Functional assays for BRCA1 and BRCA2 variant analysis: new tools and their applications

Abstract

Genetic variants perturbing BRCA1 and BRCA2 function are linked to hereditary breast and ovarian cancers. As BRCA1 and BRCA2 are critical for homologous recombination-mediated DNA repair, interpretation of variants of uncertain significance is essential. Here, we review established and emerging assays used to functionally characterize BRCA1 and BRCA2 variants and highlight the strengths and limitations of each approach, as well as their general relevance to variant classification and clinical applications.

Introduction

The BRCA1 and BRCA2 genes are characterized by a high degree of genetic variants, some of which are clinically significant and predispose to cancers or occur sporadically within somatic tissues. Pathogenic variants in either gene are strongly associated with elevated lifetime risks of developing particularly breast, ovarian, prostate, and pancreatic cancers^1–8. In clinical practice, genetic testing laboratories interpret BRCA1 and BRCA2 variants based on the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines⁹. These guidelines have been further refined by the Clinical Genome Resource (ClinGen) Evidence-based Network for the Interpretation of Germline Mutant Alleles (ENIGMA) BRCA1 and BRCA2 Variant Curation Expert Panel (VCEP) to provide standardized guidance for the clinical interpretation of BRCA1 and BRCA2 variants^10–12. According to the ClinGen ENIGMA BRCA1 and BRCA2 VCEP, population frequency, computational predictions, functional assays, and various case-level data are applied at calibrated strengths to achieve accurate classification of BRCA1 and BRCA2 variants^13–15. Since many observed variants are rare, informative case-level data are often unavailable or limited, making variant classification heavily dependent on population frequency, computational predictions, and functional assays^16–18. The lack of case-level data often leads to the classification of genetic variants as variants of uncertain significance (VUS), creating a major barrier to clinical decision-making. This uncertainty leaves patients, families, and clinicians with inconclusive and non-actionable results. To address this challenge, it is essential to incorporate complementary sources of evidence. Among these, functional assays have emerged as valuable tools for supporting accurate classification of genetic variants.

Functional assays assess whether a variant impacts the protein’s ability to perform its biological roles relevant to cancer. For BRCA1 and BRCA2, this primarily involves evaluating their contribution to DNA double-strand break (DSB) repair by homologous recombination (HR)^19–22, which is considered their main tumor suppressive function. Given that BRCA1 and BRCA2 are essential in tissue culture systems, evaluation of fitness is another frequently used measure for the impact of variants. However, functional assays may also address other likely cancer-relevant functions of BRCA1 and BRCA2, such as their roles in replication fork protection^23–26, transcription^27–30, and cell cycle checkpoint control^31–34. According to the latest ACMG/AMP framework (with ClinGen Sequence Variant Interpretation (SVI) refinements), well-validated functional assays are accepted as strong evidence when genetic data are inconclusive³⁵. A variety of in vitro and cell-based assays have been developed to assess BRCA1 and BRCA2 function, including HR reporter, transcriptional activation, protein interaction assays, ubiquitin ligase activity, and complementation tests in knockout cells^36–46. Moreover, the emergence of genome editing technologies, especially CRISPR/Cas9, is transforming the functional interpretation of genetic variants^{14,15,47–49}. CRISPR-based approaches now allow precise engineering of mutations directly into endogenous loci, preserving native regulatory and chromatin contexts. These innovations further enable functional assessment in physiologically relevant systems, such as human-derived cell lines or organoids.

This review provides an overview of the current toolbox of functional assays used to study BRCA1 and BRCA2 variant effects, with emphasis on assays emerging in the post-genome editing era. We begin by revisiting established functional assays and then focus on newer, CRISPR-based technologies, including multiplexed variant testing and quantitative functional studies. By outlining the key strengths and limitations of the currently available assays, we aim to support both assay development and the interpretation of functional evidence in variant classification.

Assays to decode the functional impacts of BRCA1 and BRCA2 variants

HR reporter assays to monitor variant impact on DNA repair activity

Method description

HR relies critically on the coordinated action of BRCA1 and BRCA2. Because of their central role in this pathway, assays designed to probe HR function have become a cornerstone for studying the consequences of BRCA1 and BRCA2 mutations. These approaches focus on measuring how efficiently cells repair DSBs through HR and have been widely applied to the evaluation of VUS^{39,42,50–59} as reviewed in Guidugli et al.⁶⁰. Commonly used HR reporter systems include the Direct Repeat-Green Fluorescent Protein (DR-GFP) and Traffic Light Reporter (TLR) assays, which quantitatively measure HR efficiency in living cells.

The DR-GFP reporter system utilizes a stably integrated reporter construct containing two defective GFP gene fragments (SceGFP and iGFP) arranged as direct repeats (Fig. 1a). The upstream SceGFP is a mutated GFP gene containing a recognition site for the I-Sce I endonuclease, while iGFP is a downstream truncated version of the GFP gene that lacks a promoter and does not express GFP but shares sufficient sequence homology to serve as a repair template. When I-Sce I is introduced into cells (typically via plasmid transfection), it creates a site-specific DSB within the SceGFP sequence. If the cell repairs this break via HR, it uses the downstream iGFP sequence as the template to restore a functional GFP gene. This results in GFP expression, which can be easily detected and quantified by flow cytometry. The proportion of GFP-positive cells reflects the efficiency of HR-mediated DNA repair^{22,50,51,61–63}.

Fig. 1. Functional assays for BRCA1 and BRCA2 variants.

For simplicity, only BRCA2 is depicted in the schematic, except for the BRCA1 TA assay. a DR-GFP reporter assay for measuring HR efficiency of BRCA2 variants. BRCA2 mutant construct and DR-GFP reporter construct are introduced into BRCA2-deficient cells. The DR-GFP reporter construct contains two non-functional GFP gene fragments SceGFP and iGFP arranged as direct repeats. Upon expression of the I-Sce I endonuclease, a site-specific DSB is introduced at SceGFP. If the break is repaired via HR using iGFP as a template, a functional GFP gene is restored, and GFP-positive cells can be quantified by flow cytometry. The percentage of GFP-positive cells reflects the efficiency of HR-mediated repair. The nature of the BRCA2 mutation determines the HR outcome. The star symbol indicates the variant in the schematic. b Complementation assay in Brca2-deficient mouse ES cells. Human BRCA2 variants are generated using a BAC construct and electroporated into mouse ES cells carrying a conditional allele of Brca2. Cre-mediated deletion of this allele activates a human HPRT1 mini-gene, enabling selection of recombinant clones in hypoxanthine-aminopterin-thymidine (HAT) medium. Wild-type human BRCA2 restores cell viability and resistance to DNA-damaging agents. Failure to rescue these functions indicates a deleterious effect of the variant. The star symbol indicates the variant in the schematic. c GAL4–luciferase TA assay for BRCA1 variants. A GAL4 DNA-binding domain (GAL4-DBD) fused to a BRCA1 mutant construct and a firefly luciferase reporter construct containing GAL4-binding sites are co-transfected into HEK293T cells. If the BRCA1 mutant retains transcriptional activation function, the GAL4-DBD–BRCA1 fusion protein binds the GAL4 promoter, driving firefly luciferase expression and light emission. Renilla luciferase is co-expressed as an internal control for transfection efficiency (not shown). Firefly luciferase activity is normalized to Renilla luciferase, and normalized values for BRCA1 mutants are compared with those for BRCA1 wild type. d SGE approach. A pooled SNV library and Cas9 guide RNAs are transfected into HAP1 cells, targeting the endogenous BRCA2 loci. Homology-directed repair introduces each SNV into the genome in parallel. Following editing and selective pressure (e.g., drug treatment or long-term culturing), deep sequencing before and after selection quantitatively tracks the frequency of each variant. e CRISPR-Select assay for functional assessment of single-nucleotide or InDel variants within the endogenous gene locus. Cells are co-transfected with Cas9, a guide RNA targeting the BRCA2 locus, and single-stranded donor oligonucleotides (ssODNs) encoding either a mutant (Mut, red) or wild-type synonymous (WT*, blue) allele of BRCA2. Over time, the ratio of Mut to WT* alleles is monitored via deep sequencing. Deleterious variants are depleted due to reduced cell fitness or defective DNA repair capacity, whereas benign variants show neutral behavior.

The TLR assay is based on the DR-GFP assay, but it allows simultaneous detection of both HR and non-homologous end joining (NHEJ) outcomes⁶⁴. The TLR system is composed of a mutant GFP (containing an I-Sce I site) and mCherry genes separated by a self-cleaving T2A peptide. Upon DSB induction via I-Sce I, a functional GFP is restored if the break is repaired by HR, whereas mCherry is expressed if the repair is mediated by non-homologous end joining.

Strengths and limitations

HR reporter assays are highly specific for measuring HR-dependent DNA repair activity. They are highly sensitive and provide live-cell quantitative readouts. Given that HR-deficient cells are generally hypersensitive to poly (ADP-ribose) polymerase (PARP) inhibitors, HR reporter assays are widely used to correlate HR function with drug responses, particularly in the context of BRCA1- and BRCA2-associated cancers. However, these assays are limited to cells that can stably integrate and express the reporter construct. As they are performed in overexpression systems, they may not fully recapitulate endogenous regulation. Despite these limitations, HR reporter assays remain a robust and widely adopted tool for functional characterization of BRCA1, BRCA2, and other HR gene variants.

Complementation and drug sensitivity assays to monitor variant impact on cell fitness

Method description

Complementation and drug sensitivity assays enable the functional characterization of BRCA1 and BRCA2 variants by testing whether they can restore cellular viability in Brca1- and Brca2-deficient mouse embryonic stem (ES) cells, as well as their response to DNA-damaging agents^{41,43,65–78}. They typically employ the Bacterial Artificial Chromosome (BAC) system to accommodate the exceptionally large size of BRCA2. Complementation with wild-type human BRCA2, delivered via a BAC construct, restores HR function and drug resistance, thereby restoring cell survival. To test the impact of specific BRCA2 variants, a BAC construct carrying the desired mutation in the BRCA2 gene is delivered to Brca2-deficient mouse ES cells. After subjecting the cells to selection, they are monitored for survival and sensitivity to therapeutic drugs (Fig. 1b). If the BRCA2 variant complements the loss of endogenous Brca2 and promotes resistance to DNA-damaging drugs, it is likely functionally intact. Conversely, failure to rescue cell survival or increased drug sensitivity indicates the variant is deleterious or functionally compromised⁶⁵.

Another assay that utilizes a similar approach is the mixed-all-nominated-in-one BRCA (MANO-B) method⁷⁹. Instead of using mouse ES cells, this method employs a BRCA2-knockout human colorectal adenocarcinoma cell line, DLD1 BRCA2^(−/−), which is co-transfected with individual BRCA2 variant cDNAs (generated via site-directed mutagenesis and subcloned into a piggyBac vector containing a barcode) along with a transposase expression vector. After puromycin selection, the cells are cultured in the presence of drugs, and deep sequencing is performed to quantify the abundance of each variant.

Strengths and limitations

Complementation and drug sensitivity assays provide a reliable and physiologically relevant method for variant analysis, with direct implications for therapeutic relevance. Since these assays mimic endogenous expression, they limit artifacts associated with overexpression. Another key strength of these assays is that they can accommodate multiple variant types, including insertions–deletions (InDels). However, there are some limitations. The assays are technically demanding and time-consuming, requiring the use of genetically engineered mouse ES cells, large BAC constructs, and precise genome editing. Although the mouse ES cell system is a powerful model, human BRCA2 may not completely recapitulate its interactions in the mouse background. Additionally, the throughput is limited, making it less suitable for large-scale screening of variants. Finally, variable expression levels of reconstituted proteins may confound functional interpretation.

BRCA1 transcriptional activation (TA) assay to monitor variant impact on transcription

Method description

The BRCA1 TA assay relies on BRCA1’s role in activating gene transcription through its C-terminal transcriptional activation domain^{40,42,80–84} as reviewed previously^36,85. In the assay, BRCA1 constructs containing wild-type or variant sequences are fused to a heterologous DNA-binding domain (DBD) such as GAL4-DBD or LexA-DBD (Fig. 1c). The fusion protein is expressed in cells along with a reporter gene (e.g., luciferase or β-galactosidase), which is placed under the control of a promoter containing the corresponding GAL4 or LexA binding sites^36,86. Benign variants retain the transcriptional activation function of BRCA1 and thus drive strong expression of the reporter gene. In contrast, pathogenic variants disrupt this function and result in reduced or no reporter gene expression^40,42.

Strengths and limitations

The BRCA1 TA assay is relatively simple and cost-effective, and it can be implemented in both yeast and mammalian systems^40,86. However, this assay is limited to the C-terminus domain of BRCA1 (BRCT domain). Since the method relies on fusion proteins and reporter constructs, the assay may not fully represent physiological conditions or native protein interactions. Additionally, some pathogenic variants may retain transcriptional activity but still disrupt other BRCA1 functions, leading to false negatives in this assay.

Saturation genome editing (SGE) to monitor fitness impacts of multiple variants

Method description

SGE is a high-throughput CRISPR/Cas9-based approach designed to systematically assess the functional impact of all possible single-nucleotide variants (SNVs) across targeted genomic regions. In studies of BRCA1 and BRCA2, SGE is most often performed by introducing variants directly into the endogenous gene locus of haploid human cell lines, such as HAP1^14,47,87 (Fig. 1d). The approach has also been extended to engineered non-haploid cell lines, in which one allele of BRCA1 or BRCA2 is deleted or humanized, thereby enabling functional assessment of individual variants^88–92. These cell systems are particularly suited for large-scale functional assays as each gene exists in a single functional copy, either naturally (haploid cells) or via targeted allele deletion (engineered diploid cells), allowing variant effects to manifest without compensation from a second allele. Using CRISPR/Cas9, guide RNAs are designed to induce targeted DSBs in specific exons of interest. A specific gRNA is matched with a library of donor templates encoding all possible SNVs in the vicinity, followed by their introduction to cells. The templates drive homology-directed repair, which rewrites the locus in parallel so that each cell carries a different variant. After genome editing, cells undergo a selection process, such as prolonged culture or exposure to DNA-damaging agents, to enrich for BRCA-relevant functional outcomes. Deep sequencing before and after selection enables quantitative tracking of variant frequencies over time. Deleterious variants, particularly those disrupting protein function or splicing, are depleted, whereas benign variants remain stable relative to synonymous controls. Notably, SGE analysis will generally also include known benign and pathogenic variants observed in patients, which serve as internal controls to calibrate these assays and enable classification of variant pathogenicity¹⁴. Loss-of-function variants often lead to growth disadvantage or increased drug sensitivity, providing a clear functional readout^47,87.

Strengths and limitations

The key strengths of SGE are its scalability and comprehensiveness. It enables relatively unbiased functional assessment of thousands of missense and synonymous variants in a single experiment, directly within the endogenous genomic context. This allows for high-resolution functional maps across coding exons, helping to clarify the clinical significance of rare or novel variants. However, SGE has limitations as it requires strong selection-based readouts, which may not be available for all BRCA1 and BRCA2 variant types. The approach is technically demanding, relying on high editing efficiency and deep sequencing infrastructure. Furthermore, SGE is typically constrained to specialized, genetically tractable cell lines, potentially limiting generalizability to other cellular contexts. It also focuses primarily on missense variants, offering less insight into other alteration types, such as InDels. Nonetheless, SGE is a powerful tool for systematically decoding missense variant functions in BRCA1 and BRCA2.

CRISPR-Select to quantify fitness impacts of variants

Method description

CRISPR-Select is a cell-based genome editing assay that enables quantitative functional characterization of genetic variants directly within their native genomic context in multiple cell model systems, including breast epithelial cells^{15,48,93–97}. The method uses CRISPR/Cas9 to introduce single-nucleotide or InDel variants into human cells, followed by a quantitative readout of variant impact through competitive growth assays or DNA damage response phenotyping. To decode variants in BRCA1 and BRCA2, guide RNAs are first designed to target the precise genomic locus of interest (Fig. 1e). A pool of donor templates is then co-delivered into the cell system, introducing the test variant and synonymous ‘neutral’ controls. Known loss-of-function alleles can be added as further internal benchmarks. Following genome editing, cells are expanded over several generations to allow functional differences to influence cell fitness. Allele-specific deep sequencing at early and late time points enables quantitative tracking of each variant’s abundance, revealing its relative effect on cell survival relative to synonymous control variants^15,48. Because BRCA1 and BRCA2 play essential roles in HR repair, functionally deleterious variants often result in impaired growth or hypersensitivity to agents like PARP inhibitors.

Strengths and limitations

CRISPR-Select is precise and accurate due to variants and internal controls being created simultaneously in the same culture system. Thus, the only variable is the oligonucleotide repair templates that differ between the test and synonymous variants. Due to efficient controls, this assay yields quality results in difficult-to-edit diploid cell types, such as breast epithelial cells. Moreover, it tests variants in situ, preserving endogenous regulation and splicing. Thus, it can resolve both severe and subtle (hypomorphic) functional effects and allows multiplexed testing of multiple variants in parallel. Nevertheless, limitations exist. The method is difficult to scale up, and it requires deep sequencing infrastructure. Nonetheless, CRISPR-Select represents a powerful tool for functional genomics in BRCA1 and BRCA2.

Clinical implementation of functional assays

The ClinGen ENIGMA BRCA1 and BRCA2 VCEP employs a quantitative, point-based classification system to integrate multiple lines of evidence, including functional assay scores, for accurate variant interpretation^98,99. Each criterion, ranging from population frequency and computational predictions to functional assays and segregation data, is assigned a point value reflecting its strength toward pathogenicity or benignity. These points are summed to generate an overall score, which determines the clinical classification of a variant (e.g., Pathogenic, Likely Pathogenic, VUS, Likely Benign, or Benign). This systematic approach enhances consistency, transparency, and clinical utility in variant assessment.

As mentioned above, since many of the observed variants are rare, variant classification is heavily dependent on population frequency, computational predictions, and functional assays. Under the ClinGen ENIGMA BRCA1 and BRCA2 VCEP guidelines, population frequency (PM2, BA1, BS1) and computational prediction (PP3, BP4) criteria are applied at supporting strength toward either pathogenicity or benignity¹². Within the Bayesian point-based classification framework, each supporting-level criterion contributes just +1 point toward pathogenicity; therefore, population frequency and computational predictions together yield only +2 points^99,100. Thus, while population frequency and computational prediction criteria contribute to variant assessment, their limited weight underscores the need for stronger evidence sources. To achieve a clinically actionable classification of Likely Pathogenic (6–9 points) or Pathogenic (≥10 points), additional evidence is required, which further highlights the importance of functional assays during variant classification.

Results from functional assays (PS3/BS3) are weighted as strong evidence, contributing +4 points toward pathogenicity. By combining functional assays (+4 points) with the readily accessible population frequency (+1 point) and computational predictions (+1 point), many VUS can reach the 6-point threshold required for a clinically actionable Likely Pathogenic classification. Therefore, incorporating functional assay data could significantly enhance variant classification efforts. However, a recent study reported that functional assay data are used in only an underwhelming 10% of diagnostic variant classifications¹⁰¹, highlighting a major gap between the potential utility of these assays and their actual adoption in clinical practice.

Broader adoption in clinical practice remains challenging due to several factors, including: (1) varying levels of familiarity with functional assay design and application; (2) evolving standards for assay validation; (3) rapid evolution of variant classification guidelines that outpaces clinical uptake; (4) limited availability of personnel with specialized training in assay execution and interpretation; (5) absence of clear clinical frameworks for interpreting intermediate (partial-loss-of-function) variants and for harmonizing outputs from multiple functional assays. Systematically addressing these challenges will significantly advance the use of functional assays in variant classification. Below, we elaborate on several of these issues and propose potential solutions to help overcome current limitations in their application.

Varying levels of familiarity with functional assay design and application

Broadly, functional assays are designed to investigate the impact of BRCA1 and BRCA2 genetic variants on either impaired protein function or splicing defects. As discussed earlier, these assays include biochemical assays (e.g., enzyme activity tests), transcript assays (e.g., splicing assays or transcriptome analysis), cell-based models (e.g., in vitro functional assays), animal models (e.g., mouse models), and high-throughput approaches such as multiplexed assays of variant effect (MAVEs).

With the growing interest in developing BRCA1 and BRCA2-based functional assays, it has become increasingly challenging for clinicians and variant curation scientists to assess their appropriateness. This challenge is further compounded by the use of diverse cell and animal models, along with varying assay readouts—such as drug sensitivity, HR repair proficiency, and cell viability—applied in both low- and high-throughput formats. Reflecting this complexity, a recent survey of diagnostic professionals found that fewer than 50% felt comfortable using functional assays in variant interpretation, largely due to limited familiarity with specific assay systems¹⁰².

To address these challenges, the routine reporting of a minimum set of essential information is crucial for assessing the validity and relevance of each published functional assay. Importantly, the ClinGen SVI Working Group has already developed a standardized tool called Functional Assay Documentation Worksheet (https://clinicalgenome.org/docs/svi-functional-assay-documentation-worksheet/) to support this effort³⁵. However, recent reports indicate that only 10–40% of diagnostic professionals are aware of this resource, significantly limiting the utility and outreach of functional data^102,103. Compounding this issue is the fact that much of the functional assay development is conducted by basic researchers, who may be even less familiar with this clinical framework. Encouraging both basic researchers and diagnostic professionals to adopt and complete the SVI functional assay documentation worksheet could enhance transparency, facilitate variant interpretation, and help bridge the gap between assay development and its clinical application.

Evolving standards for assay validation

The initial ACMG/AMP guidelines proposed the use of functional assay results (PS3/BS3) as strong evidence in variant interpretation⁹; however, the guidelines lacked specific direction on how such evidence should be interpreted and applied in practice. A key contributor to this ambiguity is the lack of clarity around which functional assays should be considered as ‘well-established’. Ideally, assays calibrated against a diverse set of known benign and pathogenic variants (truth sets) offer greater reliability and interpretive power than those validated with only a limited number of variants or with exclusively benign or pathogenic variants. Nevertheless, the initial guidelines treated both types of assays equivalently, allowing their results to be applied at the same evidence strength. This lack of distinction led to widespread inconsistency in the application of PS3/BS3 among clinical laboratories¹⁰⁴. To address this, the ClinGen SVI Working Group proposed a set of guidelines to evaluate the technical rigor and reproducibility of functional assays by estimating the Odds of Pathogenicity (OddsPath)³⁵. OddsPath is calculated based on the number of known benign and pathogenic variants included in the assay’s validation and its ability to distinguish between them using predefined thresholds for functional readouts. Results from these estimations are then used to assign corresponding levels of evidence strength for each functional assay: for benign evidence—BS3 (OddsPath < 0.053), BS3_moderate (<0.23), and BS3_supporting (<0.48); and for pathogenic evidence—PS3_supporting (OddsPath >2.1), PS3_moderate (>4.3), and PS3 (>18.7)⁹⁹. Assays yielding OddsPath values between 0.48 and 2.1 are considered indeterminate and do not provide reliable support for variant classification³⁵.

Practically, a functional assay that yields a clear binary readout can support different levels of evidence strength depending on the number and quality of well-characterized controls included. Supporting-level evidence (PS3_supporting or BS3_supporting) requires at least three benign (OddsPath = 0.333) and three pathogenic (OddsPath = 3.0) variants. Moderate-level evidence (PS3_moderate or BS3_moderate) is achieved with a minimum of five benign (OddsPath = 0.200) and five pathogenic (OddsPath = 5.0) variants. Strong evidence (PS3 or BS3) demands at least 19 benign (OddsPath = 0.053) and 19 pathogenic (OddsPath = 19.0) variants³⁵. Scientists aiming to develop new functional assays for clinical variant interpretation, such as for BRCA1 and BRCA2, are encouraged to incorporate these minimum numbers of control variants to support clinical validation. Given that variant classification often relies heavily on full-strength PS3/BS3 evidence (+4 points), optimally designed assays include at least 19 benign and 19 pathogenic variants, assuming a clear and reproducible binary readout. To further strengthen the role of functional assays in variant classification, the routine reporting of OddsPath metrics should be included as a minimum standard for all published functional studies. Brnich et al. provided a detailed framework for calculating OddsPath for individual assays³⁵. Encouraging both basic researchers and diagnostic professionals to adopt this approach will help accelerate the clinical translation of functional assay findings. The corresponding evidence strength, as determined by OddsPath, can also be documented using the SVI functional assay documentation worksheet, which is designed to capture this information³⁵.

The complexity of intermediate variants

Variants that show an intermediate effect in functional assays warrant special consideration³⁵. The observed intermediate effect could arise because the assay fails to capture the protein’s complete biological role, or because the variant causes a partial loss of function associated with a moderate increase in disease risk, consistent with low-penetrance or risk alleles^35,105,106. Although current ClinGen ENIGMA BRCA1 and BRCA2 VCEP variant interpretation guidelines do not allow such intermediate results to contribute toward PS3 or BS3 classification under the ACMG/AMP framework, these findings may still hold clinical relevance¹². However, their potential utility in patient care remains uncertain, largely due to the rarity of bona fide low-penetrance BRCA1 and BRCA2 alleles and the lack of established risk estimates. To address this, broader collaborative efforts, such as those led by ENIGMA, are combining data from multiplexed assays (e.g., MAVEs) with large case-control studies to more accurately define the clinical significance of these variants. These efforts may ultimately support more informed risk assessment and tailored management for individuals carrying low-penetrance or risk alleles who currently receive inconclusive results and lack clear clinical guidance.

Harmonizing outputs from multiple functional assays

Combining data from different orthogonal or complementary functional assays can significantly improve the clinical classification of BRCA1 and BRCA2 genetic variants³⁵. Each assay might measure a distinct aspect of gene function, such as cell fitness, drug sensitivity, or HR activity, which means that when the results are integrated, the assays offer a more comprehensive view of BRCA1 and BRCA2 functional integrity in the presence of a variant. Integrating data from multiple functional assays increases confidence in variant interpretation by reducing the impact of limitations or artifacts associated with any single assay. When different types of assays consistently suggest that a variant is deleterious, the overall evidence becomes more robust. Conversely, if one assay produces conflicting or uncertain results, data from additional assays can help resolve these discrepancies and support the reclassification of VUS. Brnich et al. proposed a general framework for integrating results from multiple functional assays, outlining three key scenarios³⁵. First, if assay results are consistent, the PS3 or BS3 criteria should be applied at the strength corresponding to the assay with the highest level of validation³⁵. Secondly, if results are conflicting, the assay that has a better level of validation and more closely reflects the disease mechanism should be prioritized, with PS3 or BS3 applied at the strength according to that assay’s validation parameters³⁵. Thirdly, if the assays yield conflicting results and have similar levels of validation, the functional evidence should not be used in the interpretation of the variant³⁵.

However, applying this framework in practice presents challenges, particularly for genes like BRCA1 and BRCA2, where over 150 functional assay datasets have been reported for each gene¹⁰⁷. The manual collection, curation, and integration of data from multiple assays for each variant remains a complex and resource-intensive process^103,108. Fortunately, this need has been addressed through comprehensive harmonization efforts, most notably by Lyra et al., who curated and standardized functional data for 2701 BRCA1 missense variants, representing 24.5% of all possible missense changes in the gene^106,109. Similar efforts are ongoing for BRCA2, as well as the expansion of BRCA1 data beyond the initial study¹⁰⁷. We believe this harmonized resource offers a one-stop solution for time-constrained diagnostic professionals aiming to incorporate PS3/BS3 evidence into variant classification.

Maximizing ClinGen ENIGMA BRCA1 and BRCA2 VCEP and ClinVar to overcome current barriers

Many of the issues outlined above could be constructively addressed through increased engagement from the ClinGen ENIGMA BRCA1 and BRCA2 VCEP. This might include a systematic review of existing functional assays to evaluate their suitability for variant classification, calibration using OddsPath metrics, and guidance on the appropriate strength of PS3 or BS3 evidence. As discussed earlier, these details can be documented using the SVI functional assay documentation worksheet, offering greater assurance to diagnostic professionals when applying VCEP-approved functional evidence^102,108. This is particularly important given that, in a recent survey, over 90% of diagnostic professionals cited insufficient information on the quality and accuracy of functional data as a major barrier for its use in variant interpretation¹⁰³. Fortunately, the BRCA1 and BRCA2 VCEP have already taken proactive steps in this direction by evaluating the appropriateness of several BRCA1 and BRCA2 functional assays for use in variant classification¹⁰². Furthermore, it has also calibrated these assays using OddsPath metrics and assigned the corresponding PS3 or BS3 strength, as documented in the SVI functional assay documentation worksheet outlined by Villani et al.¹⁰². The continued and routine accreditation of functional assays by the BRCA1 and BRCA2 VCEP can significantly accelerate and strengthen the use of functional evidence in clinical interpretation.

We also note that the dissemination of functional assay data is currently suboptimal, clearly limiting its effective use in variant classification. Improving how these data are shared could significantly enhance the integration of functional evidence into clinical interpretation. One promising approach is to incorporate functional data into widely used public databases such as ClinVar, which is already a key resource for clinicians and researchers^103,108,110. In fact, a recent survey among diagnostic professionals found that 94% believed the presence of functional data in ClinVar would enhance its use in variant interpretation¹⁰³. As suggested by Park et al., a dedicated functional data dashboard could be implemented in ClinVar for each functionally assessed variant¹⁰³. This dashboard could present essential details such as variant effect scores and their distribution, classification as functionally abnormal or normal, study references (e.g., PMID), and the recommended PS3 or BS3 strength¹⁰³. To ensure reliability, we recommend that only functional assays approved by the BRCA1 and BRCA2 VCEP be included in such a system. This would allow diagnostic professionals to quickly check for the presence of functional evidence for a given variant and apply it effectively in their interpretation. Most importantly, adapting ClinVar for this purpose would be particularly useful for disseminating functional data on intermediate or discordant variants, where clear guidance on PS3 or BS3 strength is most needed.

Conclusions

Functional assays for BRCA1 and BRCA2 are rapidly evolving, with CRISPR-based approaches such as SGE and CRISPR-Select offering powerful, context-specific insights into variant pathogenicity. Although each method has limitations, combining orthogonal assays with standardized functional evidence frameworks, as well as available clinical data, will be essential for translating assay results into patient care.

Author contributions

Conception and design: M.R. and C.S.S. Prepared figure: E.E.F. Manuscript writing: M.B., E.E.F., M.R., and C.S.S. Final approval of manuscript: all authors.

Data availability

No datasets were generated or analyzed during the current study.

Competing interests

A patent application (EP20211801) for CRISPR-Select was submitted by the University of Copenhagen and its inventors (including C.S.S.) on December 4, 2020. C.S.S. is a cofounder of BioPhenyx, a biotechnology company that uses CRISPR-Select assays. All other authors have no competing interests.