lines) and SH-modified poly T (T1 to T20) domain (red lines, sequences listed in Table S1).The SH-modified poly T -DNAs were assembled with AuNPs (~15 nm diameter) via a freezing method as described 1 .
UV absorption spectrum of the DNA-AuNP conjugates show characteristic absorption peaks of nucleic(λ260nm) acid and AuNP(λ520nm) at the same time (Supplementary Fig. 2c) and transmission electron microscopy (TEM) images of the DNA-AuNP conjugates show a thin and low-contrast shell coating on the AuNPs (Supplementary Fig. 2d), suggesting that the polyT-DNAs were successfully attached to AuNPs.From the TEM of InCasor-SNADNF, the SNADNF were successfully labelled on InCasor probe (Supplementary Fig. 2e).Then, Bio-TEM was performed to show the location of InCasorCyt C apt+ or InCasorCyt C apt-labelled by SNADNF or SNARNP in the HepG2 cells.
As you can see from the following Supplementary Fig. Notes to Supplementary Fig. 3 d-l.
A padlock probe based rolling circle amplification (RCA) methods was firstly applied based on the principles described in Supplementary Fig. 3c 2 ,3 .Padlock probes hybridize to their target sequence (crRNA) and the probe ends are joined through ligation, locking the probe onto the target molecule.After ligation, the RCA is initiated by the Phi 29 DNA polymerase, which turns the target molecule into a primer by 3`-5` exonucleolytic cleavage of any 3`end protruding from the hybridization site of the padlock probe.The padlock probe then serves as the template for DNA synthesis.The RCA product (gray) is detected through agarose gel electrophoresis.After verifying the feasibility of this method (Supplementary Fig. 3d), we examined whether total RNA of mitochondria in HepG2 cells with different treatments can triggered RCA responses.
Like the positive control crRNA group, total RNA in InCasorCyt C apt+ group was also able to elicit RCA responses, while neither InCasorCyt C apt-group nor untreated group showed RCA products (Supplementary Fig. 3e).These results suggested that there was crRNA of InCasor probe in the total RNA sample extracted from mito of InCasorCyt C apt+ group.
Moreover, a CRISPR-Cas13a based fluorescence assay was employed to detect the crRNA in total RNA sample.A CRISPR guide RNA (crRNA) complementary to 20 nucleotides of the crRNA of Cas12a was used to direct Cas13 from Leptotrichia wadei (LwaCas13a) to the target sequence.Detection of the target resulted in Cas13 activation and subsequent collateral cleavage of an oligonucleotide carrying a quenched fluorophore that exhibits fluorescence when cleaved, correlating with the initial concentration of the crRNA in the total RNA sample from mito (Supplementary Fig. 3f).As shown in Supplementary Fig. 3g, only the target crRNA can trigger the transcleavage activity of Cas13a, resulting in an obvious fluorescent signal.We, therefore, continued to use this method to detect the presence of crRNA in different RNA samples.
The significantly increased fluorescence signal in InCasorCyt C apt+ group demonstrated the presence of crRNA (Supplementary Fig. 3h).Together, the results indicated that our InCasor probe can delivery crRNA to mitochondria in living HepG2 cells.
In addition, we analyzed the DNA (equal mass) from mito in HepG2 cells by agarose gel electrophoresis (Supplementary Fig. 3k).The InCasorCyt C apt+ group had diffuse bands around 500 bp, which were similar to the DNF positive control group after acid treatment 4 , indicating that these fragments may be produced by InCasor probe in mitochondria.To further verify that these bands were derived from InCasor probe, we performed PCR with DNA extracted from mitochondria using DNF-specific primers.
As shown in Supplementary Fig. 3l, the diffuse band circled by the red box indicated that the DNF of InCasor probe was in the mitochondria.
treated with 50 nM InCasorNT or InCasorND4 (Supplementary Fig. 8a).Despite the presence of one region with significant homology to the mtDNA target site in the nuclear genome, no evidence for off-target effects exerted by InCasorND4 could be detected at these sites (Supplementary Fig. 8 b-c).In addition, no evidence for nonhomologous end joining at the target site in mtDNA could be detected, confirming previous data that InCasorND4 induced double-stand DNA breaks do not result in nonhomologous end joining activity (Supplementary Fig. 8d).Furthermore, we analyzed the single-nucleotide variants (SNVs) and indels at an average depth of 40.Notes to Supplementary Fig. 13 We created cells with heteroplasmic mtDNA, i.e., containing both mitochondria with wild-type mtDNA (WT-Mito) and mitochondria with mutant mtDNA (MT-Mito), by isolating mitochondria from MDA-MB-231 cells and transfecting them into HepG2 cells.Before that, we first examined mitochondrial DNA heterogeneity in HepG2 cells and MDA-MB-231 cells separately by pyrosequencing (Fig. 5a).WT-Mito were labeled with MitoTracker (red), and MT-Mito were pre-stained with CellMask (green).
During 24 hours of internalization, labeled MT-Mito first bound to the cell surface (3-6 hours), then entered the cytoplasm (12 hours), and finally localized near the nucleus (18-24 hours) (Supplementary Fig. 13a).The spatial co-localization of nuclei (blue) and MT-Mito (green) from static analysis in Fiji also indicated that the distribution, or motility, of labeled MT-Mito in HepG2 cells was nucleotropic (Supplementary Fig. 13b).Meanwhile, no co-localization of MT-Mito and WT-Mito was observed, and the PCC between them was always less than 0 (Supplementary Fig. 13c).We speculated that MT-Mito is compromised during extraction and fluorescent labeling to tend to the vicinity of the nucleus where mitophagy occurs (Supplementary Fig. 13d).
To further determine mitochondrial DNA heterogeneity and its stability in heterogeneous cells, we investigated its heterogeneity using in situ RCA and pyrosequencing, respectively.We used a pair of padlock probes specific for the two sequence variants to genotype the mitochondrial 13105A>G point mutation in situ in heterogeneous cells.RCA products were detected by hybridization of two fluorescent oligonucleotide probes with sequences identical to distinct tag sequences in the two variant-specific padlock probes.As shown in Supplementary Fig. 13 e-f, the strong discrete signal shows the distribution of the two mitochondrial genome variants in the heterogeneous cells.

Supplementary Fig. 1 .
Preparation and characterization of InCasor.a. Particle sizes and Zeta potential of DNF, DNF/CLR and InCasor probe.(n = 3).Data are presented as mean ± SD. b.SEM characterization of DNF, DNF/CLR and InCasor probe.Scale bar: 1 μm.Inset shows the corresponding TEM image.Scale bar: 200 nm.The experiments were repeated three times independently.c.In vitro fluorescence assay to analyze the stability of InCasor in 90% serum over time.n=6 biologically independent experiment, data show mean ± SD. d.DLS analysis of particle size changes of InCasor before and after recognition of target mtDNA in vitro.n=3 biologically independent experiment, data are analyzed by two-sided Student's t-test and shown as mean ± SD. e-i.Comparison of the mtDNA detection efficiency of the SSR (e), CLR (f), DNF/SSR (g), DNF/CLR (h), and InCasor (i).n=3 biologically independent experiment, data show mean ± SD. j.Variation of particle size and potential of InCasor with time at 4 °C storage conditions.n=3 biologically independent experiment, data show mean ± SD. k.The ability of InCasor to recognize target DNA to produce a fluorescent signal over time under storage conditions at 4 °C.n=3 biologically independent experiment, data show mean ± SD.Source data from (a, and c-k) are Supplementary Fig. 2. Subcellular localization of InCasor.a. Confocal imaging and flow cytometry analysis of HepG2 cells incubated with 50 nM FAM-InCasorSgc8+ or FAM-InCasorSgc8-.Scale bar: 10 μm. b.Fluorescent imaging of the subcellular localization of InCasor in HepG2 cells after different incubation time.InCasor was labelled with FAM (green), and lysosomes were stained with Lysotracker (red).Scale bar: 10 μm.The spatial co-localization of InCasor and lysosomes were statically analyzed by Fiji.c. UV absorption spectrum of Au NPs, SNADNF and SNARNP.d.TEM characterization of Au NPs and SNADNF.Scale bar: 20 nm.e. TEM characterization of InCasor and InCasor-SNADNF.f.Bio-TEM of HepG2 cells treated with SNADNF labelled InCasorCyt C apt+ or InCasorCyt C apt-.g.Bio-TEM of HepG2 cells treated with SNARNP labelled InCasorCyt C apt+ or InCasorCyt C apt-.Mitochondria are circled with white dotted lines.The red arrow points to SNA.The scale bars are 2 μm, 500 nm and 500 nm, respectively.The experiments were repeated three times independently.Source data from (b, c) are provided as a Source Data file.
2 f-g, numerous SNA labelled InCasorCyt C apt+ probe was located mitochondria and part of it was in the endosomes/ lysosomes, cytoplasm.Meanwhile, there are no Au NPs in the mitochondria of untreated cells or SNA labelled InCasorCyt C apt-treated cells.These results reveal that InCasor probe effectively targets mitochondria in living HepG2 cells.Supplementary Fig. 3. Quantitative characterization of mitochondria targeting of InCasor.a. Workflow for quantitative characterization of mitochondria targeted by InCasor.First, InCasorCyt C apt+ or InCasorCyt C apt-probe were incubated with HepG2 cells for 6 h.After that, the cells were collected and the mitochondria in them were isolated.Proteinase, RNase, and DNase protection assays were used to degrade InCasor adsorbed on the outer mitochondrial membrane, beforehand extracting DNA, protein or RNA from the mitochondria for the next step of quantitative analysis.b.Analysis of proteins in Mito extracted from HepG2 cells after incubated with InCasorCyt C apt+ or InCasorCyt C apt-probe through SDS-PAGE gel.The proteins were stained by coomassie brilliant blue.c.Schematic diagram of RCA method for specific crRNA detection.d.Feasibility analysis of crRNA triggering RCA reaction.e. Agarose gel electrophoresis analysis of the product of the RCA reaction initiated by crRNA in Mito extracted from HepG2 cells after incubated with InCasorCyt C apt+ or InCasorCyt C apt-probe.f.Schematic diagram of Cas13a method for specific crRNA detection.g.In the presence of target crRNA, the trans-cleavage activity of Cas13a is able to in to generate fluorescence signal.n=3, data show mean ± SD. h.Fluorescence analysis of the Cas13a RNP incubated with RNA in Mito extracted from HepG2 cells after incubated withInCasorCyt C apt+ or InCasorCyt C apt-probe.n=3, data are analyzed by two-sided Student's t-test and shown as mean ± SD. i.The Pearson's correlation coefficient (PCC) of Cas12a RNP on InCasor and mitochondrial in Fig.2ewere statically analyzed by Fiji.n=3, data show mean ± SD. j.Quantification of Cy5-Cas12a/ FITC-crRNA in mitochondria using flow cytometry.HepG2 cells were incubated with InCasorCyt C apt+ or InCasorCyt C apt- probe (both loading Cy5-Cas12a/ FITC-crRNA).Cells without any transfection were used as the negative control (NC).Crude mitochondria were isolated from cells at 12 h after transfection, and were subjected to flow cytometry.k.Agarose gel electrophoresis analysis of DNA in Mito extracted from HepG2 cells after incubated with InCasorCyt C apt+ or InCasorCyt C apt-probe.l.Agarose gel electrophoresis analysis of PCR product of DNA in Mito extracted from HepG2 cells after incubated with InCasorCyt C apt+ or InCasorCyt C apt-probe.The experiments were repeated three times independently.Source data from (b-e, g-I, and k-l) are provided as a Source Data file.
Genotyping of the HepG2 and MDA-MB-231 cell lines, homoplasmic for the two different genotypes.Nuclei were stained with DAPI (blue), Cy3-labeled padlock probes on wild-type mitochondrial DNA (red) and AF488-labeled padlock probes on mutant mitochondrial DNA (green).Scale bar: 10 µm.i.In situ genotyping of the heterogeneous cells.Nuclei were labelled by DAPI (blue), AF488-labeled padlock probes on wild-type mitochondrial DNA (red) and Cy3-labeled padlock probes on mutant mitochondrial DNA (green).Scale bar: 10 µm.g.Analyzing heteroplasmic mtDNA mutations in live cybrid cells using InCasor with crRNA2 or crRNA3.Scale bar: 2.5 μm.Cell nucleus are shown in blue, WT-Mito were labelled by Mito-Tracker (red), InCasor probe would generate fluorescent signal (green) when binding targeted mtDNA.Scale bar: 10 μm.h.PCC of purple signal with green or red fluorescence in panel g were statically analyzed.n=3,datashow mean ± SD. i. Semi-quantitative statistics are performed on the three types of fluorescence intensities (MT-Tracker, cell mask deep red and FAM) of each mitochondrial in the live cybrid cells.(n=50).Source data from(b, c, d, h, and i)are provided as a Source Data file.
group (Supplementary Fig.8e).We then filtered SNVs and indels using the 104 Cas-OFFinder predicted off-target sites and the WGS sequence of NC group.None of them were located in the coding region, suggesting no functional off-target sites in