Large-scale preparation of fluorescence multiplex host cell reactivation (FM-HCR) reporters

Abstract

Repair of DNA damage is a critical survival mechanism that impacts susceptibility to a variety of human diseases and represents a key target for cancer therapy. A major barrier to applying this knowledge in research and clinical translation has been the lack of efficient, quantitative functional assays for measuring DNA repair capacity in living primary cells. To overcome this barrier, we recently developed a technology termed fluorescence multiplex host cell reactivation (FM-HCR). We describe a method for using standard molecular biology techniques to generate large quantities of FM-HCR reporter plasmids containing site-specific DNA lesions and using these reporters to assess DNA repair capacity in at least six major DNA repair pathways in live cells. We improve upon previous methodologies by i) providing a universal workflow for generating reporter plasmids, ii) improving yield and purity to enable large-scale studies that demand milligram quantities, and iii) reducing preparation time more than 10-fold.

INTRODUCTION

Repair of DNA damage from both endogenous and environmental sources is essential for maintaining genomic stability and reducing the risk of mutations and cell death. Furthermore, unrepaired damage in surviving cells can lead to cellular dysfunction including replication stress ¹, transcriptional stress ², senescence ³, and cellular transformation. DNA repair capacity (DRC) varies from person to person, and inefficient DNA repair is associated with cancer predisposition, premature aging, immunodeficiencies, and radio-sensitivity ^4-7. As a result, there is great interest in identifying susceptible individuals on the basis of their DRC in normal tissues, with the goal of personalized disease prevention. Likewise, as many cancer therapies are more effective in DNA repair deficient cancer cells, there is great interest in predicting patient responses based on DRC in cancer cells⁸. For example, DNA repair defects in specific pathways are associated with better responses to several drugs including temozolomide (direct reversal of alkylation damage) ⁹, poly (ADP-ribose) polymerase (PARP) inhibitors (homologous recombination, HR) ¹⁰, cisplatin (inter-strand crosslink repair) ¹¹, and PD-1/PD-L1 immune checkpoint blockade therapy, (mismatch repair, MMR)¹². Thus, the identification of novel functional biomarkers for DNA repair processes could allow physicians to tailor therapies for individual patients.

Fluorescence multiplex host cell reactivation (FM-HCR) is a powerful technology for measuring the DRC of living cells (Figure 1). One major advantage of this technology is that it directly reports the functional status of DNA repair in living cells, rather than relying on cell-free extracts. As these reporter plasmids are prepared in vitro, they can be engineered to contain one or more chemically defined lesions with total control over sequence context. FM-HCR allows for the simultaneous analysis of several DNA repair pathways (multiplexing), by leveraging the expression of multiple fluorescent proteins (FPs) and the discriminatory power of flow cytometry^13,14. These advantages could not be realized with earlier host cell reactivation assays that utilized enzymatic activity as a reporter^15-17. More recently, the pioneering work of Paul Doetsch has revealed the misincorporation of nucleotides by RNA polymerase during transcriptional bypass of a variety of lesions ¹⁸. This process is termed transcriptional mutagenesis and can be measured using reporters wherein the altered RNA sequence encodes a functional reporter protein (Figure 2)¹⁹.

Figure 1. Schematic Representation of FM-HCR Workflow.

a, Multiple plasmids encoding different fluorescent reporter proteins, indicated here by representing the circular reporter plasmid DNA in four different colors, enable multiplexing. b, Each type of damage is generated in vitro and assigned to a plasmid encoding a different fluorescent reporter protein. DNA lesions are marked with broken red circles, c, Cells of interest are transfected with reporter plasmids, allowed time for repair, and analyzed for fluorescence by flow cytometry. Created using biorender.com.

Figure 2. Schematic Representation of Using Transcriptional Mutagenesis as a Basis for FM-HCR Reporter Plasmids.

a, Transcription of plasmids in the absence of damage yields an RNA transcript encoding a non-fluorescent protein. In this case, the undamaged base is guanine, and the transcript contains a cytosine at the corresponding position b, An unrepaired DNA lesion results in RNA polymerase producing a transcript that encodes the wild type fluorescent protein. In this case, an O⁶-MeG lesion results in occasional misincorporation of uracil at the corresponding position of the transcript.

We and others have demonstrated that fluorescent host cell reactivation assays can measure DRC in at least six major pathways, which include nucleotide excision repair (NER) ^13,20, base excision repair (BER) initiation and intermediate processing²¹, non-homologous end joining (NHEJ) ^22,23, HR ^24,25, direct reversal (DR) by methylguanine methyltransferase (MGMT) ^8,13,26,27, and MMR ^13,28. The key features of FM-HCR that distinguish it from conventional HCR assays are: i) the ability to multiplex and ii) the use of transcriptional mutagenesis to monitor repair of DNA lesions that neither block transcription nor result in a change in sequence of the DNA. Because transcriptional mutagenesis has now been characterized for many DNA lesions¹⁸, it is possible to design novel plasmids that report the presence of unrepaired DNA lesions as a fluorescent signal.

By measuring the efficiency of multiple DNA repair pathways at once, one can achieve a comprehensive characterization of the functional DNA repair landscape. FM-HCR assays hold great promise for both large-scale population studies and screens of cell lines, small molecule libraries, and the identification of functional biomarkers for predicting health outcomes. Until now, the most important obstacle in carrying out ambitious functional studies lay in generating sufficient quantities of reporter plasmid. The main difficulties in substrate preparation previously included: i) a large time commitment in synthesising reporter substrates (weeks to months), ii) a large financial investment due to costly specialized reagents and equipment, and iii) the inclusion of multiple gel purification steps at intermediate stages of preparation, substantially reducing the yield and limiting purity. The protocol described here presents a major advance in the generation of site-specifically modified plasmid substrates by providing a unified workflow while dramatically reducing the costs and hands-on time, improving purity, and enhancing the yield of the preparation.

Overview of the procedure:

This protocol describes the preparation of reporter plasmids with chemically defined site-specific DNA lesions at microgram to milligram scale. Minimal expertise required to carry out the protocol includes experience with molecular biology, culture of bacteria and mammalian cells, and familiarity with flow cytometry. The plasmids are mammalian expression vectors that encode FPs and enable quantitation of multiple DNA repair transactions in living cells. The protocol describes the steps required to incorporate a chemically modified oligonucleotide into a dsDNA plasmid in vitro. This is accomplished in three ‘Stages’; ‘Stage 1’ (steps 1 → 41), ‘Stage 2’ (steps 42 → 69), and ‘Stage 3’ (steps 70 → 101) using both in vitro and in vivo analyses (see Figure 3). Briefly, once a sufficient quantity of starting plasmid is obtained and validated (steps 1 → 7), the starting plasmid is nicked with a strand-specific endonuclease, which is then validated and recovered (steps 8 → 21). The resultant single stranded closed circular plasmid DNA (ssDNA) nick creates a substrate suitable to permit exonuclease III (Exo III) digest of the nicked strand, yielding closed circular ssDNA (steps 22 → 29). A 5’ phosphorylated oligonucleotide containing the site-specific lesion of interest is then annealed to the ssDNA, creating the necessary substrate to permit extension by T4 polymerase and ligation by T4 ligase (steps 30 → 41). The remainder of the protocol is focused on recovering the newly synthesized highly pure dsDNA plasmids containing the modification of interest. This is accomplished by heat inactivation (HIA) of the extension reaction, followed by the removal of any open circular, linear or single stranded reaction intermediates with T5 exonuclease (T5 Exo). Additionally, proteinase K and phenol:chloroform extractions serve to remove any remaining protein contaminants (steps 42 → 46). The remaining closed circular DNA (ccDNA) species represent the desired final modified plasmid and is subsequently purified via a modified polyethylene glycol (PEG) precipitation (steps 47 → 58). In steps 59 → 68, we describe the preparation of two reporter plasmids for double strand break repair that do not require the primer extension strategy. In all cases, once the final reporter plasmid is recovered, it is subjected to quality control (QC) measures to ensure: (i) the correct sequence, (ii) the expected size (using agarose gel electrophoresis), (iii) presence of the desired DNA lesion (using analytical digests via agarose gel electrophoresis), and (iv) in vivo functional validation (steps 70→96) using flow cytometry in appropriate cell line controls and finally quantitation of repair capacity using the Eq’s 1 → 3 (Step 97). Lastly, we include an example multiplexed cocktail setup for the simultaneous measurement of MMR, DR, and BER DRC’s (steps 98 → 101).

Figure 3. Schematic Representation of the FM-HCR Reporter Preparation Protocol and Gel Electrophoretic Analysis of Intermediates and Products.

a, (Steps 1-7) Production of the substrate begins using a ccDNA with the oligonucleotide-binding sequence of interest. (Steps 8-21) A nick is introduced into the plasmid on the strand that will modified with a site-specific DNA lesion. (Steps 22-29) The nicked strand is digested using Exo III to generate circular ssDNA. (Steps 30-33) An oligonucleotide containing the lesion of interest is annealed to the ssDNA. (Steps 34-42) The annealed plasmid:oligo duplex is incubated with T4 Polymerase and T4 Ligase produce ccDNA containing the lesion of interest. (Steps 42-69) Pure closed circular product is obtained through incubation with T5 Exo followed by Proteinase K digest, DNA extraction and precipitation, and dissolution in TE buffer. Note, procedures in steps 59-68 result in linearized plasmids. All yields refer to reporter plasmids generated using the oligonucleotide extension methodology. (Steps 70-101) substrates are subjected to a battery of in vitro and in vivo QC steps. b, Representative agarose gel electrophoretic analysis of starting ccDNA (Lane 2), ocDNA (Lane 3), circular ssDNA (Lane 4) and the final purified product (Lane 5) are shown. c, Representative gel analysis showing evidence of incomplete digest of ccDNA with nicking enzyme (Lane 3) and incomplete digest of nicked DNA with Exo III (Lane 5). d, Overexposed representative gel analysis showing evidence of residual ssDNA due to incomplete extension when the molar ratio of oligonucleotide to ssDNA is too low (1:1, Lane 5 and 1:2 Lane 6) and an absence of ssDNA when the ratio is optimal (4:1, Lane 7). e, Representative gel analysis showing accumulation of open circular DNA in a preparation where the proteinase K step was excluded (Lane 5), and the near absence of nicked product when proteinase K is included (Lane 6). In panels b-e, the relevant steps of the protocol are indicated below the gel.

Development of the protocol:

The initial reports of FM-HCR were primarily focused on demonstrating a proof of concept by confirming the presence of the intended DNA lesions by validating the assays in cell lines with defined DNA repair defects ^13,14. Some key applications that originally motivated the development of these assays included their potential use in large-scale studies, such as cell line screens and molecular epidemiology, as such studies require large (microgram to milligram) quantities of the reporter plasmids. They also require highly purified and consistently produced materials given the goal of detecting relatively small differences in repair capacity, which could be confounded by potential plasmid-batch dependent effects. Thus, the relatively low (nanogram to microgram) yields were originally a critical limiting factor in FM-HCR assays. Beyond these considerations, the FM-HCR assays have generated considerable interest among our colleagues in the broader genome integrity research community. We developed this protocol to meet the demands of these applications, and to provide a complete set of procedures and guidelines for others who are interested in applying or generating FM-HCR reporter plasmids in their own laboratories. The protocol synthesizes and refines previously published reports of the assays, their application, and their interpretation, as well as recommended best practices for their future use.

Traditional host-cell reactivation assays utilize reporter plasmids that have been treated with DNA-damaging agents (radiation and chemical treatments) that induce randomly-distributed DNA lesions²⁹. This approach is sometimes advantageous as these reporters can be generated very easily and in large quantities, and we have used UV irradiation to generate NER reporters¹³, as well as treatment with ionizing radiation and DNA-damaging chemicals to generate FM-HCR reporters (unpublished). However, a significant drawback of this approach is that the plasmids contain heterogeneous mixtures of DNA lesions, making it difficult to relate a repair phenotype to a specific DNA lesion. Therefore, the majority of FM-HCR reporter plasmids contain chemically defined site-specific DNA lesions that are introduced into DNA during solid-phase oligonucleotide synthesis. In contrast with chemical or radiation treatments, preparation of these reporter plasmids requires more specialized methodologies. The essential principles of reporter plasmid preparation were adapted from those in earlier reports ^30-32 and are summarized in Figure 3a, along with associated gel intermediates, troubleshooting, and final substrates (Figure 3b-e). Previously, a relatively time-consuming and laborious procedure was employed to generate reporter plasmids. Drawbacks included the need for both T4 and Pfu DNA polymerases and the use of multiple gel extraction and purification steps. Furthermore, a universal procedure for plasmid preparation was not possible. Instead, each reporter plasmid was prepared using slightly different methods, with different incubation times and temperatures that were empirically determined to maximize yield.

Seeking to improve yield, purity and reproducibility, we developed a user-friendly universal workflow for more rapidly and efficiently generating FM-HCR substrates. Major advantages over our previously described methodology and those described elsewhere in the literature include the use of a single reaction vessel for “one-pot” synthesis, the absence of gel extraction, chromatography, or density gradient centrifugation to separate plasmid DNA isoforms, and a single set of conditions that can be generalized to the production of reporters with any DNA lesion that can be incorporated into an oligonucleotide. As a result, the procedure requires fewer toxic chemicals and can be performed using standard molecular biology tools and reagents by most scientists with basic laboratory training. We furthermore have provided a streamlined protocol for the collection, analysis, and interpretation of FM-HCR data.

Applications of the method:

Despite the steadily falling cost and expanding breadth of analyses that can be conducted using modern ‘omics approaches, functional assays continue to fill an important niche in experimental biology. Functional assays are particularly valuable in the case of variants of unknown significance, and in scenarios where incomplete information precludes inferences of function from assays that report transcript levels, protein levels or localization, and post-translational modifications. FM-HCR assays have successfully been employed in immortalized human and rodent cell lines¹³, primary human fibroblasts ¹³ , primary human peripheral blood mononuclear cells ¹⁴, and patient derived xenograft (PDX) models of glioblastoma multiforme⁸. Recently, FM-HCR has successfully been implemented to determine how chromatin remodelling proteins affect DRC in the context of oxidative stress³³, to measure the DNA repair landscape in triple negative breast cancer cell lines³⁴, and to understand the role of DRC in the sodium sulphide-induced radiosensitivity in GBM and endothelial cell lines³⁵. MMR, as measured by FM-HCR, has successfully complemented next-generation sequencing (NGS) methods for assessing microsatellite stability/instability (MSS/MSI) in cancer cell lines and aided in the identification of novel synthetic lethal interactors ³⁶. Furthermore, FM-HCR assays have successfully detected transient changes in MMR capacity that occur as part of an adaptive mutator phenotype in response to combinatorial treatments with epidermal growth factor receptor (EGFR) and BRAF inhibitors in metastatic colorectal cancer cell lines models ³⁷. FM-HCR assays have untapped potential for use in classifying variants of unknown significance and expanding the scope of functional analyses in human populations. Since it is a fluorescent cell-based assay, FM-HCR also presents an opportunity to sort cells based on DNA repair phenotype. Although we have primarily used FM-HCR in flow cytometric assays, repair kinetics and single cell analyses can also potentially be carried out using dynamic imaging. Finally, our generalized methodology for preparing large quantities of plasmids with site-specific DNA lesions may provide a useful source of material for biochemical and structural studies that require them ^38-40.

Comparison with alternative methods

Non-HCR assays:

The absence of a single methodology for measuring the activity of multiple DNA repair pathways has been a major barrier to comprehensively analyzing DRC in large numbers of samples. Existing methodologies are often limited to interrogating a single repair pathway or type of damage and are usually too cumbersome to be applied to more than 5-10 samples ⁴¹. Examples include chromosomal double strand break (DSB) repair assays such as the EJ5-GFP and EJ2-GFP assays (along with other assays operating under similar principles) ²². These assays involve viral-mediated integration of a reporter construct into chromosomal DNA. Expression of the I-Sce-I homing endonuclease induces a site-specific DSB that gives rise to green fluorescent protein (GFP) expression when repaired by a specific end joining pathway. These assays are very powerful for mechanistic investigations but are more cumbersome than FM-HCR due to the need for viral integration of the reporter construct and cannot easily be used for population studies. Furthermore, the scope of the assay is limited to DSBs formed by the endonuclease I-Sce-I. Other approaches, such as comet assays, allow researchers to interrogate repair of a broader range of DNA lesions. Cells are treated with a DNA-damaging agent, lysed, and then subjected to single cell electrophoresis, which leads to “tailing” of damaged DNA that can be visualized by microscopy ^42,43. Traditional comet assays are laborious and low-throughput, making them inappropriate for large-scale studies; however, recent advances have greatly increased the throughput and reproducibility of comet assays ⁴⁴. Additionally, strategies have been devised for measuring repair of certain classes of DNA lesions based on their reactivity with different DNA repair enzymes, improving upon a major limitation of the conventional comet assay ^45,46. However, even with these recent advances, comet assays generally report on the repair of mixtures of DNA lesions and require several pathway-specific sample processing steps, precluding multiplexing of different repair pathways. In our experience, the time commitment for high-throughput comet assays is similar to that of FM-HCR, and we note that comet assay is conceptually simple and offers a powerful complementary approach to FM -HCR.

In order to measure repair of a specific DNA lesion, as opposed to the often heterogeneous mixtures generated by DNA-damaging agents, the lesion must be specifically incorporated into DNA. This is typically done by solid-phase oligonucleotide synthesis, but other methods exist such as incorporation of a modified deoxynucleotide triphosphate by a DNA polymerase. Repair of the DNA lesion can then be measured in a cell-free extract via different techniques. One method involves labeling the DNA with a fluorophore or a radioisotope and then visualizing repair by gel electrophoresis. Although straightforward, this method is typically limited to a single DNA lesion and is too cumbersome for large-scale studies. Molecular beacons are an adaptation of this approach that allows higher throughput by using fluorescence detection⁴⁷. A recently developed platform (termed LXRepair) employs damaged oligonucleotides in a microarray, allowing for high-throughput analysis of certain DNA repair pathways ^48,49. While powerful, this method does not cover all repair pathways and, like all DNA repair assays involving cell-free extracts, it lacks the biological context of an intact cell. Cell based assays using chemically defined DNA lesions exist for several DNA repair pathways, including an oligonucleotide retrieval method for NER⁵⁰, and a more recent variation for measuring BER⁵¹. Finally, most of the aforementioned assays are destructive and live cells cannot be monitored over the course of an experiment or recovered at the end of the analysis.

HCR assays

Host-cell reactivation technology predates the elucidation of the structure of DNA ⁵², yet still today it remains a powerful methodology for measuring and understanding the mechanisms of DNA repair. The basic principle, wherein DNA is damaged in vitro, and then transferred into host cells that repair the damage and restore the native structure to the DNA, has not changed. However, the methods for detecting this repair have advanced dramatically as reporter systems have become more sophisticated. While the earliest work was originally performed in E. coli, advances in cloning technology later enabled pioneering host-cell reactivation experiments in human cells ^16,53. This work laid the foundation for FM-HCR and established several key features that remain a centerpiece in the current assays. These features include the use of transient transfection, plasmid-based reporter systems expressing a foreign protein under a mammalian promoter, and detection of repair based on alterations in expression brought about by the presence of a DNA lesion. More recently, the relatively cumbersome chloramphenicol acetyltransferase reporters have been replaced with luciferase and FP reporters. The development of efficient methods for oligonucleotide synthesis in the 1980s and 1990s opened the door to site-specific incorporation of DNA lesions, in lieu of the randomly distributed DNA lesions induced by radiation or chemical treatments.

A key advance has been the development of strategies for detecting DNA lesion-induced transcriptional mutagenesis in cells ¹⁹. By overcoming the requirement that DNA lesions block transcription, this approach has vastly expanded the range of repair pathways that can be monitored by host cell reactivation assays to include those that change the sequence of the reporter plasmid ^21,28, or remove a DNA lesion that induces transcriptional errors ¹³. FM-HCR achieves multiplexing by utilizing several of the scores of different FPs that are available thanks to the Nobel Prize winning work of Roger Tsien⁵⁴. Transient transfection methods have also dramatically improved, with lipid mediated transfection and nucleofection enabling the use of FM-HCR in a wide variety of cell types, including primary cells.

Assays for generating site-specific DNA lesions in plasmids:

To take advantage of one of the major strengths of HCR assays, site-specific DNA lesions must be incorporated into plasmids. Early work by Essigmann and coworkers introduced methods for incorporating site-specific lesions into viral genomes ³². This approach continues to be used. For example, a fluorescent reporter assay for MUTYH activity was generated using an oligonucleotide with a site-specific 8-oxoguanine (8oxoG) that was annealed to single stranded phage DNA followed by polymerase extension and ligation to produce ccDNA ²¹. Variations of this approach have included using gel extraction or density gradient centrifugation to purify the final ccDNA from other molecular species³¹. Non-circularized intermediates and other DNA fragments can interfere with assays and reduce reproducibility by leading to errors in the estimated reporter plasmid DNA concentration. However, gel extraction can lead to shearing of DNA, severely limits yields, and risks carryover of agarose, DNA binding dyes, or reagents that are used to separate DNA from agarose. These contaminants may inhibit enzymes in subsequent preparation steps or interfere with the assay. Cesium chloride density gradient centrifugation has also been employed in the purification of plasmid DNA^16,55. This method is time intensive, requiring long centrifugations (> 24 hrs) at high speeds (100,000 g) that require use of an ultra-centrifuge. Notably, the protocol described here does not require the use of phage, gel extractions, or resin/column elution, and effectively eliminates any un-ligated intermediate species using enzymatic digests. The ability to generate highly purified synthetic DNA constructs (at near milligram scale) enables a broad range of experimental approaches that require large amounts of material.

Assays for assessing mismatch repair

A comparison of FM-HCR with other strategies for assessing MMR capacity highlights some of the strengths of the technique. Currently, there is a clinical demand for generating accurate and reproducible systems to classify cells according to their MMR status. NGS is an excellent means of classifying cells as MSI/MSS based on the sequence of repetitive microsatellite regions. However, this approach remains relatively expensive and time-consuming, and provides only an indirect qualitative assessment of MMR capacity. Similarly, PCR-based analysis of MSI status is qualitative, requires a reference sample, and involves a subjective assessment of peak structures observed upon electrophoretic analysis of PCR products. Although genomic signatures of microsatellite instability (MSI) are widely accepted surrogates for MMR deficiency, some MMR defects do not result in MSI, and MSI signatures remain intact in cells that restore MMR through reversion of inactivating MMR gene mutations or reversal of gene silencing due to promoter methylation. NGS also remains relatively expensive and requires several days of sample processing and data analysis. Immunohistochemistry may also be useful clinically for assessing the expression of MMR proteins, but this methodology lacks the ability to predict protein function, particularly in samples with variants of unknown significance, and it is highly dependent on sample quality.

These considerations motivate efforts to develop functional assays that can directly assess DRC. A recently reported biochemical assay from the research group of Christopher Pearson has demonstrated that the repair of a functional plasmid-based T:G mismatch using whole cell extracts from human cell lines and primary patient samples is capable of diagnosing constitutional MMR deficiency syndrome. For this system, MMR activity is estimated by recovering plasmid DNA from cell extracts, incubating the DNA with a restriction enzyme that only recognizes the sequence of the repaired plasmid, and quantifying the ratio of repaired versus unrepaired plasmid by analyzing the products of the digest using a gel-based assay⁵⁶. Another recent study has demonstrated that these powerful cell-free in vitro MMR activity (CIMRA) assays are capable of identifying MMR-related gene variants of core MMR proteins. This work has shown promise that CIMRA can be reproduced across several labs using standard laboratory equipment in only a few days⁵⁷. Similarly, FM-HCR has successfully been utilized to corroborate NGS^58,59 and traditional PCR-based classification of MSI/MSS status in colorectal cancers, but this assay can provide a quantitative assessment of MMR activity in living lymphocytes or cancer cells in as little as 18 hours⁶⁰. Although we have focused here on MMR, the same principles apply to assays for other DNA repair pathways.

FM-HCR assays for double strand break repair

We previously described FM-HCR reporter assays for NHEJ and HR ¹³. These assays utilize site-specific DSBs formed within reporter plasmids by restriction enzymes. Each reporter gives rise to FP expression only when the DSB is repaired by the pathway the assay was designed to measure. These assays have allowed us to interrogate DSB repair efficiency and pathway choice in large sets of samples; however, they are subject to important caveats. Like other FM-HCR reporters, these assays measure repair in the context of a plasmid instead of genomic DNA. Furthermore, DSBs formed by restriction enzymes are unlikely to be representative of physiological DSBs. However, FM-HCR assays for DSB repair can potentially be modified by treatment with DNA repair enzymes or by incorporation of modified nucleotides to more closely recapitulate the complex end chemistry of physiological DSBs, improving upon an important limitation of I-Sce-I based genomic DSB repair assays ²².

Limitations:

Sample Limitations:

FM-HCR assays require live cells that can be transiently transfected with plasmid DNA. Primary samples are sometimes difficult to transfect, but we have found that most cells can be transfected either by lipofection or electroporation. For instance, we routinely transfect peripheral blood mononuclear cells with FM-HCR plasmids using electroporation. Additionally, manufacturers such as Lonza and ThermoFisher describe methods for transfecting different primary samples. However, cells may occasionally be refractory to transfection with the available methods. Transient transfection induces an interferon response in cells⁶¹, and some cells are more difficult to transfect due the activation cGAS-STING signalling that diminishes transgene expression, but this may be overcome with inhibitors ⁶². Inefficient transfection can preclude analysis and may require extra steps, such as the use of cell surface markers, to confirm that the transfected cells are representative of the population of interest. Multiplexing requires transfection of multiple plasmids, and while we generally observe that most cells positive for one reporter are also positive for other co-transfected plasmids, this is not strictly necessary because reporter expression is averaged over the entire cell population. With regard to population studies, tissue banks often preserve materials by snap freezing, embedding in paraffin, and other methods that are not compatible with recovery of live cells and thus are not amenable to FM-HCR.

Technical limitations:

Although the starting plasmids for FM-HCR are readily expanded in E. coli, the reporters containing a DNA lesion must be prepared in vitro, since copying the damaged DNA would result in the daughter strand, and therefore the replicated plasmid, lacking the DNA lesion. This creates the need for the synthetic procedure we describe herein, but also restricts the acceptable methods of transfection. Because viral mediated transfection involves copying of the DNA, it is not compatible with FM-HCR. An additional technical limitation of FM-HCR arises from spectral overlap of the FPs expressed from the reporter plasmids. In general, unless permitted by flow cytometers with multiple lasers and excitation/emission narrow filter range, the typical multiplex FPs scheme is suitable for up to 5 colors. We have found that the relatively short stokes shifts of FPs impose gating and compensation requirements that limit the practical use of larger numbers of FPs in a single experiment. A 5-color experiment would be sufficient to assess the DRC of 4 DNA repair pathways/proteins, with one FP as a transfection control. One final but notable limitation of FM-HCR is the need to transduce repair of a single DNA lesion into a change in fluorescent reporter expression. Although the principle of the assay is relatively simple, either the DNA lesion blocks transcription or induces transcriptional mutagenesis, engineering reporters that convert these signals into a change in FP expression can be challenging. This is particularly true for transcriptional mutagenesis.

Biological limitations:

Fundamentally, by using synthetic plasmid-based reporters in living cells, there are also several biological limitations. Synthetic plasmids are limited in evaluating DNA repair in a chromatin context. Previous reports have described a histone deposition into closed circular plasmid DNA⁶³, and we have demonstrated that some FM-HCR substrates may be sensitive to alterations in chromatin remodeller expression³³. However, these effects could be due to changes in DNA repair gene expression rather than direct chromatin remodelling effects. Valuable insights into the role of chromatin context and remodelling in DNA repair could potentially be gained by introducing a nucleosome positioning sequence into FM-HCR reporter plasmids.

Experimental design:

Reporter construct design

A wide variety of mammalian expression constructs are compatible with FM-HCR, but at a minimum, the reporter construct must include a fluorescent reporter gene, a mammalian promoter sequence, a bacterial origin of replication, and a bacterial antibiotic resistance gene. For the majority of our FM-HCR assays, we have used the pMax Cloning vector (Lonza Cat #VDC-1040) as the base vector for our FM-HCR reporter plasmids. This vector includes a cytomegalovirus (CMV) promoter, a pUC origin of replication, and a kanamycin resistance cassette. This plasmid does not have an origin such as SV40 or EBV and therefore does not replicate in mammalian cells. We have subcloned several fluorescent reporter genes into this vector, and also engineered a series of point mutations that express non-fluorescent mutant proteins in order to generate reporters that use transcriptional mutagenesis to report repair of DNA lesions that do not block RNA polymerase. The relatively small size of the pMax backbone (2.8 kb) makes it easier to overexpress in E. coli and transfect into mammalian cells. The vector also includes an intron that may enhance gene expression ⁶⁴, as well as the Kozak consensus sequence to maximize translation efficiency ⁶⁵. Taken together, these features help to maximize the chances of detectable reporter expression in cells that are difficult to transfect, and generally reduce the amount of material required by increasing signal.

Incorporation of DNA lesions

Because the introduction of randomly distributed damage, such as those induced by radiation or chemical treatments are described in detail elsewhere, this protocol is focused on the introduction of site-specific DNA lesions using a primer extension methodology that we have recently refined (Figure 3a). The concentration and purity of the starting materials: ccDNA, ssDNA, and oligonucleotide should be determined spectrophotometrically prior to starting this procedure.

Key steps of the procedure

The protocol begins with a strand-specific digest of the ccDNA starting material with the appropriate enzyme for nicking either the transcribed strand or the non-transcribed strand (Figure 3b), depending on the reporter (Table 1 and Table 2). Although the digest will generally be complete within 4 hours, we observe comparable yields following overnight digests, which help to ensure that no closed circular DNA is carried into subsequent steps. If present, unmodified closed circular plasmid DNA will survive all of the remaining steps in the protocol and can thus interfere with assays. At several points in the protocol, purification of DNA must be carried out with ultrapure phenol chloroform isoamyl alcohol in order to avoid oxidative damage to DNA. Subsequent digest of the cleaved strand in the resulting open circular double stranded plasmid DNA (ocDNA) with Exo III is more time-sensitive, as we find that single stranded DNA can be degraded upon extended incubation with commercial Exo III. Reaction products should be tested for possible incomplete digest by agarose gel electrophoresis following each enzymatic digest step (Figure 3c). During oligonucleotide annealing, we recommend a molar ratio (ssDNA:oligo) of at least 1:4, as lower ratios often result in unbound ssDNA, decreasing the final product yield (Figure 3d). Annealing must be carried out using a slow-cooling setup wherein the reaction is cooled from approximately 85°C to room temperature (approximately 20°C) over the course of several hours. Although this can be achieved using a thermal cycler, we have found a simple 3L water bath prepared in a large Pyrex beaker on a magnetic stirring hot plate to be effective and convenient, especially for larger reaction volumes. When making up the reaction for primer extension, it is important to incubate the mixture of all components excluding the enzymes at 37°C for approximately 15 minutes to minimize any secondary structures that may distort the helix due to improper oligonucleotide binding. We have found that omission of this step may lead to a higher proportion of unligated ocDNA. The enzymes are then added last, after mixing all other ingredients to avoid potential inactivation of ligase and polymerase by high local concentrations of reaction components. Dimethyl sulfoxide (DMSO) is also added to relax secondary structures that can otherwise impede T4 polymerase leading to incomplete extension. Although we recommend overnight incubation of the extension/ligation reaction at 37°C, this step is largely complete within 4 hours. PEG is included to increase the stability of the enzymes and affords a convenient desalting strategy at the end of the protocol. The final cleanup of the reaction using T5 Exo (37°C for 3 hours) and then Proteinase K (55°C for 1 hour) are critical to maximize purity and yield (Figure 3e). T5 Exo removes all DNA species other than ccDNA (the desired product), and proteinase K removes residual enzymes that could degrade the product or interfere with downstream assays.

Table 1.

Summary of established FM-HCR reporter plasmids; required starting materials, and methods for quality control and validation.

Reporter Plasmid	Repair Pathway (Protein)	Reporter Protein	Substrate Starting Plasmid	Control Plasmid
mOrange_GG	MMR	mOrange	pMax_mOrange_G299C	pMax_mOrange
mOrange_8oxoG-C	BER (OGG1)	mOrange	pMax_mOrange A215C	pMax_mOrange
mPlum_A-8oxoG	BER (MUTYH)	mPlum	pMax_mPlum	pMax_mPlum
mPlum_O6MeG	DR (MGMT)	mPlum	pMax_mPlum C207G T208C	pMax_mPlum
GFP_Hx	BER (MPG)	eGFP	pMax_GFP C289T	pMax_GFP
GFP_THF	BER (APE1)	eGFP	pMax_GFP	pMax_GFP
BFP_U	BER (UDG)	BFP	pMax_BFP A191G	pMax_BFP
BFP_NHEJ	NHEJ	BFP	pMax_ScaI_BFP	pMax_ScaI_BFP
mOrange_UV	NER	mOrange	pMax_mOrange	pMax_mOrange
Δ5GFP-StuI-Linear + Δ3GFP	HR	GFP	pCX-NNX-Δ5GFP	pCX-NNX-GFP

Table 2.

Reporter plasmids with site-specific DNA lesions and the starting materials, enzymes and oligonucleotides required for generating them. The position of the site-specific DNA lesion is indicated in red text. 8oxodG refers to 8-oxoguanine, O⁶MeG refers to O⁶-methylguanine, idexyl refers to deoxyinosine, and idSp refers to a tetrahydrofuran spacer.

Reporter Plasmid	Starting Plasmid	Nicking Enzyme (Nicking Strand)	Primer Sequence
mOrange_GG	mOrange G299C	Nt.BsPQI (+)	GCTTCAAGTGGGAGCGCGTGATGAACTTCG
mOrange_8oxoG-C	mOrange A215C	Nb.BtsI (−)	GTAGGCCTTGGAGCCG [8oxodG] AGGTGAACTGA GG
mPlum_A-8oxoG	mPlum WT	Nt.BsPQI (+)	GGACATCCTGTCCCCTCAGATCATG [8oxodG] AC GGCTCCAAGGCCTACGTG
mPlum_O6MeG	mPlum C207G T208C	Nb.BtsI (−)	CACGTAGGCCTTGG [Me-G] CCCGTACATGATCTG
GFP_Hx	GFP C289T	Nb.BtsI (−)	GAAGAAGATGGTGC [idexyl] CTCCTGGACGTAGCC
GFP_THF	GFP WT	Nb.BtsI (−)	GCTCAGGGCG [idSp] ACT GGGTGCTCAGGTAGTG
BFP_U	BFP A191G	Nb.BtsI (−)	GGTCTTGCTGCCGUACAGGAAGCTAGTAGC

Reporter construct quality control

Final purity and yield should be determined by gel electrophoresis and spectrophotometric analysis. In addition to running a DNA ladder for estimating the size of plasmids, we strongly recommend including closed circular, open circular, and linearized plasmid controls in each agarose gel. The overwhelming majority of the final product should be ccDNA as other species will be undetectable in a well-executed preparation. A total yield of approximately 10-20%, relative to the starting amount of homogenously pure closed circular plasmid DNA, is commonly expected. The presence of the desired DNA lesion must be confirmed through methods specific to the DNA lesion as described in Table 1. Some DNA lesions can be confirmed by incubating the reporter plasmid product in vitro with purified enzymes that process the lesion of interest. For example, Fpg specifically generates single strand breaks at the site of oxidative damage, and thus converts closed circular reporter modified with a single site specific 8oxoG into open circular DNA that can readily be detected by agarose gel electrophoresis. In other cases, such as O⁶-methylguanine (O⁶-MeG), the lesion is incorporated into a restriction recognition site (PspOMI) that is blocked when the lesion is present. Thus, the presence of the lesion is confirmed by the absence of linear plasmid DNA upon agarose gel electrophoretic analysis of the restriction digest products. Because DNA lesions may not completely block restriction digest, excess enzyme may still yield some linearized DNA. For this reason, we recommend optimizing this digest using unmodified plasmid by determining the minimum number of units of enzyme activity needed to digest the desired amount of ccDNA (typically 500 ng) in one hour at 37°C. For reporters with mismatches, bidirectional Sanger sequencing confirms the sequence of the mismatch, and provides a limited measure of sequence purity ⁶⁶.

Transient transfection of cells

FM-HCR analysis requires transfection of both undamaged and damaged plasmid “cocktails”. The damaged plasmid cocktail contains the desired lesion-bearing plasmids along with one undamaged reference plasmid that is a transfection control. The undamaged cocktail (Figure 1A) contains the same transfection control plasmid along with the wild type (WT) damage-free counterpart of each plasmid used in the damaged cocktail (Figure 1B). Importantly, each cocktail needs to conform to the following specifications: 1) the undamaged and the damaged cocktails must contain the same WT FPs for normalization and quantification purposes, 2) the transfection volume for each cocktail should be equal, 3) the quantity of DNA to be transfected should not result in an abundance of off-scale fluorescent events after setting flow cytometer laser voltages, and 4) “carrier plasmid” may be used to enhance the transfection efficiency, but should be added such that the total amount of DNA transfected into cells from each cocktail is the same. Carrier plasmid refers to any plasmid that does not encode a FP. The purpose of the carrier plasmid is to facilitate the use of smaller amounts of DNA that might be adsorbed to reaction vessel surfaces or otherwise sequestered so as to be unable to enter cells during transfection. In the case of electroporation, we have found that, especially for small amounts of DNA (less than 50 ng), including carrier DNA leads to higher transfection efficiency that manifests as both a higher percentage of fluorescent positive cells and brighter FP expression. In the case of lipid mediated transfection, a fixed ratio of DNA:lipid is required. We have observed that transfecting small amounts of lipid-DNA complexes reduces transfection efficiency (as expected); however, the cells that do transfect tend to be brightly fluorescent. By contrast, diluting reporter plasmids in carrier DNA allows one to keep both the lipid:DNA ratio constant and FP expression on scale during flow cytometric analysis.

Multiplexed transfections:

Here we describe an FM-HCR assay for MMR, BER of hypoxanthine (Hx), and DR of O⁶-MeG. Although FM-HCR can include at least 5 FPs, gating and compensation become more technically challenging as the number of fluors increases, and we therefore favour multiple four-color experiments when sample quantity is not a limiting factor. However, the principles described throughout this protocol can be extended to larger numbers of FPs. We describe the composition of cocktails of plasmids for electroporation of TK6 and TK6+MGMT cells cultured in a 12-well plate. The first plasmid cocktail contains only undamaged plasmids: pMax_GFP, pMax_mPlum, pMax_mOrange, and a transfection control plasmid pMax_BFP. The damaged cocktail mix also contains undamaged pMax_BFP, which serves as the control for transfection efficiency, as well as mOrange_GG, which reports repair of a G:G mismatch, GFP_Hx which reports excision of Hx, and mPlum_O⁶MeG, which reports repair of O⁶MeG:C. In this experiment, 1 microgram of total DNA is transfected per well. To conserve materials and to ensure that fluorescence intensity remains on scale, we dilute the reporter plasmids using carrier DNA. For the “undamaged” reporter plasmid cocktail, each transfection requires 50 ng pMax_BFP, 50 ng pMax_mOrange, 50 ng pMax_GFP, 50 ng of pMax mPlum and 800 ng of carrier DNA. For the “damaged” reporter plasmid cocktail, each transfection requires 50 ng pMax_BFP, 50 ng mOrange_GG, 50 ng GFP_Hx, 50 ng mPlum_O⁶MeG, and 800 ng of carrier DNA. For our experiments, we anticipate needing to perform 9 transfections. So, each cocktail is prepared with enough material for 9 + 1 extra for a total of 10 measurements. This comes to 500ng pMax_BFP, 500 ng pMax_mOrange (or mOrange_GG), 500 ng of pMax mPlum (or mPlum_O⁶MeG), 500 ng of pMax GFP (or GFP_Hx) and 8 micrograms of carrier DNA. As a final step, if needed we adjust the total volume of the cocktails by adding Tris-EDTA Buffer Solution (TE) to a final volume of 15 μl (totalling 10 transfections, 1.5 μl each). Although equal cocktail volumes are not strictly necessary for lipid-mediated transfection, it reduces the risk of mistakes by ensuring the pipetting volumes are the same for both cocktails. Because equal volumes are important when transfecting cells by electroporation, we recommend adjusting cocktail volumes with TE buffer as a general best practice.

Introduction of FM-HCR reporter plasmids into the cells can be achieved by several means. The choice of transfection method is based on the type of cells to be analyzed. Lipid mediated transfection is efficient for many cell lines and primary adherent cells, such as skin fibroblasts ⁶⁷. Electroporation may be required for cell types that are more difficult to transfect, such as primary lymphocytes, and immortalized cell lines that grow in suspension, such as lymphoblastoid cell lines. The manufacturers of commercial transfection reagents and electroporators generally publish specific guidelines for transfecting individual cell types. To limit costs, many types of cells can also be transfected using inexpensive and widely available reagents, such as calcium phosphate or DEAE/Dextran ^16,68. We recommend optimizing transfection efficiency for difficult-to-transfect cells to minimize the amount of FM-HCR reporter plasmid consumed. We note that FM-HCR assays remain accurate under conditions of low transfection efficiency, particularly for homogenous cell cultures ¹⁴. In cases where there are mixed populations of cells that may have different repair capacities, transfection conditions could in principle affect the assay, especially if one subpopulation were more difficult to transfect than the other. FM-HCR analysis can be restricted to a subpopulation of cells using fluorescent surface markers that are specific to the cell type of interest in conjunction with FM-HCR. The excitation and emission properties of the fluorescent surface markers must be compatible with the fluorescent reporter proteins. This can be checked using online spectral viewers, or by seeking technical support from the manufacturer of the fluorescent surface markers.

Training and instrument requirements for flow cytometric analysis

Analysis of cellular fluorescence by flow cytometry provides the raw data needed for calculating repair capacity in FM-HCR assays. We have successfully conducted the assays on a variety of flow cytometry platforms. It is critical that the flow cytometer is compatible with the fluorescent reporters to be used. In order to take full advantage of multiplexing, an instrument with multiple excitation lasers and detection filters is required. Unlike many flow cytometric analyses that rely solely on the percentage of positive cells, calculation of repair capacity with FM-HCR also requires the mean fluorescence intensity of the positive cells. Although we provide substantial detail in the procedure section below concerning our recommended setup for flow cytometry worksheets, gating, and compensation, for those who are not experienced with flow cytometry, we recommend working with a core facility, and familiarizing oneself with the fundamentals of flow cytometry. A particularly useful resource is Practical Flow Cytometry (4th edition) by Howard Shapiro ⁶⁹. Users should also become familiar with the key steps for performing FM-HCR assays, which are described in more detail below and in Extended Figures 1-3. These include setting up a gating workspace to detect cells; setting up the gates for detecting FP expression; setting photomultiplier tube (PMT) voltages; using single color controls to set compensation; and setting up a gating hierarchy for excluding false positives that may arise due to spectral overlap that cannot be eliminated with compensation.

Quantitation of repair capacity

FM-HCR relies on a “ratio of ratios” for robust and reproducible calculation of DRC. The two ratios come from two transfections, one with undamaged and one with damaged cocktails (see Transient Transfection of Cells). The first ratio arises from normalizing expression of each reporter to the expression of an undamaged plasmid that is included as a transfection control in every plasmid cocktail. Repair capacity is then calculated from a second ratio: normalized fluorescent reporter expression from the damaged plasmid divided by normalized expression of the corresponding undamaged plasmid. This system affords estimates of DRC that are robust to variations in transfection efficiency. Although the computations are relatively simple (Equations 1-3) and can be semi-automated using a spreadsheet or a statistical package, we recommend spot-checking results to make sure the correct fields in the source file are being used to calculate repair capacity.

Cell line controls

Although quality control steps confirming DNA purity and the presence of the lesion are key steps, the gold standard for validation of FM-HCR reporters is the use of cell lines with defined DNA repair defects to confirm that the reporter plasmid is working as expected. We provide a list of recommended positive and negative controls in Table 3. We also present typical results for QC, analytical digests, and cell-based transfection validation in the Extended Data (Extended Figures 1-3). For isogenic cell line controls for each respective reporter, we provide the expected % reporter expression range (Calculated according to Equations 1-3). We also present an alternative representation of data as the relative reporter expression among positive and negative cell line controls. It is imperative to follow all QC, analytical digests, and validate % reporter expression levels in cells to have confidence that each reporter plasmid performs as expected.

Table 3.

Reporter plasmids with site-specific DNA lesions and the cell lines and quality control measures required for validating them.

Reporter Plasmid	Positive control cell line	Negative control cell line	Quality Control Measures
mOrange_GG	TK6	MT1, HCT116	Sequencing. Cell line verification (Extended Figure 3b)
mOrange_8oxoG-C	WT MEF	OGG1−/− MEF	Analytical digest: 250 ng plasmid with Fpg endonuclease (8 U) for 45 min in 37°C water bath (Extended Figure 1c)
mPlum_A-8oxoG	WT HAP	MUTYH−/− HAP	Analytical digest: 250 ng plasmid with Fpg endonuclease (8 U) for 45 min 37°C water bath (Extended Figure 1b)
mPlum_O6MeG	TK6+MGMT	TK6	Analytical digest: 250 ng plasmid with PspOMI RE (20 U) for 45 min 37°C water bath (Extended Figure 2a)
GFP_Hx	WT HAP	MPG−/− HAP	Analytical digest: 250 ng plasmid with ApaLI RE (10 U) for 45 min 37°C water bath (Extended Figure 1a).
GFP_THF	CH12F3	Ape1−/−/− CH12F3	Analytical digest with APE1.
BFP_U	WT MEF	UNG−/− MEF	Analytical digest: 250 ng plasmid with UDG (5 U) for 5 min 37°C water bath, followed by APE1 (10 U) for 30 min 37°C water bath. (Extended Figure 2b).
BFP_NHEJ	TK6	PKcs−/− TK6	Agarose Gel. Confirm linearization by upward mobility shift.
mOrange_UV	TK6 XP2OS + XPA	XPA−/− TK6, XP2OS	T4 PDG digest. Cell line verification.
Δ5GFP-StuI-Linear + Δ3GFP	U2OS UWB1.289+BRCA1	U2OS + RAD51 siRNA KD, UWB1.289	Agarose Gel. Confirm linearization by upward mobility shift.

Customization:

The procedures described in this protocol for generating DNA damage-containing constructs can be modified and adapted to accommodate a wide range of workflows and applications. All materials needed for this protocol are commercially available, including oligonucleotides containing site-specific lesions, but the procedure can also be readily adapted to the use of custom synthetic oligonucleotides and enzymes expressed and purified in house. When customizing this procedure to generate novel lesion-containing reporters, ensure that the DNA lesion is stable to the reaction conditions. In general, lesions that are labile to heat or moderately alkaline conditions may present challenges. Although we have not tested our reaction conditions using more than 1 mg of ssDNA, the procedure can in principle be carried out on any scale if the appropriate molar quantity of single-stranded DNA and oligonucleotide are available. Within each stage, the protocol can be paused at numerous steps (where indicated) at the discretion of the user. Furthermore, screening points utilizing gel electrophoresis to monitor progression of the protocol are included. When preparing a reporter for the first time, we recommend running a small pilot reaction to confirm all steps are successful before scaling up.

Example and generalization of the procedure:

This protocol describes the synthesis and validation of an FM-HCR reporter plasmids using a site-specific G:G mismatch containing reporter as an example that can be generalized to the other reporters with site-specific DNA lesions in Table 1 and Table 2. Extensive troubleshooting recommendations are organized by protocol step number in Table 4. The starting material for generating this reporter is pMax_mOrange_G299C, which expresses a non-fluorescent mutant mOrange protein, and has been described previously¹³. If the G:G mismatch is repaired, this plasmid expresses WT mOrange FP. As discussed above (Quantitation of Repair Capacity), FM-HCR involves calculating ratios of FP expression relative to undamaged controls. In our present example, the undamaged (lesion-free) control plasmid expresses the wild type FP (mOrange) in an otherwise identical vector. The transfection control plasmid must express a different FP that can be measured at the same time as the FH-HCR reporter protein. In our example, this will be a blue fluorescent protein expressing plasmid, pMax_BFP.

Table 4.

Troubleshooting tips for potential problems.

Step	Problem	Possible Reason	Solution
STAGE 1
1-2	No colonies observed after transformation.	Not enough plasmid DNA used, non-viable E.coli stock, or used incorrect selection antibiotic.	Appropriately dilute plasmid DNA according to manufacturer recommendations, use fresh E.coli stock, and use correct selection reagent.
3	Plasmid dimer/trimer observed via agarose gel electrophoresis (little/no monomer observed).	Extended incubation time, incorrect media/media volume used, incubated at wrong temperature, the starting plasmid material was incorrect size, utilized incompatible shaking speed for E.coli strain used.	Follow manufacturer recommendation for culture conditions for the E.coli strains used, and transform E.coli with confirmed monomeric plasmid DNA.
4	Incorrect sequence detected.	Sequenced the wrong strand or used incorrect sequencing primer.	If site-directed mutagenesis was performed, confirm primers and expected product. Ensure to sequence + and − strands with appropriate primer sets.
11	Residual ccDNA is observed after digest with nicking enzyme (as in Figure 3c, Lane 3)	Incomplete digest due to use of incorrect buffer, loss of enzyme activity in storage, failure to add enough enzyme, and/or insufficient incubation time.	Confirm assay conditions and obtain fresh enzyme. Incomplete nicking can be confirmed by digesting 10 μg of the nicked plasmid with 100 U T5 Exo in 1X NEB Buffer 4 for 2 hours at 37°C. This degrades all open circular DNA, making it easier to visualize residual closed circular DNA on an agarose gel. (i.e. repeat steps 10 & 11).
12-21 (Phenolchloroform extraction)	Phenol-chloroform contamination carry-over when measured by spectrophotometer.	Disruption of the top aqueous layer with the bottom organic layer. This may occur when carrying the tube from centrifuge to fumehood or while pipetting up the aqueous layer.	Carefully remove tube from centrifuge, carefully transport tube without mixing aqueous and organic layers, and carefully pipette off designated aqueous volume without disrupting organic layer. If contamination occurs, re-centrifuge tube to re-settle layers and try again carefully.
12-21 (Phenolchloroform extraction)	Unexpected coloration of the organic phase or diminished yield or integrity of DNA	Expired or impure phenolchloroform solution	Use fresh ultrapure molecular biology grade reagent
27	Residual ocDNA is observed after digest with ExoIII (as in Figure 3c, Lane 5)	Incomplete digest due to use of incorrect buffer, loss of enzyme activity in storage, failure to add enough enzyme, and/or insufficient incubation time.	Confirm assay conditions and obtain fresh enzyme. If using smaller or larger plasmids, incubation times may need to be reduced or extended (to be determined experimentally).
29	If the ssDNA concentration is lower than 1000 ng/μL, the reaction volumes in the subsequent annealing and extension steps will not accommodate the larger volume of ssDNA	Too much TE volume added, resulting in too low concentration of ssDNA.	Adjust the volumes in steps 26 and 28 as detailed in the troubleshooting notes below. The minimum ssDNA concentration that is compatible with subsequent steps in the protocol is approximately 500 ng/μL. If the concentration is less than 500 ng/μL, the DNA should be concentrated by ethanol precipitation followed by dissolution in a smaller volume.
32	More than 12 μL of NEBuffer 2.1 is needed to achieve 1X concentration	If the ssDNA concentration is less than 1000 ng/μL, the total volume in this step will exceed 120 μL	The volume of 10X NEBuffer 2.1 should be increased to ensure a final concentration of 1X, and the volume of water should be adjusted to maintain a 120 μl volume.
34	More than 48 μL of NEBuffer 2.1 is needed to achieve 1X concentration	The annealing volume exceeded 120 μL either because of low ssDNA concentration or a miscalculation	To determine the volume of buffer to be added, subtract the volume of NEBuffer 2.1 added in step 32 from 60 μl (as the final volume for the Primer Extension should be 600 μl; yielding a final working concentration of 1X NEBuffer 2.1.
41	Residual ssDNA after oligonucleotide extension as in Figure 3d, Lanes 5 and 6.	Incomplete extension due to use of insufficient amount of oligonucleotide, high oligonucleotide Tm, and/or insufficient time allowed for annealing step	Ensure that the recommended 4:1 molar ratio of oligonucleotide:ssDNA is used. For oligonucleotides that are highly destabilizing, such as those modified with helix distorting bulky DNA adducts, it may be helpful to increase the molar ratio to 10:1, or to increase the length of the oligonucleotide.
41	Poor ligation efficiency observed after extension.	Use of a non 5’-phosphorylated oligo, secondary structure formation in the oligo, use of incorrect ssDNA strand.	Use 5’ phosphorylated oligos (through vendor or in house), do not select oligos with repetitive sequences or oligos which may self-anneal. Ensure the correct ssDNA strand is used during annealing. SDS-PAGE purified oligos are recommended. We have also observed that for some oligonucleotides, the conversion of ocDNA to ccDNA is superior if the extension step is initiated by addition of polymerase to samples pre-equilibrated to 37°C (not shown).
STAGE 2
45	ssDNA and/or intermediates persist.	Incomplete/partial ssDNA digestion of intermediates.	If using smaller or larger plasmids, incubation times may need to be reduced or extended (to be determined experimentally). Add 1/10th of the amount of T5 Exo originally used in Step 42, mix the reaction by pipetting, and incubate for another hour. Analyze a 17 μL aliquot on a 1% agarose gel as described in step 10 to confirm that only ccDNA is detectable.
46	Low yield in subsequent steps	Accidental use of NEBuffer 2.1 (contains BSA)	Be sure to use NEBuffer 2.0. If present, BSA may inhibit proteinase K resulting in incomplete inactivation of T5 exonuclease
50	Unable to see precipitated pellet at bottom of tube.	Small amount of DNA. Fixed rotor centrifuges do not concentrate all DNA at the bottom of the tube, and plasmid DNA may spread out to the wall of the tube.	For smaller-scale preparations, it is easier to visualize the pellet in microfuge tubes. For 15 mL conical tubes, a swinging bucket centrifuge is recommended because it ensures the pellet will be easier to visualize as it is concentrated at the bottom of the tube. The pellet may be glassy and difficult to visualize at this step but typically becomes more visible after 70% ethanol wash.
69	Uncertain status of reporter plasmids after storage	Plasmids have been through multiple freeze-thaw cycles or were stored for more than 3 months at 4 °C.	We have successfully used reporter plasmids that have been stored at 4 °C for over 2 years; however, we observe slow nicking of the DNA backbone at this temperature. We recommend consuming aliquots of plasmid within 6 months at 4 °C or else analyzing by gel electrophoresis to confirm the integrity of the DNA.
Stage 3
70	Substantial fraction of final product is ocDNA, as in Figure 3e, Lane 5.	Insufficient proteinase K was used in step 46, and/or contaminating nucleases are present	In the final steps of the reporter plasmid preparation, proteinase K is used to digest residual protein from previous steps. We found that this step is critical for recovering pure ccDNA in subsequent precipitation steps. If proteinase K is excluded, a substantial fraction of the DNA becomes nicked over time in storage at 4°C.
72	Incomplete analytical digest	Enzyme used for in vitro AD is inactivated or the incubation period was too short. This can also indicate that the desired lesion is not present in the reporter plasmid.	Obtain fresh enzyme and confirm reaction conditions match those in supplementary figures and accompanying tables. Ensure correct enzymes were utilized are compatible for AD of desired substrate.
74	Uncertain whether training is sufficient for running flow cytometric analysis	Limited previous experience	General procedures common to all flow cytometric analyses are described in textbooks such as Practical Flow Cytometry (4th edition) by Howard Shapiro, and the details for carrying them out may vary depending on the instrument and software package used for analysis. We recommend first-time users obtain hands-on training from an experienced user or flow cytometry core
75	Unsure how to identify the main population	Multiple populations are evident in the scatter plot	Cell debris will generally appear to the upper left of the main living cell population, while doublets and higher order cellular aggregates will be located to the upper right portion of the scatter plot at a lower density than that of the main population of interest.
78	Uncertain how many fluorescence scatter plots are needed	Using a combination of fluorescent proteins different from the one described	We considered detection of the mOrange FP when multiplexing with BFP, GFP, and mPlum because we use this combination frequently, and mOrange provides an opportunity to illustrate our recommended strategy for dealing with spectral overlap between multiple FPs. The approach here can be generalized to any number N of fluorescent proteins by making an NxN grid of each FP against each other.
78	The diagonal plots of FP do not seem to contain useful information	The plots have been included as placeholders that facilitate downstream analysis	No action is needed; the diagonal plots (each FP versus itself) are not needed, and the plots on either side of the diagonal contain equivalent information. However, we recommend keeping all plots, as this visually intuitive spatial arrangement facilitates the process of establishing gating and setting compensation.
81	Linear range of PMT voltages is not known	Fluorescent beads are not available or making this determination is not practically feasible	This step is recommended because quantitation with FM-HCR assumes a linear relationship between the mean fluorescence intensity of the cells and the amount of FP present in the cells (See Equations 1-3). Failing to keep the instrument within the linear PMT range could lead to systematic errors in estimating absolute FP expression, but would not likely preclude comparisons between samples as long as they are analyzed using the same instrument settings.
84	Brightly fluorescent cells are off-scale	Too much plasmid has been transfected into cells	Although lowering the voltage is a common practice for bringing brightly fluorescent cells on scale, whenever possible we recommend reducing the amount of transfected reporter plasmid to conserve materials and minimize the need to revisit gating and compensation between experiments.
84	Single color controls are only weakly transfected	Transfection conditions were not optimal for the cells of interest	If feasible, increase the number of cells and/or the amount of plasmid used in the transfection. Then, optimize transfection conditions and use the maximum possible amount of the single color control plasmid. Refer to manufacturer recommendations regarding optimization of transfection conditions. During optimization, we recommend working with a fixed number of cells, and optimizing the absolute number of fluorescent positive cells. Optimizing the percentage of positive cells occurs at the expense of cell survival, diminishing the amount of data obtained from the experiment. Note, if cells in samples to be analyzed by FM-HCR are brighter than the brightest cells in the single-color control, compensation may not be accurate.
85	Uncertain why only some colors require compensation	User has limited experience working with fluorescent proteins and/or flow cytometry	Some colors don’t require compensation because the excitation and emission spectra do not overlap. For example, the excitation maxima of BFP and mOrange are separated by over 100 nm. The need for compensation between GFP and mOrange arises because there is more spectral overlap between FP with similar emission wavelengths. Online spectral viewers provide a useful guide for minimizing spectral overlap when designing experiments.
85	Uncertain whether compensation has been applied correctly	User is unable to tell visually whether the ideal scenario in Figure 5q-ff has been achieved	Correct compensation can be confirmed quantitatively by adjusting compensation so that the GFP Mean Fluorescence Intensity (MFI) in the GFP channel is the same for cells expressing high levels of mOrange and cells that are no brighter than the unstained control in the mOrange channel.
88	Autocompensation yields oddly shaped or counterintuitive fluorescence scatter plots	Cells with off-scale fluorescence do not respond to compensation	Some software packages for analysis of flow cytometry data perform autocompensation, however we recommend setting compensation manually to avoid pitfalls associated with autocompensation that can interfere with the accuracy of FM-HCR data analysis.
89	After applying compensation, cells adopt a funnel shape with some cells taking negative fluorescence values	Extensive (25% or more) compensation is required between the two FP	Note that compensation is applied on a cell-by-cell basis. Although the same percentage of signal is subtracted, the uncertainty in fluorescence measurements is proportional to fluorescence intensity. For FPs requiring extensive (more than 25%) compensation, and especially for very brightly fluorescent cells, the uncertainty in the number that is being subtracted can be larger than the mean fluorescence intensity of the unstained control, which we set to approximately 100. As a result, for some cells, too much signal is subtracted, resulting in negative values, and for some cells, too little signal is subtracted, resulting in values that are still above background. Overall, this phenomenon manifests as a symmetrical funnel shape in the fluorescence scatter plot. It also creates the necessity for wedge shaped gates that accommodate the funnel shape to avoid counting compensated cells as false positives.
89	Compensation results in negative fluorescence values at intermediate fluorescence intensity	Brightly fluorescent cells are off-scale	Transfect a smaller amount of the plasmid that expresses the reporter protein that is resulting in off-scale fluorescence.
93	Square gates are not possible because cells adopt a funnel shape in fluorescence scatter plots	Extensive spectral overlap between two fluors	Generate wedge shaped gates to accommodate the funnel shape to avoid counting compensated cells as false positives.
96	Single color dropout controls (aka “minus one”) still register false positive events in the fluorescent channel of interest	Autofluorescence and noise can result in some false positives even when gating and compensation have been performed correctly. Autofluorescence appears as broad-spectrum fluorescence (detectable in all channels) and often reflects the presence of dead cells in the analysis	Autofluorescence due to dead cells can be excluded from flow cytometric analysis using a dead cell stain such as TO-PRO-3 (Thermo Fisher Cat. no. T3605). For some experiments, it may not be feasible to reach zero false positive events without making the gating so stringent as to result in an unacceptably high number of false negatives. In these cases, we recommend keeping the false positive percentage at least 10-fold lower than the lowest observed percentage of positive cells for that FP in any experimental sample. For example, if at least 1% of cells are GFP positive in all experimental samples, then a 0.1% false positive rate is tolerable in the GFP dropout control.

MATERIALS

REAGENTS:

CRITICAL Wherever available, molecular biology grade reagents should be used.

Plasmid encoding a FP of interest under a mammalian promoter

5’ phosphorylated, HPLC/PAGE purified oligonucleotide modified with DNA lesion of interest. Typical lengths range from 20-40 nucleotides

Human CMV immediate early promoter, forward primer [5’-CGCAAATGGGCGGTAGGCGTG-3’]

Human CMV immediate early promoter, reverse primer [5’-GCAATAGCATCACAAATTTCACA-3’]

1 kb DNA ladder (NEB, cat. no. N3232S)

Agarose LE (Cole-Parmer, cat. no. VV-00141-03)

Appropriate single strand nicking enzyme:

• Nt.BspQI (NEB, cat. no. RO644S)

• Nb.BtsI (NEB, cat. no. R0707S),

• Alternative restriction enzyme which targets only one strand of the DNA

DNase/RNase free microcentrifuge tubes

• 1.8 mL (VWR, cat. no. 87003-294)

• 5 mL (Genesee Scientific, cat. no. 86-902)

Nucleotides:

• ATP (NEB, cat. no. P0756L)

• dNTP mix (NEB, cat. no. N0447L)

DNA Modifying Enzymes:

• T4 DNA polymerase (NEB, cat. no. M0203L)

• T4 DNA ligase (NEB, cat. no. M0202L).

Enzymatic buffers:

• CutSmart™ Buffer, (NEB, cat. no. B7204S),

• NEBuffer™ 1 (NEB, cat. no. B7001S)

• NEBuffer™ 2 (NEB, cat. no. B7002S)

• NEBuffer™2.1 (NEB, cat. no. B7202S)

• NEBuffer™ 3.1 (NEB, cat. no. B7203S)

• NEBuffer™4 (NEB, cat. no. B7004S).

Exonucleases:

• Exonuclease III (NEB, cat. no. M0206L)

• T5 Exonuclease (NEB, cat. no. M0363L)

Dimethyl sulfoxide (DMSO; VWR, cat. no. BDH1115-1LP)

Ethanol, 200 proof, molecular biology grade (Fisher Scientific, cat. no. BP2818100) “CAUTION” Flammable. Wear protective clothes and eyewear

Orange G gel loading dye- (0.2% Orange G, 30% Glycerol)

Isopropyl alcohol (VWR, cat. no. VWRV0918-4L) “CAUTION” Flammable. Wear protective clothes and eyewear

LB Agar Kanamycin Plates (Sigma-Aldrich cat. no. L0543)

Lipofectamine 3000 (Invitrogen, cat. no. L300000)

Luria Broth (LB) base Miller (VWR, cat. no. 90003-252)

Max efficiency DH5α competent cells (Invitrogen, cat. no. 18258012)

Monarch Plasmid Miniprep Kit (NEB, cat. no. T1010S)

Nalgene PPCO centrifuge bottles (Thermo Scientific, cat. no. 3120-9500PK)

Nuclease-free water diethylpyrocarbonate (DEPC)-treated water (Invitrogen, cat. no. AM9916)

Polyethylene glycol (PEG)-8000 (Millipore Sigma, cat. no. 89510-250G-F)

Proteinase K, molecular biology grade (NEB, cat. no. P8107S)

PureLink endotoxin-free giga plasmid purification kit (Invitrogen, cat. no. A31233)

Sodium acetate buffer solution (3M, pH 5.2, 0.2μm filtered; Millipore Sigma, cat. no. S7899-100 mL)

Sodium chloride (NaCl)- Dnase, Rnase, and protease free (Millipore Sigma, cat. no. S3014-1KG)

15 mL sterile conical centrifuge tubes (Thermo Scientific, Nunc™ cat. no. 339651)

50 mL sterile conical centrifuge tubes (Thermo Scientific, Nunc™ cat. no. 339652)

Tris-EDTA buffer solution (TE, pH8.0; Invitrogen, cat. no. 12090015)

Tris-acetate 10X (TAE, pH 8.6; VWR, cat. no. 10128-406)

UltraPure phenol:chloroform:isoamyl alcohol (25:24:1, v/v) (Invitrogen, cat. no. 15593031). CRITICAL: High-purity reagent is necessary to avoid oxidative damage to the DNA.

• “CAUTION” Toxic, corrosive, can result in reproductive harm and/or cancer development. Avoid contact with skin, as will cause severe skin irritation and burning. Only use with appropriate eye and hand protection in a fume hood which has been inspected and maintained in working order

REAGENT SETUP:

PEG filter and precipitation buffer – Mix PEG 8000 (15%), NaCl (1.1M), and NEBuffer 2 (1X) in deionized water. CRITICAL Mix thoroughly and pass through a 0.45μm filter. Ensure that the PEG 8000 solution is made fresh for each prep. Precipitation Buffer should only be prepared on the day it will be used.

Orange G loading buffer- prepare 0.2% Orange G by mass, 30% Glycerol by volume in deionized water

70% Ethanol – prepare in deionized water

EQUIPMENT:

Hot Plate Stirrer (VWR, Dyla,Dual cat. no. 12620-970)

3 L beaker (VWR, cat. no. 13912-320)

Flow Cytometer (Core Facility-dependent selection)

NxT Attune™ Flow Cytometer (Invitrogen) or comparable multi-laser flow cytometer

Gel Imaging Systems

ChemiDoc™ (Bio-Rad, cat. No. 12003153) or iBright FL1500 (Invitrogen, A44241), or comparable gel documentation platform

Owl easycast B1A mini gel electrophoresis system (Thermo Scientific, cat. no. B1A)

PowerPac™ basic power supply (Bio-RAD, cat. no. 1645050)

Neon™ transfection system (Invitrogen, cat. no. MPK5000), Gene pulser MXCELL™ (Bio-RAD) or comparable alternative electroporation unit.

SOFTWARE:

Attune NxT Software Version 2.7

FloJo

PROCEDURE:

CRITICAL All steps and centrifugations are carried out at room temperature unless otherwise indicated.

Stage 1: Plasmid generation: TIMING Approximately 2-3 days depending on sequencing turnaround and verification of FP expression in cells.

1. Transform pMax_mOrange_G299C into competent DH5-α E. coli and spread the transformed bacteria on LB agar supplemented with kanamycin.

2. Pick 2 colonies for 6 mL liquid culture in antibiotic supplemented LB media and carry out plasmid minipreps according to manufacturer’s instructions.

3. Confirm the plasmid size and isoform via gel electrophoresis (Figure 3b, lane 2).

4. Submit plasmid for Sanger sequencing using the CMV-F sequencing primer to confirm the desired construct, and expand additional colonies if the correct sequence is not obtained.

5. Transiently transfect the plasmid into cells of interest using the desired transfection method.

6. Confirm expression of the expected FP (or lack thereof for non-fluorescent protein expressing mutant plasmids) 24 hours after transfection using flow cytometry. CRITICAL STEP: Validating the size, sequence and reporter expression is necessary before proceeding to relatively resource-intensive large scale plasmid purification.

7. Scale up plasmid purification to allow for sufficient starting material for planned reporter plasmid preparation; this typically includes performing a milligram-scale preparation by using the Giga Plasmid purification Kit (Invitrogen, cat. no. A31233). CRITICAL STEP: The product of this preparation must also be confirmed by following steps 3-6. Before generating a site-specific reporter, a 5’-phosphorylated synthetic oligonucleotide must be obtained (Table 1).

Plasmid nicking preparation: TIMING Approximately 20 minutes active time + 8-16 hours incubation time.

• 8.
  Digest 3 mg pMax_mOrange_G299C with nicking enzyme in a 15 mL conical tube as follows:

  • Start with 3 mg of pMax_mOrange_G299C (600 ng/μl)
  • Add 500 μl 10X NEBuffer 3.1 (1X)
  • Add 150 μl Nt.BspQI nicking enzyme (1500 U enzyme:3 mg ccDNA)
  • Add DEPC water to a total reaction volume of 5 mL
  • Gently pipette up and down to ensure thorough mixing and collect by brief centrifugation at 500g if needed.
• 9.Incubate the nicking reaction overnight (8-16 hours) in a 37°C water bath.
• 10.
  Confirm the nicking digest (Figure 3b, lane 3) by analyzing 1 μl of the digested material on a small (~50 mL) 1% agarose gel prepared with 1X TAE (pH 8.6) by following the instructions below. As controls, we recommend including an equivalent amount (600 ng) of closed circular starting material and 5 μL of a 1kb ladder (NEB, cat. no. N3232S).

  • Treat samples with proteinase K (1 U) for 15 minutes at room temperature before loading to remove bound protein, which can alter the migration of DNA.
  • Add 1/10^th volume of 10X Orange G loading dye to each sample (or 1/6^th volume of 6X NEB loading buffer). Mix thoroughly by pipetting, and load samples onto the gel. Run the gel at no more than 120V until the dye front reaches the end of the gel for optimal resolution.
• 11.If the nicking reaction is confirmed to have reached 100%, save a 30 μl aliquot at 4°C for downstream QC, and proceed to the phenol-chloroform extraction and ethanol precipitation in the next step. If there is evidence of ccDNA remaining (Figure 3c, lane 3), centrifuge the nicking reaction for 1 min (2000 g) to ensure the contents are collected at the bottom of the conical tube. Add an additional 20 μl of the appropriate enzyme (Nt.BspQI), mix thoroughly by pipetting, and incubate at 37°C in the water bath for an additional hour. Repeat as needed until 100% nicking is achieved (Figure 3c, lane 4).

Phenol-chloroform extraction: TIMING Approximately 1 hour

• 12.Add 5 mL (one volume) of UltraPure phenol:chloroform:isoamyl alcohol to the 15 mL conical tube in which the nicking reaction was carried out. CAUTION – this step involves a hazardous material. Consult the manufacturer’s SDS to determine appropriate PPE and complete the following steps in a fume hood.
• 13.Ensure that the tube is tightly closed. Vortex vigorously for 15 seconds (solution will turn milky white). CAUTION Additional precautions, such as the use of a face shield or fume hood sash, should be taken while vortexing a tube containing phenol-chloroform.
• 14.Centrifuge for 10 minutes (5000 g). A swinging bucket centrifuge is preferred as this will minimize disruption of the phase separation after centrifugation. CRITICAL STEP take extreme care to maintain distinct phase separation layers when removing from centrifuge and transporting to the fume hood.
• 15.Carefully pipette 4.5 mL from the aqueous phase (top layer) and transfer to a new 15 mL conical tube without disturbing the bottom layer or organic interphase layer.
• 16.OPTIONAL STEP: To increase yield, add 1 volume of TE to the remainder of the material in the tube from step 15, repeat steps 12-15, and combine the aqueous layer with the previously collected aqueous layer in the 15 mL conical tube from step 15.
• 17.Ethanol precipitate the nicked DNA obtained from the phenol-chloroform extraction by adding 1/10^th volume (500 μL) of 3M sodium acetate (pH 5.2) and then 3 volumes (15 mL) of 200 proof/100% ethanol. CRITICAL STEP: If optional step 16 was carried out, these volumes should be doubled.
• 18.Centrifuge for 30 minutes (At least 5000 g at 4°C).
• 19.White DNA pellet should be visible at the bottom/side of the conical tube, carefully decant supernatant.
• 20.Add 5 mL 70% ethanol, and centrifuge for 10 minutes (5,000 g at 4°C). CRITICAL STEP Carefully vacuum aspirate off the 70% ethanol using a sterile glass aspirator without dislodging or disturbing the DNA pellet.
• 21.Dissolve the pellet in 500 μl TE for 10 minutes at room temperature. It may be beneficial to pipette gently to dislodge DNA that may be stuck to the wall of the conical, however be careful to avoid touching the sticky pellet with the pipette tip, as this can result in sample loss. PAUSE POINT: Dissolved DNA can be stored at 4°C for several days if needed.

Single stranded DNA preparation: TIMING Approximately 4 hours

• 22.Measure and record the yield and the A₂₆₀/A₂₈₀ for the nicking reaction. Typical yields are 80% or better, and A₂₆₀/A₂₈₀ should be at least 1.8.
• 23.Determine the number of units (U) of Exo III needed for the digest. We use 5U for each microgram of DNA. For example, a 2 mg digest (2,000 micrograms) requires 10,000 U.
• 24.
  Prepare the digest in a 15mL conical tube as described in the following example (for generating ssDNA from 2 mg of ocDNA in a 2 mL reaction):

  • Add 1,000 μL nicked DNA (initial concentration 2.0 μg/μL, final concentration 1.0 μg/μL)
  • Add 200 μL 10X NEBuffer 1 (1X)
  • Add 100 μL Exo III (initial concentration 100,000 U per milliliter; final concentration 5U/μg)
  • Add 200 μL DEPC-treated water
  • Mix thoroughly.
• 25.Aliquot into 2X 1.8 mL microcentrifuge tubes, or incubate the entire reaction in a 15 mL conical tube
• 26.Incubate at 37°C in a water bath for 1.5 hours.
• 27.After the incubation, aliquot 10-15 μL of the reaction and analyze on a 1% agarose gel as described above in step 9 to assess completion of digest. Place the reaction tube(s) at 4°C while the gel is running.
• 28.If nicked DNA is present in the gel analysis (Figure 3c, lane 5), incubate the single strand reaction tubes at 37°C for 20 minutes with an additional 1/10^th of the original amount of Exo III (in this example we would add an additional 10 μL) and repeat the gel analysis in step 17 to confirm that only single stranded DNA product is present (Figure 3c, lane 6).
• 29.Once no nicked DNA remains, save a 5 μL aliquot at 4°C for downstream QC. Bring the volume up to 5 mL by adding 3 mL TE buffer and proceed to phenol-chloroform extraction and ethanol precipitation as described in steps 12-21 CRITICAL STEP: After dissolving the ssDNA pellet, the final concentration should be at least 1000 ng/μL. For every mg of nicked DNA, the expected yield of ssDNA is approximately 400 μg. If the yield of the nicking reaction was lower than expected, reduce the volume of TE used to dissolve the pellet. PAUSE POINT: Dissolved DNA can be stored at 4°C for several days if needed.

Oligonucleotide annealing and extension: TIMING Approximately 20 minutes to set up reaction + 3 passive hours for annealing, and approximately 20 active minutes and 16 passive hours for extension.

• 30.Determine the molecular weight (MW) of the ssDNA (1.1 MDa for pMax_mOrange_G299C) and calculate the number of picomoles (pmol) of plasmid to be annealed. We recommend annealing 100 micrograms of ssDNA:

For 100 μg of pMax_mOrange_G299C ssDNA, use the following calculation: 100 x 10⁻⁶ g / 1.1 x 10⁶ g/mol = 9.1 x 10⁻¹¹ mol, or 91 pmol.

• 31.Calculate the number of pmol of oligonucleotide to be annealed by multiplying the pmol of plasmid by 4. For 91 pmol (100 μg) of pMax_mOrange_G299C number of pmol of oligonucleotide to be annealed = 364 pmol.
• 32.In a 1.8 mL microcentrifuge tube, prepare an annealing reaction between the custom 5’-phosporylated oligo and the ssDNA, using a 4:1 molar ratio of phosphorylated oligo to single stranded DNA with 1X NEBuffer 2.1 as shown in the example below:
  Example: To anneal phosphorylated oligo to 100 μg (~91 pmol) ssDNA

  • Add 50 μL single stranded DNA (initial concentration 2000 ng/μL; final concentration 833 ng/μL)
  • Add 12 μL 10X NEBuffer 2.1 (1X)
  • Add 3.6 μL 100μM custom oligonucleotide (360 pmol; 16 μM)
  • Add 54 μL DEPC-treated water (up to a total volume of 120 μL)
  • Mix thoroughly by pipetting.
• 33.Heat 1.6 L of water in a 2 L beaker to 85°C on a hot-stir plate with moderate stirring. Turn off the heat and float the tubes in the heated water. Allow the reaction to cool to room temperature (~20°C, takes approximately 3 hours) in the water.
• 34.
  Prepare the primer extension reaction. The extension reaction is typically carried out with 100 μg of ssDNA per tube, but this can be scaled up or down as required. CRITICAL Add in sequential order (as noted below). Final concentrations are given in parentheses:

  • Start with 120 μL of the annealing reaction (167 ng/μL)
  • Add 60 μL 10mM ATP (1 mM)
  • Add 60 μL 10mM dNTP (1 mM)
  • Add 48 μL 10X NEBuffer 2.1 (1X)
  • Add 150 μL 30% PEG-8000, sterile filtered (7.5%)
  • Add 30 μL neat DMSO (4%)
  • Add 113 μL DEPC-treated water
  • Mix thoroughly by pipetting
  • Incubate at 37°C for 15 minutes to equilibrate reaction for “hot start” conditions
  • Add 16.6 μL 3 U/μL T4 DNA polymerase (50 U; 0.5 U /μg ssDNA)
  • Add 2.5 μL 400 U/μL T4 DNA ligase (1000 U; 10 U /μg ssDNA)
  • After addition of the enzymes, the final volume is 600 μL.
• 35.Mix thoroughly by pipetting up and down and incubate the reaction at 37°C overnight (4-16 hours)
• 36.Heat 1.6L of water on the hot-stir plate (Step 33 to 85°C. Heat inactivate enzymes by floating the extension/ligation reactions in the water bath for 20 minutes, and then remove and allow to cool to room temperature.
• 37.Analyze a 10 μL aliquot of the overnight extension/ligation reaction on a 1% agarose gel with closed circular DNA as a control.
• 38.Centrifuge the remainder of the samples for 10 minutes (5,000 g) to precipitate denatured proteins.
• 39.Use a pipette to transfer the supernatant from each extension reaction into a fresh microcentrifuge tube.
• 40.Save a 10 μL aliquot at 4°C aliquot from each reaction for downstream QC.
• 41.Perform visual confirmation of ccDNA (Figure 3d, lane 7). Typically, aligation efficiency above 50% is observed, but this may vary depending on oligo design and annealed sequence context).\ Upon confirmation, continue to the enzymatic cleanup stage (Stage 2).

Stage 2 Enzymatic cleanup: TIMING Approximately 10 minutes to set up reaction and 5 hours passive time.

• 42.Add T5 Exo (2 U per μg of ssDNA) to the heat inactivated extension reactions as follows:

For example:

• Start with 740 μL of HIA overnight extension reaction (100 μg ssDNA)

• Add 20 μl T5 Exo (100 U/μL)

• Add 150 μL 10X NEBuffer 4 (1X)

• Add 590 μL DEPC water

• 1.5 mL total volume.

• 43.Mix thoroughly by pipetting.
• 44.Incubate for 3.5 hours at 37°C. Save 10 μL of the enzymatic cleanup reaction at 4°C for QC.
• 45.Analyze 10 μL of the reactions on a 1% agarose gel as described in step 10 and include the closed circular starting material as a control to confirm complete T5 Exo digest of nicked intermediates. If only closed circular DNA remains, proceed to the next step. Store remaining reaction tube at 4°C until gel confirmation of complete T5 digest.
• 46.To inactivate and digest T5 Exo along with other enzymes present, add 1 μL proteinase K per 50 μg starting ssDNA and incubate for 1 hour at 55°C. Cool to room temperature and proceed to desalting/precipitation steps.

Substrate precipitation and final cleanup: TIMING Approximately 4 hours

• 47.Prepare Precipitation Buffer: PEG 8000 (15%), NaCl (1.1M), and NEBuffer 2 (1X) (see reagent setup).
• 48.Determine the total volume of the proteinase K treated reaction and add the Precipitation buffer at a 1:1 v/v ratio. Mix thoroughly by pipetting.
• 49.Centrifuge for 30 minutes (12,000 g at 4°C).
• 50.Carefully remove the supernatant by pipetting. CRITICAL STEP- If residual supernatant is present, carefully use a pipette to remove the residual supernatant. If the pellet has become dislodged, centrifuge again for 1 min (12,000 g), and carefully remove remaining supernatant.
• 51.Wash the pellet with 1 mL 70% ethanol, and centrifuge for 10 minutes (12,000 g at 4°C) in a swinging bucket centrifuge.
• 52.Carefully aspirate off the ethanol without disturbing the pellet and dissolve the pellet at approximately 1 μg/μL, assuming a 50% yield relative to the amount of ssDNA starting material (100 μL TE buffer for this example where 100 μg of ssDNA was used, and a 100% yield would be 200 μg). Allow the pellet to dissolve in TE for 1 hour. CRITICAL STEP Avoid pipetting to solubilize the pellet as this can diminish yield through possible DNA shearing and adhesion to the pipette tip if contact is made.
• 53.Determine the DNA concentration by UV-VIS spectrophotometry. PAUSE POINT: Dissolved DNA can be stored for several days at 4°C.
• 54.Phenol-chloroform extraction and ethanol precipitation. Add 100 μL of phenol-chloroform, vortex for 10 s, and centrifuge for 10 min (16,000 g). Remove 80 μL of the upper aqueous phase (or as much as possible without contaminating with the lower organic phase) and transfer to a clean 1.8 mL microcentrifuge tube.
• 55.To the tube containing the organic phase (as well as residual aqueous phase not removed in step 54), add a volume of TE buffer equal to the volume removed in step 54 (80 μL). Vortex for 10 s and centrifuge for 10 min (16,000 g) at room temperature.
• 56.Remove 80 μL from the aqueous phase and combine with the aqueous phase from step 54 (160 μL combined volume).
• 57.Perform an ethanol precipitation as described in step 17. Briefly, add 20 μL of sodium acetate (3M, pH 5.2), 20 μL of DEPC-treated water, and 600 μL of absolute ethanol. Mix by vortexing for 10 s, centrifuge for 30 min (16000 g, 4°C), and remove the supernatant. Wash the pellet with 500 μL 70% ethanol and then dissolve the pellet in nuclease-free TE buffer, aiming for a concentration of 1 μg/μL.
• 58.Determine final DNA concentration and purity (A₂₆₀/A₂₈₀) using a UV-Vis spectrophotometer.

Preparing the FM-HCR reporter plasmid for NHEJ: TIMING approximately 4 hours

• 59.Generate the pMax_ScaI_BFP plasmid according to Stage 1 (Steps 1-6 above) and amplify by Maxi Prep or Giga Prep according to the desired scale.
• 60.Digest the plasmid with a restriction enzyme that recognizes a unique site in the 5’-UTR, such as ScaI-HF (New England Biolabs Catalog No.: R3122) by treating 100 μg of plasmid with 150 units of ScaI-HF in the provided CutSmart buffer in a reaction volume of 500 μL at 37 °C. for 1.5 hr.
• 61.Purify the digested plasmid by UltraPure phenol-chloroform extraction followed by ethanol precipitation (Steps 12-21 above).
• 62.Confirm complete linearization of the plasmid by 1% agarose gel electrophoresis (described in Steps 10-11). Include DNA ladder and closed circular controls.
• 63.Determine final DNA concentration and purity (A₂₆₀/A₂₈₀) using a UV-Vis spectrophotometer.

Preparing the FM-HCR reporter plasmid for HR: TIMING approximately 4 hours

• 64.Generate the pCX-NNX-Δ5GFP and pCX-NNX-Δ3GFP plasmids, originally described by Kiziltepe et al. ⁷⁰ and subsequently modified by Nagel et al. ¹³ according to Stage 1 (Steps 1-6 above) and amplify by Maxi Prep or Giga Prep according to the desired scale.
• 65.Digest 100 μg of the pCX-NNX-Δ5GFP plasmid with 150 units of StuI (New England Biolabs, Catalog No.: R0187) in the provided CutSmart buffer in a reaction volume of 500 μL at 37 °C. for 1.5 hr.
• 66.Confirm complete linearization of the plasmid by 1% agarose gel electrophoresis (described in Steps 10-11). Include DNA ladder and closed circular controls.
• 67.Purify the digested plasmid by phenol chloroform extraction followed by ethanol precipitation as described in steps 12-21.
• 68.Determine final DNA concentration and purity (A₂₆₀/A₂₈₀) using a UV-Vis spectrophotometer.

Plasmid storage:

• 69.Aliquot plasmids and maintain at −80 °C for long-term storage. Avoid repeated freeze-thaw cycles. Once an aliquot of reporter plasmid has been thawed, it can be stored at 4 °C for at least 6 months.

Stage 3 Substrate validation: TIMING Approximately 3 hours active time and 24 hours inactive time

• 70.Analyze the final substrate using a 1% agarose gel with appropriate controls and intermediates collected for QC (QC aliquots from steps 11, 29, 40, and 45 as well as steps 62 and 66 if applicable), as illustrated in Figure 3.
• 71.For the mismatch-containing reporter, confirm the presence of the site-specific modification by sequencing as described previously ¹³.
• 72.For other reporters where a chemical modification has been introduced, perform an analytical digest in which treatment with a purified enzyme(s) will alter the plasmid DNA sufficiently to result in a notable shift on an agarose gel. The recommended analytical digest to be used for each of the FM-HCR reporter plasmids is listed in Table 3.
• 73.Carry out functional validation of newly prepared reporter plasmids via transfection into cell lines that are either known to repair the damage efficiently (positive control) or known to have a defect in repair of the lesion (negative control), as described previously¹³. Cells used for validation of each of our substrates can be found in Table 3.

Controls and experimental setup:

Setting up the analysis workspace with scatter plots for detecting cells: TIMING Approximately 5 minutes unless starting from a previously generated template

CRITICAL See the Experimental Design section for information on the training required before starting this section.

• 74.Establish a preliminary scatter plot including all events and adjust instrument settings such that the main cell population is located within the log range of 10³ and 10⁵ along the y-axis (side scatter-area, SSC-A), and within the linear range of 200-500 along the x-axis (forward scatter-area, FSC-A).
• 75.Set a gate that selects for the main cell population of interest (see gate “Cells”-Figure 4a).
• 76.Set a second gate, which should be nested within the “Cells” gate and is the first “Single Cell-1” gate (Figure 4b), where cell doublets are excluded from analysis by plotting the “Cells” population within a new scatter plot of FSC-H vs. FSC-W.
• 77.Next, set a second “Single Cell-2” gate that is nested within the “Single Cell-1” gate by plotting all cells events in “Single Cell-1” in a plot of SSC-H versus SSC-W. Cells in the “Single Cell-2” gate (Figure 4c) will subsequently be used in all proceeding steps when generating fluorescence scatter plots (Figure 4d-4f).

Figure 4. Gating Scheme for Identifying Single Cells and Representative Scatter Plots of Samples Requiring Compensation.

a, Representative primary gate for initially identifying the main live cell population. b, Exclusion of doublets using a plot of the forward scatter pulse width versus forward scatter pulse height. c, Exclusion of doublets using a plot of the side scatter height versus side scatter pulse width versus side scatter pulse width d, All laser voltages are adjusted to center the background fluorescence of non-transfected cells at a mean fluorescence intensity of approximately 100. e, Representative scatter plot showing the signature associated with overcompensation (left panel) f, Representative scatter plot showing the signature associated with or under-compensation.

Setting up the workspace with fluorescence scatter plots for analyzing FP expression: TIMING Approximately 20 minutes unless starting from a previously generated template

• 78.Set up a grid of fluorescence scatter plots that are systematically organized such that each fluor is plotted against each other fluor (Figure 5).
• 79.Transform the scatter plots to the hyper log scale, which allows for visualization of cells with negative fluorescence values that can arise when applying compensation.
• 80.Organize the plots in ascending order according to their fluorescence emission maxima so that FPs that have the most spectral overlap appear next to each other in the grid.

Figure 5. Scheme for Setting Compensation in a Four-Color Experiment.

a-p, Representative grid plotting each fluorescent against the others in a four color experiment. Curved patterns in the scatter for each single color identifies improper compensation. q-ff, Representative plots in the same grid following application of compensation.

Setting PMT voltages: TIMING Approximately 10 minutes

• 81.Use fluorescent beads to determine the PMT voltages that fall within the linear response range for the instrument as previously described ⁷¹; for accurate quantitative analyses, PMT voltages outside this range should not be used.
• 82.Adjust the PMT voltages such that the unstained control (cells that are not expressing fluorescent proteins) are centered at approximately 10² in every fluorescent channel of interest as shown in Figure 4d. CRITICAL STEP: This step generally needs to be performed each time a new cell type is analyzed, but once set, it is essential to use the same PMT voltages for all samples analyzed in an experiment.

Setting compensation: TIMING Approximately 20 minutes

• 83.Run single color controls; in our example, this requires four individual transfections using reporter plasmids encoding BFP, GFP, mOrange, or mPlum.
• 84.Confirm that cells yield strong fluorescence, ideally spanning the entire range of detection.
• 85.
  Determine whether compensation is needed for each reporter by carefully inspecting each of the fluorescence scatter plots as described here for mOrange; the key information can be read directly off the plots in the row of panels i, j, k, and l in Figure 5.

  • Consider first the FP with the shortest emission wavelength, in this example by examining the plot of mOrange versus BFP (Figure 5i),
  • If, as in this case, the signal in the BFP channel appears to be completely independent of signal in the mOrange channel (mOrange signal spreads in a straight line along the mOrange axis, without advancing along the BFP axis), then no compensation is needed.
  • Next consider the FP with the second shortest emission wavelength, by examining the plot of mOrange versus GFP (Figure 5j).
  • If, as in this case, brightly fluorescent mOrange positive cells appear as false GFP positive cells, then compensation is needed. This is accomplished by subtracting a percentage of the mOrange fluorescent signal from the GFP fluorescent signal until fluorescence in the GFP channel is independent of fluorescence in the mOrange channel (Figure 5z).
  • Finally, consider the FP with the longest emission wavelength by examining the plot of mOrange versus mPlum (Figure 5l).
  • If, as in this case, brightly fluorescent mOrange positive cells appear as false mPlum positive cells, then compensation is needed. This is accomplished by subtracting a percentage of the mOrange fluorescent signal from the mPlum fluorescent signal until fluorescence in the mPlum channel is independent of fluorescence in the mOrange channel (Figure 5bb).
• 86.Repeat the previous step for each of the remaining FPs.
• 87.Review all fluorescence scatter plots to check whether too much signal has been subtracted (referred to as over-compensation) for any FP pair. In the case of over-compensation, the scatter profile will extend backwards along the x-axis, taking negative values for the brightest cells (Figure 4e).
• 88.Review all fluorescence scatter plots to check whether too little signal has been subtracted (under-compensation) in any cases. In this situation, the scatter profile extends toward the upper right-hand corner of the plot (Figure 4f).
• 89.CRITICAL: Check compensation by confirming a nearly linear scatter profile (or a symmetrical funnel profile for FP requiring extensive compensation) for each FP plotted against each other filter to avoid spectral leakage into another filter.

Setting up a gating hierarchy for detection of true fluorescent positive events: TIMING Approximately 20 minutes unless starting from a previously generated template

• 90.Identify all FPs that required compensation due to spectral overlap with the FP of interest.
• 91.Set a gate in each scatter plot of the FP of interest versus the overlapping FPs identified in the previous step.
• 92.Name the gates for the FP of interest following the example in Figure 6, where FPs of interest are plotted on the X-axis, and overlapping FPs are plotted on the Y-axis.
• 93.Use the single color control for each overlapping FP to adjust the gates drawn in the previous step to exclude potential false positives for the FP of interest.
• 94.Define a joint gate such that true positives for the FP of interest are required to appear in all of the gates that were set in the previous steps.
• 95.Repeat the above steps for the remaining FPs.
• 96.Use cells transfected with dropout controls (where the FP of interest is excluded) to confirm the absence of false positive events.

Figure 6. Gating Scheme for Excluding False Positive Fluorescent Events in FM-HCR Experiments.

A representative gating scheme is presented for the scenario in which there is significant spectral overlap between all four colors. As in Figure 5, the rows in this grid of scatter plots are organized according to the fluorescent protein that is reported on the X-axis, from lowest to highest excitation wavelength. The columns are correspondingly organized according to the excitation wavelength of the fluorescent protein that is plotted on the Y-axis. Gates in each column are named alphabetically (e.g. Blue-A, Blue-B, and Blue-C) from top to bottom. Cells must appear in all of the gates in a column to be counted as a true positive for the fluorescent protein of interest that is plotted on the X-axis in that column. Gates A, B, and C can be joined using “AND” logic, or by setting up a gating hierarchy wherein gate C is nested into gate B, and gate B is nested into gate A.

Quantitation: TIMING Approximately 15 minutes

• 97.
  Use Eqns 1-3 below to quantify percent reporter expression as illustrated here for an example where an mOrange FM-HCR reporter is used together with a BFP transfection control.

  • Normalize expression of the DNA damage containing mOrange reporter in cells transfected with the “damaged” plasmid cocktail to expression of the co-transfected undamaged BFP:

  X=mOrange Count x Mean mOrange intensity /BFP Count x Mean BFP intensity (Eqn 1)

• $$X=\frac{mOrangeCount\ast MeanmOrangeintensity}{BFPCount\ast MeanBFPintensity}$$ (Eqn 1)

• Expression of the unmodified mOrange reporter in the “undamaged” plasmid cocktail is normalized to expression of the co-transfected undamaged BFP:

  Y=mOrange Count x Mean mOrange intensity / BFP Count x Mean BFP intensity (Eqn 2)

• $$Y=\frac{mOrangeCount\ast MeanmOrangeintensity}{BFPCount\ast MeanBFPintensity}$$ (Eqn 2)

• Divide normalized mOrange expression in cells transfected with the damaged plasmid cocktail (X) by normalized mOrange expression in cells transfected with the undamaged plasmid cocktail, and multiply by 100 to obtain percent reporter expression (Z).

  Z=100x(X/Y) (Eqn 3)

$$Z=100\ast X∕Y$$ (Eqn 3)

Multiplexed analysis of DNA repair capacity using GFP_Hx, mPlum_O⁶MeG, and mOrange_GG: TIMING Approximately 2 hours + 24 hours passive time

• 98.Prepare the undamaged plasmid cocktail. Each transfection contains 50 ng pMax_GFP, 50 ng pMax_mOrange, 50 ng pMax_mPlum, 50 ng pMax_ScaI_BFP (transfection control plasmid), and 800 ng Δ3GFP (Carrier DNA; i.e. non-fluorescent protein expressing carrier plasmid).
• 99.Prepare the damaged plasmid cocktail. Each transfection contains 50 ng GFP_Hx, 50 ng mOrange_GG, 50 ng mPlum_O⁶MeG, and 50 ng pMax_ScaI-BFP, and 800 ng Δ3GFP (Carrier DNA).
• 100.Transfect cells by electroporation or lipid-mediated transfection.
• 101.Analyze by flow cytometry following procedures detailed below and illustrated in Figure 7 (derived from gating scheme described in Figure 6).

Figure 7. Representative Four Color FM-HCR Experiment with Three FM-HCR Reporters.

a, fluorescence scatter plots showing expression of BFP, GFP, mOrange and mPlum in TK6 cells transfected with mOrange_GG, GFP_Hx, and mPlum_O⁶MeG. b, fluorescence scatter plots showing expression of BFP, GFP, mOrange and mPlum in TK6+MGMT cells transfected with mOrange_GG, GFP_Hx, and mPlum_O⁶MeG. c, fluorescence scatter plots showing true positive cells that appear in the joint gates (A, B, and C) for each of the fluorescent reporters in TK6 cells. d, fluorescence scatter plots showing true positive cells that appear in the joint gates (A, B, and C) for each of the fluorescent reporters in TK6+MGMT cells. e, Similar expression the mOrange_GG MMR reporter. f, high mPlum_O⁶MeG reporter expression in MGMT deficient TK6 cells and low mPlum_O⁶MeG reporter expression in MGMT proficient TK6+MGMT cells. g, Similar expression of the GFP_Hx reporter.

TIMING:

See approximate time estimations in procedure above.

Stage 1

Plasmid generation – Approximately 2-3 days.

Plasmid nicking preparation – Approximately 16 hours

Phenol-chloroform extraction – Approximately 1 hour

Single stranded DNA preparation – Approximately 4 hours

Oligonucleotide annealing and extension – Approximately 1 day

Stage 2

Enzymatic cleanup – Approximately 6 hours

Substrate precipitation and final cleanup – Approximately 4 hours

Preparing the FM-HCR reporter plasmid for NHEJ – Approximately 4 hours

Preparing the FM-HCR reporter plasmid for HR – Approximately 4 hours

Stage 3

Substrate validation – Approximately 1 day

Controls, experimental setup, setting up flow cytometry workspaces, and calculating DRC – User-dependent

ANTICIPATED RESULTS:

Substrate preparation:

A well-executed reporter plasmid preparation should yield approximately 50 μg for every 250 μg of ccDNA starting material. The final product should be nearly 100% closed circular modified plasmid DNA. A typical agarose gel electrophoretic analysis of the material at each step of the protocol is presented in Figure 3b. The starting material should be mostly closed circular DNA (Figure 3b, lane 2), but a minor open circular band is expected. Following enzymatic nicking, there should be no detectable closed circular band, and a uniform open circular band is expected (Figure 3b, lane 3), although minor higher MW bands may appear if plasmid dimers were present in the starting material. Digest with Exo III should completely eliminate the ocDNA band. Circular ssDNA runs ahead of ccDNA (Figure 3b, lane 4). Following the annealing and extension steps, a dominant closed circular band should be evident, but will likely be accompanied by other bands (Figure 3d, lanes 5-7). The other bands include failed ligations (open circular band), bands that run at intermediate sizes and correspond to stalled extensions, and a smear that may reflect partial extensions. All except the closed circular band should be eliminated during the enzymatic cleanup step, and little or no open circular DNA should be present in the final preparation (Figure 3b, lane 5).

Flow cytometry and analysis:

Although results will vary widely depending on the cell type, transfection method, and cell culture conditions, we can provide rules of thumb for assessing whether an experiment has yielded satisfactory results. Transfection efficiency, measured as the percentage of single live cells that are fluorescent positive, should be at least 1% for each FP in the cells transfected with the “undamaged” cocktail, but in most cases will exceed 10% for optimized transfection conditions. For a properly calibrated experiment, it is more important that fluorescence remains on scale than to maximize the percentage of fluorescent cells. The percentage of live cells should be at least 10%, but in most cases will be 50% or more. Low percentages of live cells can result for cell types that are difficult to transfect and thus require harsher transfection conditions that are more toxic to cells. It is generally observed that reporter expression from the DNA damage containing reporter plasmids is lower than expression from their undamaged counterparts. We recommend identifying experimental conditions that result in at least 30 total fluorescent positive cells for each damaged reporter plasmid as a minimum threshold for robust results. In situations where transfection efficiency is low, or repair capacity is such that it yields very low percent reporter expression, it may be necessary to increase the number of cells analyzed or the amount of damaged reporter plasmid in the cocktail in order to meet this threshold. In all cases, we recommend collecting flow cytometric data for the entire volume of the transfected sample to maximize the number of fluorescent events that are used to calculate DRC. Reporter expression generally falls between 0 and 100% but can in some cases exceed 100% for one of several reasons. This is most likely to result from errors in the composition of reporter plasmid cocktails, but may also reflect changes in the efficiency of transcription resulting from DNA damage or processing of DNA damage⁷². We have previously found that multiplexing has little or no effect on the readouts of FM-HCR assays¹³, however we recommend confirming this whenever a new application is developed. For example, reporter expression from an MMR reporter should be the same whether an NER reporter plasmid was included in the assay or not. If this is not the case, it may reflect issues with gating and compensation. By following the approach in Figure 6, it should be possible to set gating such that a false positive rate of less than 0.1% is obtained when cells are transfected with a “minus one” cocktail (where the color of interest has been excluded).

Interpretation of FM-HCR results.

The range of typically observed values for % reporter expression from each reporter depends upon the integrated action of a large number of proteins. Although this is one of the strengths of functional assays, it also makes absolute determinations of “high” or “low” repair capacity elusive. A general range of the expected values for percent reporter expression can be gathered from positive and negative control cell lines (Table 1, and Extended Figures 1-3). However, we recommend applying FM-HCR as a comparative method; the effect of knocking out a gene of interest, treatment with a drug, or the health status of an individual from whom a cell sample is obtained should be compared against an appropriate control. The control might be untreated cells in the case of a drug, an isogenic wild type control in the case of a genetic model, or a healthy individual in the case of a disease susceptibility study. This comparative approach is similar in principle to RNAseq or proteomics approaches utilizing isotopic labeling, in that the unperturbed control should be analyzed in parallel with the samples of interest.

For reporters that make use of transcriptional mutagenesis to report repair of DNA lesions that do not block transcription, it is important to note that there is an inverse relationship between FP expression and repair capacity. This relationship arises because it is the presence of the DNA lesion, not the removal of the DNA lesion that gives rise to the fluorescent signal. Thus, the assays turn from “on” to “off” as repair proceeds. In principle, it would be possible to design reporters that turn from off to on, as is the case for assays that measure repair of transcription blocking lesions. However, this approach was not adopted because DNA lesions do not generally induce transcriptional mutagenesis with 100% efficiency. For some DNA lesions, the correct base is occasionally incorporated into transcripts, and as a result the reporter would not be 100% “off” in the absence of repair. For a DNA lesion that induces transcriptional mutagenesis with 50% efficiency, this would result in an assay with a dynamic range of only 2 (100/50). By using an “on-to-off” strategy for the same DNA lesion, the dynamic range can be quite large (50/X). The minimum value of X can be extremely small, because for cells that repair the DNA lesion with very high efficiency reporter expression will approach zero. However, a technical disadvantage of this approach is that larger amounts of reporter plasmid or larger numbers of transfected cells may be needed to obtain sufficient signal to distinguish between two samples with high DRC. Indeed, this consideration was an additional motivation for the development of the strategies presented herein for increasing the yield of our reporter plasmid preparations.

Generating novel FM-HCR reporter plasmids

Researchers may wish to use this methodology to develop new FM-HCR reporters, and here we provide some general recommendations for strategies that can be used. The fundamental requirement for FM-HCR assays is that a DNA lesion is incorporated into the plasmid and repair of that lesion in cells results in a change in FP expression. To study repair of transcription-blocking lesions, that lesion is simply placed into the transcribed strand (within the reporter gene or the adjacent 5’ and 3’ untranslated regions). In this case, cells produce fluorescent protein only if the lesion has been repaired. We are actively using this method to develop new reporter assays.

Alternative strategies are necessary when the desired DNA lesion does not block transcription. For DNA lesions that induce transcriptional mutagenesis, the lesion is incorporated at a position where ribonucleotide misincorporation during transcription changes the sequence in a manner that alters the function of the translated protein, thereby turning the reporter off or on. We previously used this approach to make a reporter for repair of 8oxoG ¹³, but in principle, this transcriptional mutagenesis strategy can be employed for any lesion that causes significant base misincorporation during transcription. When generating novel FM-HCR reporter plasmids, it is necessary to identify the sequence of a fluorescent gene wherein disruption of the coding sequence would result in loss of a fluorescent protein. Knowledge of the transcriptional mutagenesis induced by DNA lesions is required to design the oligonucleotide. DNA lesions that are not miscoding can also be studied using FM-HCR assays, but the approach is slightly different. In this case, the DNA lesion is placed into the transcribed strand in place of a correct nucleotide. For example, we have used this approach to generate a reporter assay for a T:G mismatch. Cytosine (in the transcribed strand) was replaced with T such that a 3’-ACC-5’ codon was changed to 3’-ACT-5’. Transcription of the unrepaired reporter plasmid results in a UGA stop codon in the reporter transcripts, abolishing fluorescent protein expression. Repair of T using the opposing dG as a template restores the wild type sequence (3’-ACC-5’ in the transcribed strand), which gives rise to the correct TGG codon in the RNA transcripts, turning fluorescence on. A caveat of this strategy is that MMR may recognize the mismatch and compete with the desired pathway. Finally, we have also incorporated modified oligonucleotides that can act as a chemical handle for post-synthetic modification of the plasmid. Obvious examples include alkyne and biotin substituents, but numerous post-synthetic modifications are possible.

Extended Data

Extended Figure 1.

Gel electrophoretic analysis and flow cytometric validation of GFP_Hx, mPlum_A-8oxoG, and mOrange_8oxoG-C reporter plasmids. a, Analytical digest and flow cytometric validation of GFP_Hx plasmid. (Lane 1) NEB 1kb MW Ladder (Lane 2) pMax_GFP_C289T ccDNA (Lane 3) HIA overnight extension reaction for GFP_Hx (Lane 4) pMax_GFP_C289T ccDNA (Lane 5) pMax_GFP_C289T after 45 min ApaLI digest at 37°C, which cleaves 2 restriction sites, resulting in 2 linear DNA fragments (Lane 6) GFP_Hx ccDNA after T5 Exo and PEG purification steps (Lane 7) GFP_Hx after 45 min ApaLI digest at 37°C, in which the Hx lesion blocks ApaLI cleavage of 1 restriction site, leaving a single linearized fragment. At right: Flow cytometric quantitation of % reporter expression and normalized relative reporter expression in WT HAP cells compared to MPG^−/− HAP cells. b, Analytical digest and flow cytometric validation of mPlum_A-8oxoG plasmid. (Lane 1) HIA overnight extension reaction for mPlum_A-8oxoG (Lane 2) pMax_mPlum ccDNA (Lane 3) pMax_mPlum after 45 min Fpg endonuclease digest at 37°C (Lane 4) mPlum_A-8oxoG after T5 Exo and PEG purification steps (Lane 5) mPlum_A-8oxoG after 45 min Fpg endonuclease digest at 37°C, resulting in plasmid nicking and upward mobility shift. At right: Flow cytometric quantitation of % reporter expression and normalized relative reporter expression in WT HAP cells compared to MUTYH^−/− HAP cells. c, Analytical digest and flow cytometric validation of mOrange_8oxoG-C plasmid. (Lane 1) NEB 1kb MW Ladder (Lane 2) pMax_mOrange_A215C ssDNA (Lane 3) pMax_mOrange_A215C ocDNA (Lane 4) pMax_mOrange_A215C ccDNA (Lane 5) pMax_mOrange_A215C after 45 min Fpg endonuclease digest at 37°C (Lane 6) mOrange_8oxoG-C ccDNA (Lane 7) mOrange_8oxoG-C after 45 min Fpg endonuclease digest at 37°C, which introduces a nick at the 8oxoG lesion, resulting in upward mobility shift. At Right: Flow cytometric quantitation of % reporter expression and normalized relative reporter expression in WT MEF cells compared to OGG1^−/− MEF cells. Error bars represent SEM from 3-4 biological replicates, differences of statistical significance (P < 0.05 *, P < 0.005 **, P < 0.005 ***, P < 0.0001 ****) were determined by unpaired two-tail t test.

Extended Figure 2.

Gel electrophoretic analysis and flow cytometric validation of mPlum_O⁶MeG and BFP_U reporter plasmids. a, Analytical digest and flow cytometric validation of mPlum_O⁶MeG plasmid. (Lane 1) NEB 1kb MW Ladder (Lane 2) pMax_mPlum_C207G/T208C ccDNA (Lane 3) pMax_mPlum_C207G/T208C after 45 min PspOMI digest at 37°C, resulting in linear pMax_mPlum_C207G/T208C starting plasmid (Lane 4) mPlum_O⁶-MeG plasmid after T5 Exo and PEG purification steps (Lane 5) mPlum_O⁶-MeG after 45 min PspOMI digest at 37°C, in which the O⁶ group on the guanine blocks linearization by PspOMI, leaving predominantly ccDNA product (Note: upon extended digest or when excess enzyme is present, some linearized DNA will result). At Right: Flow cytometric quantitation of % reporter expression and normalized relative reporter expression in MGMT deficient TK6 cells compared to TK6 cell complimented with stable MGMT expression. b, Analytical digest and flow cytometric validation of BFP_U plasmid. (Lane 1) NEB 1kb MW Ladder (Lane 2) pMax_BFP_A191G ccDNA (Lane 3) pMax_BFP_A191G ocDNA (Lane 4) pMax_BFP_A191G after 5 min UDG digest at 37°C, followed by 30 min APE1 digest at 37°C (Lane 5) BFP_U after 5 min UDG digest at 37°C, followed by 30 min APE1 digest at 37°C, resulting in UDG excising the incorporated Uracil, followed by APE1 nickicking the abasic site resulting in upward gel mobility shift (Lane 6) BFP_U plasmid after T5 Exo and PEG purification steps. At Right: Flow cytometric quantitation of % reporter expression and normalized relative reporter expression in WT MEF cells compared to UNG^−/− MEF cells. Error bars represent SEM from 3-4 biological replicates, differences of statistical significance (P < 0.05 *, P < 0.005 **, P < 0.005 ***) were determined by unpaired two-tail t test.

Extended Figure 3.

Gel electrophoretic analysis and flow cytometric validation of mOrange_GG plasmid. a, Gel electrophoretic analysis of mOrange_GG plasmid. (Lane 1) NEB 1kb MW Ladder (Lane 2) PMax_mOrange_G299C ccDNA (Lane 3) HIA overnight extension reaction for mOrange_GG (Lane 4) mOrange_GG after 3 hour digest with T5 Exo (Lane 5) mOrange_GG after PEG precipitation step (Lane 6) Final mOrange_GG ccDNA after T5 Exo and PEG purification steps. b, Flow cytometric quantitation of % reporter expression and normalized relative reporter expression in TK6 cells compared to MMR deficient MT1 lymphoblastoid cells. Error bars represent SEM from 3-4 biological replicates, differences of statistical significance (P < 0.005 **, P < 0.005 ***) were determined by unpaired two-tail t test.

Abbreviations:

(8oxoG)

8-oxo-Guanine

(BER)

Base Excision Repair

(BFP)

Blue Fluorescent Protein

(ccDNA)

closed circular plasmid DNA

(CIMRA)

Cell-free In vitro MMR Activity

(CMV)

Cytomegalovirus

(DMSO)

Dimethylsulfoxide

(DRC)

DNA Repair Capacity

(EGFR)

Epidermal Growth Factor Receptor

(Exo III)

Exonuclease III

(FM-HCR)

Fluorescent Multiplexed Host-Cell Reactivation

(FP)

Fluorescent Protein

(FSC-A)

Forward Scatter-Area

(GFP)

Green Fluorescent Protein

(HR)

Homologous Recombination

(Hx)

Hypoxanthine

(MGMT)

Methylguanine-DNA Methyltransferase

(MMR)

Mismatch Repair

(MSI)

Microsatellite Instability

(MSS)

Microsatellite Stable

(NER)

Nucleotide Excision Repair

(NGS)

Next Generation Sequencing

(NHEJ)

Non-Homologous End Joining

(MFI)

Mean Fluorescent Intensity

(MW)

Molecular Weight

(O⁶-MeG)

O⁶-Methylguanine

(ocDNA)

Open Circular Double Stranded Plasmid DNA

(PDX)

Patient Derived Xenographs

(PEG)

Polyethylene Glycol

(pmol)

Picomoles

(PMT)

Photomultiplier Tube

(QC)

Quality Control

(ssDNA)

Single Stranded Closed Circular Plasmid DNA

(SSC-A)

Side Scatter-Area

(T5 Exo)

T5 Exonuclease

(TE)

Tris-EDTA Buffer Solution

(WT)

Wild Type

Data Availability:

Source data are provided with this paper. All other data supporting the approach described in this protocol are available from the corresponding authors upon reasonable request. Starting plasmids will be deposited in Adgene. Small amounts of prepared FM-HCR reporter plasmids can be shared for pilot and feasibility studies upon reasonable request.