Clinical and Functional Analyses of an African-ancestry Gain-of-function HOXB13 Variant Implicated in Aggressive Prostate Cancer

Abstract

Background:

Recent reports have uncovered a HOXB13 variant (X285K) predisposing to prostate cancer in men of West African ancestry. The clinical relevance and protein function associated with this inherited variant are unknown.

Objective:

To determine the clinical relevance of HOXB13 (X285K) in comparison with HOXB13 (G84E) and BRCA2 pathogenic/likely pathogenic (P/LP) variants, and to elucidate the oncogenic mechanisms of the X285K protein.

Design, setting, and participants:

Real-world data were collected from 21 393 men with prostate cancer undergoing genetic testing from 2019 to 2022, and in vitro cell-line models were established for the evaluation of oncogenic functions associated with the X285K protein.

Outcome measurements and statistical analysis:

Genetic testing results were compared among patient groups according to self-reported race/ethnicity, Gleason scores, and American Joint Committee on Cancer stages using the exact test. Oncogenic functions of X285K were evaluated by RNA sequencing, chromatin immunoprecipitation sequencing, and Western blot analyses.

Results and limitations:

HOXB13 (X285K) was significantly enriched in self-reported Black (1.01%) versus White (0.01%) patients. We observed a trend of more aggressive disease in the HOXB13 (X285K) and BRCA2 P/LP carriers than in the HOXB13 (G84E) carriers. Replacement of the wild-type HOXB13 protein with the X285K protein resulted in a gain of an E2F/MYC signature, validated by a gain in the expression of cyclin B1 and c-MYC, without affecting the androgen response signature. Elevated expression of cyclin B1 and c-MYC was explained by enhanced binding of the X285K protein to the promoters and enhancers of these genes. The limitations of the study are the lack of complete clinical outcome data for all patients studied and the use of a single cell line in the functional analysis.

Conclusions:

HOXB13 X285K is significantly enriched in self-reported Black patients, and X285K carriers detected in the real-world clinical setting have aggressive prostate cancer features similar to the BRCA2 carriers. Functional studies revealed a unique gain-of-function oncogenic mechanism of X285K protein in regulating E2F/MYC signatures.

Patient summary:

The HOXB13 X285K variant is clinically and functionally linked to aggressive prostate cancer, supporting early disease screening of Black men carrying the HOXB13 X285K variant.

1. Introduction

Heritable genetic risk factors play a significant role in prostate cancer etiology and clinical practice [1]. Currently, the adoption of genetic testing in prostate cancer is mainly driven by the established prognostic and treatment utility of rare pathogenic variants in BRCA2 and other DNA damage response (DDR) pathway genes [2]. The underlying pathogenic mechanism for these variants is loss of DNA repair function, often due to frameshift or premature stop alterations causing protein truncation [3]. In addition to DDR genes, multiple genetic variants associated with prostate cancer risk have been found in the HOXB13 gene [4,5] in different ancestral populations, including European (eg, G84E) [6], Japanese (G132E) [7], and Chinese (G135E) [8] populations. Unlike pathogenic DDR gene variants, these HOXB13 variants result in alterations in a single amino acid, and while useful for early disease screening, there is little evidence supporting their prognostic and treatment utility. The functional implications of germline HOXB13 variants detected in European and East Asian ancestral populations remain poorly characterized, although the wild-type (WT) HOXB13 protein is known to be expressed abundantly and specifically in cells of prostatic lineage [4,9], with roles in regulating the androgen receptor (AR) axis [10,11].

Recently, a stop-loss HOXB13 variant c.853delT (referred to as X285K), in which a single base deletion within the HOXB13 stop codon (c.853delT) results in an extension of the HOXB13 protein by 96 amino acids, was identified as a risk factor for prostate cancer in men of African ancestry [12–14]. This variant was first reported to be associated with prostate cancer by Marlin et al [13] in three prostate cancer patients in Martinique out of a total of 46 prostate cancer patients diagnosed before age 51 yr. Subsequently, we reported eight X285K carriers identified by whole-exome sequencing in a cohort of 1048 self-reported African-American prostate cancer patients who had radical prostatectomy [14]. All eight carriers had prostate cancers of Gleason grade 7–9. Most recently, in a study by Darst et al [12], X285K carrier status was determined by genetic imputation and was associated with late-stage disease, with an overall carrier rate of 0.7% in prostate cancer patients of African ancestry. It was estimated that the variant emerged ~1500–4600 yr ago in West Africa [12]. Therefore, HOXB13 X285K represents a unique germline variant affecting the risk for aggressive prostate cancer, specifically in men of African ancestry.

These recent reports raise important questions about the clinical relevance of HOXB13 (X285K) in the context of other clinically established prostate cancer genetic risk factors (eg, BRCA2). In addition, there is a critical need to understand whether and how the X285K protein changes the HOXB13 function. Here, we present real-world germline genetic testing results and compare clinical characteristics between patients with HOXB13 variants (X285K and G84E) and BRCA2 pathogenic/likely pathogenic (P/LP) variants. We performed in vitro functional analyses comparing the HOXB13 (X285K) variant with its WT counterpart, revealing a unique gain-of-function oncogenic mechanism related to E2F/MYC signaling. These findings inform genetic testing strategies and potential therapeutic development targeting HOXB13 (X285K).

2. Patients and methods

2.1. Germline genetic testing data analysis

We conducted an IRB-approved (protocol 1167406) review of the deidentified data from 21 393 men with prostate cancer who received clinical-grade germline testing, as described previously [15]. Patients included in this analysis were tested through the Detect Hereditary Prostate Cancer (DHPC)-sponsored (no-charge) testing program, which ran from 2019 to 2022 (Invitae Corporation, San Francisco, CA, USA). The inclusion criteria for DHPC were low-, intermediate-, or high-risk localized, stage >IIb disease, or stage IIA disease diagnosed in <55 yr olds. Briefly, patients underwent full gene sequencing, including a deletion-duplication analysis for cancer-relevant genes, with the number of genes ordered varying at the discretion of the ordering clinician. Variants were classified using a refinement of the American College of Genetics and Genomic criteria (Invitae’s Sherloc) [16], and those classified as P/LP were orthogonally confirmed. Patients who underwent testing for HOXB13 (G84E and X285K) variants (n = 21 091) or BRCA2 P/LP variants (n = 21 362) were stratified by self-reported race/ethnicity: Black (or African American), White (non-Hispanic), and all other (including multiracial) groups. Gleason grades and American Joint Committee on Cancer (AJCC) tumor stage were clinician reported on the genetic testing requisition form. Statistical analyses were completed using R (version 4.1.2) and a significance threshold of α <0.05 was used. Additional details are described in the Supplementary material.

2.2. Functional analysis

Detailed in vitro functional analyses were conducted in stable LNCaP95 (LN95) cell clones that were engineered to express exogenous HOXB13 WT (n = 5) and HOXB13 X285K (n = 3) proteins (hereafter called LN95WT and LN95X285K, respectively) upon induction by doxycycline (Dox). By design, treatment with Dox simultaneously induces the knockdown of endogenous WT HOXB13 (Supplementary material). As a control, three clones carrying nontargeting (NT) control vectors were generated (hereafter called LN95NT). These clones (n = 11) were subjected to RNA sequencing (RNA-Seq), protein analyses, and chromatin immunoprecipitation (ChIP) sequencing (ChIP-Seq). Details are provided in the Supplementary material.

3. Results

3.1. Germline genetic testing results from the DHPC program

To determine the clinical relevance of HOXB13 (X285K) in real-world practice, we extracted genetic testing data from 21 393 patients and interrogated for HOXB13 (X285K), together with HOXB13 (G84E), as well as established BRCA2 P/LP variants. Notably, while the prevalence of HOXB13 (X285K) in prostate cancer patients was 0.18% overall, this was markedly enriched in men of self-reported Black versus White ancestry (1.01% vs 0.01%; Table 1). Conversely, the HOXB13 (G84E) variant was enriched in White versus Black patients (1.12% vs 0.11%). By comparison, the BRCA2 P/LP rates in this real-world setting was 1.58% and 1.92%, respectively, in self-reported White and Black patients (Table 1). Further, we demonstrated differences in Gleason scores (p = 2.8 × 10⁻⁶) and cancer stages (p = 6.6 × 10⁻⁴) among HOXB13 (X285K) carriers, HOXB13 (G84E) carriers, and BRCA2 carriers (Table 1). Compared with G84E carriers, X285K and BRCA2 P/LP carriers had a higher prevalence of stage IV (including metastatic disease) at presentation (93.7% vs 67.7%, p = 0.037 and 85.6% vs 67.7%, p = 2.5 × 10⁻⁴, respectively). Although statistically insignificant, compared with G84E carriers, X285K carriers also had a numerically higher prevalence of Gleason 8–10 histology (62.5% vs 49.3%, p = 0.243). Interestingly, we did not detect differences between X285K carriers and BRCA2 P/LP carriers in cases with Gleason 8–10 histology (62.5% vs 70.0%, p = 0.423) and stage IV disease at presentation (93.7% vs 85.6%, p = 0.707), although, within the Gleason 8–10 histology, numerical differences in Gleason 8 (40.6% vs 20.5%) and Gleason 9–10 (21.9% vs 49.5%) prevalence were observed between the X285K carriers versus BRCA2 P/LP carriers (Table 1). Among the 2715 self-reported Black patients (2680 tested for HOXB13), a “noncarrier” group (n = 2626) was created after excluding BRCA2 (n = 52), BRCA1 (n = 7), X285K (n = 27), and G84E (n = 3) carriers (Supplementary Table 1). Among Black patients, we found a numerically higher prevalence of Gleason 8–10 histology (63.6% vs 52.2%, p = 0.39) and stage IV disease (91.7% vs 77.1%, p = 0.32) in the X285K carriers, when compared with the noncarrier group.

Table 1 –

Summary of germline genetic testing results from the DHPC program

	HOXB13	HOXB13	BRCA2	p value for difference
	p.X285K	p.G84E	P/LP variants
N variants detected/N evaluated (%)	38/21 091 (0.18)	173/21 091 (0.82)	382/21 362 (1.79)
Black	27/2680 (1.01)	3/2680 (0.11)	52/2715 (1.92)
White	1/13628 (0.01)	153/13 628 (1.12)	218/13 821 (1.58)
Other/multiracial	10/4783 (0.21)	17/4783 (0.36)	112/4826 (2.32)
	p = 2.2 × 10−16 a	p = 1.8 × 10−12 a	p = 0.0037 a
Race/ethnicity in variant+ patients, n (%)				p = 2.2 × 10−16 b
Black	27 (71.1)	3 (1.7)	52 (13.6)
White	1 (2.6)	153 (88.4)	218 (57.1)
Other/multiracial	10 (26.3)	17 (9.8)	112 (29.3)
Gleason score, n (%)				p = 2.8 × 10−6 b
≤6	1 (3.1)	18 (11.7)	12 (3.7)
7	11 (34.4)	60 (39.0)	86 (26.3)
8	13 (40.6)	31 (20.1)	67 (20.5)
9–10	7 (21.9)	45 (29.2)	162 (49.5)
Unknown	6	19	55
Stage, n (%)				p = 6.6 × 10−4 b
Stage II	1 (6.3)	14 (14.6)	10 (3.9)
Stage III	0 (0.0)	17 (17.7)	27 (10.5)
Stage IV	15 (93.7)	65 (67.7)	221 (85.6)
Unknown	22	77	124

DHPC = Detect Hereditary Prostate Cancer; P/LP = pathogenic/likely pathogenic.

All p values were calculated with exact test. Patients with unknown or missing Gleason score and stage information were excluded from the analysis.

^aComparison of race/ethnicity distribution between patients with and without the variant.

^bComparison among the three variants.

3.2. Population-attributable risk

Results from the real-world genetic testing data are limited by the lack of a control population. Nevertheless, population-attributable risk (PAR) can be estimated for each specific genetic variant. PAR for HOXB13 (X285K) was estimated to be 0.62% in the Black/African-American population (Table 2). By comparison, PAR was 0.59% for HOXB13 (G84E) in the White population and 1.19% for BRCA2 in the entire population (Table 2).

Table 2 –

Estimation of the population-attributable risk (PAR) for HOXB13 and BRCA2 variants

	DETECT program	gnomAD a	PAR (%) b
HOXB13 X285K (Black)	0.010075	0.003924	0.62
HOXB13 G84E (White)	0.011227	0.005317	0.59
BRCA2 (entire population)	0.017882	0.006034	1.19

^aSingle variant carrier frequency data for HOXB13 X285K and G84E were from male patients in the Genome Aggregation Database (gnomAD). The carrier frequency for BRCA2 was the sum of carrier frequencies for 247 BRCA2 pathogenic/likely pathogenic variants detected in the entire population.

^bPAR was calculated by 1 – ([1 – proportion of variant+ among men with prostate cancer]/[1 – proportion of variant+ in the population]).

3.3. Establishment of in vitro models for X285K

Structurally, X285K is a single-base deletion within the HOXB13 stop codon (c.853delT), resulting in an extension by 96 amino acids of the HOXB13 protein (p.Ter285Lys ext96) at the C terminal of the DNA-binding homeobox domain, as visualized by Alphafold (Supplementary Fig. 1). To study the functions mediated by X285K, we devised a strategy (Fig. 1A) to replace the WT HOXB13 with X285K in LNCaP95, a castration-resistant prostate cancer cell line demonstrating HOXB13-dependent AR/AR-V7 functional output [17,18]. Treatment with increasing doses of Dox successfully displaced the endogenous WT HOXB13 with exogenous WT or X285K HOXB13 in a dose-dependent manner (Fig. 1B). Replacement of endogenous WT with exogenous X285 was validated by both Western blot analyses and quantitative reverse transcription reverse-transcription (RT-PCR) in all 11 stable clones following five different Dox doses (Supplementary Fig. 2).

Fig. 1 –

(A) Schematic diagram of HOXB13 bidirectional inducible vectors in stable LNCaP95 clones including five LN95WT, three LNX285K, and three LN95NT clones. (B) Western blot showing replacement of endogenous WT with exogenous WT and X285K HOXB13 in representative clones (LN95WT clone 26 and LN95X285K clone 21). Higher molecular weights for the exogenous proteins are due to tags and 96 amino acid extension. Homogeneous RFP expression was confirmed upon Dox treatment. Dox = doxycycline; RPF = red fluorescent protein; WT = wild type.

3.4. RNA-Seq analyses

We generated RNA-Seq data from 44 cell line samples after subjecting each of the 11 stable clones (five LN95WT, three LN95X285K, and three LN95NT) to four different treatment conditions (with or without Dox, in the presence or absence of androgen; Supplementary Fig. 3A). Examination of the sequencing data did not uncover a change in AR/AR-V7 expression (not shown). Surprisingly, CCNB1 (encoding cyclin B1) was identified as a top-ranked gene that was upregulated following induction of exogenous X285K by Dox treatment (RT-PCR validation of RNA-Seq data; Fig. 2A). Interestingly, this occurred only in androgen-stimulated conditions (Fig. 2A), and the degree of CCNB1 upregulation appeared to positively correlate with the abundance of exogenous X285K expression (Supplementary Fig. 2). Conversely, overexpression of WT HOXB13 suppressed CCNB1 expression (Fig. 2A). We next conducted unbiased analyses focusing on the comparison of WT and X285K in androgen-stimulated cells. Notably, a differential gene expression analysis identified MYC and CCNB1 among genes significantly upregulated in X285K clones compared with WT clones (Fig. 2B and Supplementary Fig. 3B). In addition, a gene set enrichment analysis revealed E2F and MYC targets as the top gene sets enriched in X285K clones (Supplementary Fig. 3C).

Fig. 2 –

(A) CCNB1 mRNA level was quantified by qRT-PCR in 44 samples from the 11 clones. All clones were treated with 20 ng/ml Dox and 1 nM R1881, as shown in Supplementary Figure 3A. Mean ± SD values from four technical replicate are shown (see Methods for details). The p values for ANOVA were p < 0.0001 for both R1881 (–) and 1 nM R1881. Only comparisons with p < 0.05 from Turkey’s multiple comparison tests are annotated in the graph. (B) Heatmap of 67 differentially expressed genes between five LN95WT clones and three LN95X285K clones treated with Dox and R1881. CCNB1 and MYC are highlighted in bold. (C) Heatmap of AR-induced genes in 44 samples from 11 clones. AR-induced genes were retrieved from the study of Norris et al [11] and divided into HOXB13-repressed (left) and HOXB13-activated (right) genes. ANOVA = analysis of variance; AR = androgen receptor; Dox = doxycycline; qRT-PCR = quantitative reverse transcription reverse-transcription; SD = standard deviation; WT = wild type. ** p <0.01. *** p = 0.0005. **** p < 0.0001.

As HOXB13 interacts with AR, a key therapeutic target in prostate cancer, we performed the analysis of a previously published [11] set of HOXB13-activated and HOXB13-repressed AR target genes to determine whether HOXB13 WT and X285K demonstrate differential functions (ie, functional gain or loss) either as activators or as repressors of AR target genes. Overexpression of both exogenous WT and X285K HOXB13 resulted in suppression of HOXB13-repressed genes in both androgen-deprived and androgen-stimulated conditions, confirming the largely intact AR repressor function of both WT and X285K (Fig. 2C). Similarly, HOXB13-activated genes remain equally inducible upon androgen stimulation in cells with endogenous WT, exogenous WT, or exogenous X285K (Fig. 2C). These results are consistent with the previously reported [11] bifunctional activity of HOXB13 and suggest that the function of X285K with respect to regulating multiple AR target genes is virtually indistinguishable from WT.

3.5. Validation of gain of function by Western blot analyses

To validate a gain-of-function mechanism in regulating E2F/MYC, we conducted Western blot analyses to evaluate CCNB1 and MYC protein expression following the induction of varying levels of X285K. Consistent with the RNA-Seq data, we detected a Dox dose-dependent increase of CCNB1 and MYC protein expression in X285K clones. By contrast, a dose-dependent decrease in CCNB1 and MYC was observed in WT clones (Fig. 3A and 3B, and Supplementary Fig. 4A–C).

Fig. 3 –

Western blot analyses comparing the key proteins of the E2F/MYC signature in (A) WT and (B) X285K representative clones expressing similar levels of exogenous HOXB13 (LN95WT clone 26 and LN95X285K clone 21). Cells were treated with high and low dose ranges of Dox as indicated for 2 d with 1 nM R1881; β-actin was used as a loading control. Dox = doxycycline; WT = wild type.

3.6. Validation of gain of function by ChIP-Seq

Next, we performed HOXB13 ChIP-Seq to determine genomic binding sites that differ between the WT and X285K HOXB13 proteins. Notably, ChIP-Seq detected 4612 X285K-specific, annotated binding sites (Supplementary Fig. 5A). In comparison, only 867 binding sites were specific to the WT protein. In addition, the X285K-specific binding sites demonstrated substantially higher peaks, suggesting global epigenetic changes that are specifically induced by X285K (Supplementary Fig. 5B). To determine whether this differential binding pattern affects gene expression, we conducted a combined RNA-Seq and ChIP-Seq analysis. Interestingly, CCNB1 was one of the genes showing elevated gene expression as well as X285K binding (Fig. 4A). We confirmed a higher enrichment of X285K at the 5′ upstream region of CCNB1 (Fig. 4B), suggesting that X285K elevates CCNB1 expression by direct binding. Next, we confirmed increased X285K binding to PCAT1 and PCAT2 regions (Fig. 4C and 4D). PCAT1 and PCAT2 gene loci have recently been reported as MYC superenhancers [19], suggesting that increased X285K binding to these regions contributes to elevated MYC expression and more aggressive clinical phenotypes.

Fig. 4 –

(A) Scatterplot with ChIP-Seq differential binding (log2-fold change) in the x axis and differential mRNA expression (log2-fold change) in the y axis. Upregulated genes in Dox + R1881 + LN95X285K RNA-Seq with gained peaks in Dox + R1881 + LN95X285K ChIP-Seq are presented as red dots. (B) Integrated genome browser (IGV) snapshot of HOXB13 binding peaks at 5′ upstream region of CCNB1. X285K binding peaks are marked red. (C and D) IGV snapshots showing HOXB13 binding peaks around PCAT1 and PCAT2 regions known as MYC superenhancer. X285K binding peaks are marked red. ChIP-Seq = chromatin immunoprecipitation sequencing; Dox = doxycycline; RNA-Seq = RNA sequencing; WT = wild type.

4. Discussion

The current study characterizing the clinical and functional relevance of the HOXB13 (X285K) variant was motivated by three recent publications implicating its association with aggressive prostate cancer in men of African ancestry [12–14]. We recognized that while the three studies yielded largely concordant findings, they were conducted in genetic research settings that may not reflect real-world practice settings. In addition, these single-variant studies did not capture the breadth of various rare germline variants with respect to their detection rates and associated patient characteristics. Additionally, these studies did not address whether and how the X285K protein alters the HOXB13 functional output. Lack of functional evidence demonstrating the impact on biological processes and pathways will impede variant classification and clinical translation. The current study was also motivated by the relevance to prostate cancer disparities [20,21]. While it is well conceived that socioeconomic determinants of prostate cancer disparity impact disease presentation, including clinical, pathological, and molecular characteristics [21,22], it is also recognized that genetic variants contribute to population differences in risk for prostate cancer [23]. In this regard, it would be necessary to conduct clinical and functional characterization of the ancestry-specific HOXB13 X285K variant in order to understand the disease etiology and develop clinical utility in a patient population disproportionately affected by aggressive prostate cancer.

Our findings from the analysis of real-world genetic testing data are noteworthy and clinically informative in several aspects. First, we found that the X285K variant is detected in 1.01% of self-reported Black/African-American prostate cancer patients undergoing genetic testing. This detection rate is nearly equivalent to the 1.12% detection rate for the G84E variant in self-reported White patients. In comparison with the G84E carriers, X285K carriers were enriched for prostate cancers with Gleason score 8 and AJCC stage IV diseases, whereas BRCA2 tumors are enriched for Gleason 9–10 and AJCC stage IV diseases. G84E is the most common HOXB13 variant in men of European ancestry. This variant was reproducibly associated with the risk of prostate cancer incidence [6], but the association was equally strong in men with aggressive and nonaggressive diseases [4]. As such, although G84E can be incorporated into genetic testing to enable targeted screening and family counseling [24], it currently does not have prognostic or treatment implications. In contrast, given the clinical relevance of the X285K variant established in this real-world setting, detection of this X285K variant should be interpreted in the prognostic and treatment settings. Specifically, early escalation may be justified in prostate cancer patients carrying the X285K variant, given the association of the X285K variant with aggressive prostate cancer. We detected BRCA2 variants in 1.79% of patients across the entire tested cohort. This detection rate is substantially lower than the BRCA2 P/LP detection rate (4.74%) in a previous report [15], likely reflecting the changed guidelines and inclusion of relatively lower-risk patients following that publication. Our real-world genetic data analysis was limited by a lack of a control population, a lack of data on the age of diagnosis, and often incomplete capture of clinical, pathological, and treatment data.

Our detailed functional studies uncovered a gain of function in E2F/MYC signature specific to the X285K variant. BRCA2 and HOXB13 are among the most important hereditary prostate cancer predisposition genes. Unlike many “loss-of-function” (LOF) BRCA2 variants with established diagnostic, prognostic, and treatment implications, evidence supporting the role of HOXB13 variants is lacking, especially with respect to their functional characteristics, although a few studies focused on functional aspects of G84E [25,26]. P/LP germline variants in BRCA2 are LOF variants caused by frameshift, premature stop, and splice site mutations that lead to protein truncation. These LOF variants explain the risk of developing cancer and are consistent with a tumor suppressor mechanism. They fit the Knudsen “two-hit” hypothesis because most tumors also have a second, somatic LOF alteration. In contrast, HOXB13 germline variants associated with the risk of prostate cancer do not exhibit protein truncation, and somatic LOF has not been demonstrated. Given the role of WT HOXB13 in the metastatic progression of prostate cancer [27–30], the findings support a gain of function in X285K with respect to its role in mediating an aggressive phenotype. The concept of gain-of-function properties associated with a germline HOXB13 variant is entirely novel, and the gain-of-function oncogenic properties of germline variants are also historically understudied [31–33]. Our findings also raise a question as to what other genomic alterations may act in concert with X285K to “activate” its oncogenic activity. Our functional study was limited by the use of a single cell line and the lack of data showing the E2F/MYC-mediated aggressive phenotypes. Further mechanistic analyses and development of this concept will require additional tools and resources (eg, variant-specific antibodies and animal models) when these become available.

5. Conclusions

In summary, large-scale germline genetic testing results accompanied by functional laboratory experiments suggest that HOXB13 X285K is enriched in self-reported Black patients, and the X285K variant protein mediates an aggressive prostate cancer phenotype associated with gain-of-function E2F/MYC activation. Additional clinical and laboratory studies are needed to determine the functional dependency and therapeutic vulnerability of X285K tumors to specific treatments. Nevertheless, these findings underscore the clinical utility of germline genetic testing in patients with prostate cancer and provide new information to facilitate clinical interpretation of genetic testing findings in a population disproportionately affected by excess prostate cancer mortality. Our studies lend evidence to the pathogenicity of the X285K variant, which may factor in the variant classification framework to move the clinical classification of this variant from “likely benign” to “likely pathogenic.” These findings also motivate the identification and development of effective treatments for prostate cancer patients carrying this ancestry-specific mutation.