Read the Full Video Transcript

Andrea Miyahira: Hi, everyone. I'm Andrea Miyahira at the Prostate Cancer Foundation. Today I'm talking with Dr. Mayuko Kanayama, a research associate at John Hopkins University, about her group's recent paper, "Clinical and Functional Analyses of an African-ancestry Gain-of-function HOXB13 Variant Implicated in Aggressive Prostate Cancer." This was published in European Urology. Dr. Kanayama, thanks so much for joining me today.

Mayuko Kanayama: Thank you for having me. It's a great honor.


Mayuko Kanayama: Thank you, Andrea, for that kind introduction. I'm Mayuko Kanayama. Please, let me start with a background. So, as recently reported by Dr. Michael Haffner's team, HOXB13 is specifically expressed in the prostates, and its expression is maintained until late-stage metastatic prostate cancer. And we focus on the HOXB13 stop-loss germline variant X285K, in which one nucleotide in its stop codon is deleted, resulting in a 96 amino acid extension at its C-terminal end. And this variant is specifically found in the population of African ancestry. And it was first reported in familial prostate cancer on a small Caribbean island called Martinique.

Then, doctors Bill Isaacs and Chris Haiman's team independently reported that this variant was significantly associated with more aggressive disease, such as diagnosis at an earlier age, or higher Gleason grade, or higher T-stage. And as you can see, this variant came out into the spotlight relatively recently. And just like many other prostate cancer risk-associated germline variants, the biological mechanism of this variant is unknown. So, our study aims to determine the clinical and biological significance of this variant.


So, first, to determine the clinical significance of this variant, we analyzed the real-world data in collaboration with a genetic testing company called Invitae. So, in this study, 21,393 men with prostate cancer underwent pro-gene sequencing, including the recent duplication analysis for cancer-relevant genes, using their blood or saliva samples. And patients were stratified by self-reported ethnicity, and their clinical outcomes, such as Gleason grade and T-stage, were compared among different variants.


And this is the result. In this study, we focused on X285K, as well as G84E and BRCA2, which are known prostate cancer risk-associated germline variants. And first, consistent with previous studies, X285K was significantly enriched in self-reported Black patients, whereas G84E was significantly enriched in self-reported white patients. Second, the prevalence of higher Gleason grade was higher in X285K than in G84E, 62.5% versus 49.3%. And also, the prevalence of stage four disease was significantly higher in X285K than in G84E. And there was statistical significance between these two values. And also, the stage four disease was higher in BRCA2 than in G84E. Here, I want to emphasize that X285K is as impactful as BRCA2 in terms of association with more aggressive disease.


So, moving on to table two. So, one of the limitations of our real-world data analysis is the lack of a control population. So, to address this problem, we came up with an equation to calculate PAR, population attributable risk, using given values. So, what is PAR? PAR is the proportion of the disease incidence that is due to exposure in the entire disease population, including both exposed and unexposed. So, these three values here, on the left, are carrier frequencies in our study, and these values here in the middle, are carrier frequency in the database, which are considered to be carrier frequencies in the control population. And these numbers here are PAR.


So, this 0.62% means that when you look at the entire prostate cancer population, 0.62% of them gets prostate cancer because of this variant. And this number is as significant as G84E. And we shouldn't forget that, as for BRCA2, it's not a single variant, but it's a congregation of pathogenic and likely pathogenic variants. So, the fact that X285K and G84E, as singular variants, show this high PAR is very meaningful. So, to summarize, our findings support the early disease screening for Black men carrying this variant and early treatment escalation when they develop prostate cancer. And patients and clinicians watching this channel may wonder where they can get tested for this variant. And although we don't have information about other genetic testing companies, but Invitae, the company we collaborated with, upgraded this variant from likely benign to increasingly scarier after the publication of this paper. So, if patients are positive for this variant, they'll be informed.

Next, we moved on to functional analysis to determine the biological significance of this variant. So, to do this, I generated that bi-directionally expresses RFP-tagged HOXB13 shRNA together with the wild-type or X285K HOXB13 cDNA to achieve simultaneous knockdown of endogenous wild-type HOXB13 and expression of exogenous HOXB13, and I used several clones for each genotype to ensure reproducibility. And these clones were subjected to RNA-seq, ChIP-seq, and ATAC-seq. This is just to show that our system is clean and working. These are the HOXB13 Western blot and RFP images of representative clones. So, as you can see, exogenous HOXB13 is expressed, and RFP is expressed by Dox treatment, while endogenous HOXB13 is successfully knocked down.


And these clones were subjected to RNA-sequencing first. This is the RNA-seq result. So, on the left is a volcano plot comparing X285K versus wild-type in an androgen-stimulated condition. Interestingly, CCNB1, encoding cyclin B1, and MYC, encoding c-MYC, appeared as upregulated genes by X285K. So, cyclin B1 is an important molecule for cell cycle progression, and MYC is, needlessly to say, an important oncogene for prostate cancer. And consistent with this, gene-set enrichment analysis showed E2F and MYC enrichment in X285K-expressing cells, suggesting that X285K is contributing to cell proliferation. And all these results are validated by Western blot and qRT-PCR.


Next, to further investigate the mechanism, we performed ChIP-sequencing and ATAC sequencing. The ATAC result is not included in the paper, but ATAC is an assay to evaluate chromatin accessibility. So, our ChIP-seq data show that X285K has a distinctive binding profile as compared to wild-type, and it has more binding sites in the genome. So, this is one example. This is the cyclin B1 gene and its 5' upstream region, and Cyclin B1, as I mentioned earlier, is upregulated by X285K. And as you can see in the ChIP-seq data, X285K-bound protein binds more to its 5' upstream region. And if you look at the ATAC-seq data, it opens up the chromatin of the same region. So, this is likely to be the mechanism of upregulation. And we also confirmed increased X285K binding to the MYC super enhancer region, consistent with MYC upregulation.


So, this is a summary of our functional analysis. Our message is pretty straightforward. X285K binds more to its target genes, opens up its chromatin, and increases their expressions, suggesting its role as a gain-of-function mutation, like BRCA2, known germline variants are loss-of-function mutations. So, the concept of a gain-of-function mutation is pretty novel.


As a conclusion, HOXB13 X285K is significantly enriched in self-reported Black patients. And this variant-carrier patients detected in the real-world clinical setting have aggressive prostate cancer features, similar to BRCA2 carriers, supporting early disease screening of Black men carrying this variant. And functional studies revealed a unique gain-of-function oncogenic mechanism of this variant protein in regulating E2F/MYC signatures. With that, I'd like to thank the PCF for funding my project and all the lab members and all the collaborators. Thank you so much for your attention.

Andrea Miyahira: Thank you so much, Dr. Kanayama, for sharing this with us. So, I guess, how frequent is the HOXB13 X285K variant in African ancestry populations overall? And do we know what specific regions of Africa-

Mayuko Kanayama: Yeah, yeah, thank you for asking.

Andrea Miyahira: ... these are more found in?

Mayuko Kanayama: Yeah, yeah, right, right. So, according to Dr. Chris Haiman's paper, this variant is only present in the population of West African ancestry, such as Nigeria, Senegal, and Ghana. And the carrier frequency, by which I mean the population of people who carry this variant, is, according to this paper, is 0.76% in a control population, but it goes up to 2.39% in the prostate cancer population. And our study tested the African-American population with prostate cancer. And the carrier frequency in prostate cancer patients was 1%, as I just presented. And although we don't have a control population, the estimated carrier frequency in the control population was 0.39%. So, it just goes up in the prostate cancer population. Yeah.

Andrea Miyahira: Okay. Thank you. And how should these findings change genetic testing guidelines for prostate cancer risk variants?

Mayuko Kanayama: Oh, it's such an important question. So, as I said, the company we collaborated with, Invitae, upgraded this variant from likely benign to increasingly scarier. So, the change is already happening, and there are two great preceding studies by Dr. Bill Isaacs and Chris Haiman showing the clinical association of this variant with more aggressive disease. And in addition to that, we looked into a biological mechanism, suggesting and showing the evidence of a causal relationship. So, our study, together with previous studies, provides clinical and biological rationale for early disease screening for men, specifically Black men, carrying this variant.

Andrea Miyahira: Thank you. And do you have any more insights into what mechanisms you think this variant is using to drive the increased disease aggressiveness?

Mayuko Kanayama: Yeah, thank you for asking that. So, our ChIP-seq data and ATAC-seq data show that this variant protein has a distinctive binding profile as compared to wild-type. And it binds more to some genes important for the prostate cancer population and opens up its chromatin. And also, our unpublished data suggests that that long tail is physically affecting its function, like pushing away some of the co-factors, affecting the overall epigenome. So, yeah, the mechanism is likely to be a gain-of-function. Yeah.

Andrea Miyahira: Okay, thanks. What are your next steps for these studies?

Mayuko Kanayama: So, yeah, from a clinical perspective, the biggest assignment is to collect more clinical specimens to validate our findings in human samples. And from a research perspective, just like I said, I'm suspecting that the long tail is doing something to physically affect its function. So, I'm really interested in a protein structural analysis, and also, I made a transgenic mouse model expressing this protein in the mouse prostate. So, I'm going to see what kind of phenotypes are induced by this variant protein. It's my future direction.

Andrea Miyahira: Okay. Well, thank you so much for coming on today and sharing this with us.

Mayuko Kanayama: Yeah, yeah. Thank you so much for your time.