Every fourth woman suffers from uterine leiomyomas (ULs) -benign uterine wall tumors -at some point in premenopausal life. ULs can cause excessive bleeding, pain and infertility 1 , and are the most common cause of hysterectomy 2 . They emerge through at least three distinct genetic drivers: MED12 or FH mutation, or HMGA2 genomic rearrangement 3 . Here we created large genome-wide data sets by DNA, RNA, ATAC, ChIP and HiChIP sequencing of primary tissues to profoundly understand UL genesis. We discovered somatic mutations in genes encoding six members of the SRCAP histone loading complex 4 , and found that germline mutations in SRCAP complex members YEATS4 and ZNHIT1 predispose to UL. The mutant tumors displayed defective histone variant H2A.Z deposition. In ULs, H2A.Z occupancy correlated positively with chromatin accessibility and gene expression, and negatively with DNA methylation, but these correlations were weak in SRCAP complex mutant tumors. In such tumors, open chromatin emerged at transcription start sites where H2A.Z was lost, associated with upregulation of genes. Furthermore, YEATS4 defects associated with abnormal upregulation of bivalent embryonic stem cell genes, as previously shown in mice 5 . Our work describes a potential novel mechanism of tumorigenesis -epigenetic instability caused by deficient H2A.Z deposition -and suggests that ULs arise through an aberrant differentiation program driven by deranged chromatin, emanating from a small number of mutually exclusive genetic driver mutations.

Hereditary leiomyomatosis and renal cell cancer (HLRCC) is a tumor predisposition syndrome caused by germline fumarate hydratase (FH) mutations and characterized by uterine and cutaneous leiomyomas and renal cell cancer. Currently, there is no generally approved method to differentiate FH-deficient uterine leiomyomas from other leiomyomas. Here, we analyzed 3 antibodies (S-(2-succino)-cysteine [2SC], aldo-keto reductase family 1, member B10 [AKR1B10], and FH) as potential biomarkers. The study consisted of 2 sample series. The first series included 155 formalin-fixed paraffin-embedded uterine leiomyomas, of which 90 were from HLRCC patients and 65 were sporadic. The second series included 1590 unselected fresh frozen leiomyomas. Twenty-seven tumors were from known HLRCC patients, while the FH status for the remaining 1563 tumors has been determined by copy number analysis and Sanger sequencing revealing 45 tumors with monoallelic (n = 33) or biallelic (n = 12) FH loss. Altogether 197 samples were included in immunohistochemical analyses: all 155 samples from series 1 and 42 available corresponding formalin-fixed paraffin-embedded samples from series 2 (15 tumors with monoallelic and 7 with biallelic FH loss, 20 with no FH deletion). Results show that 2SC performed best with 100% sensitivity and specificity. Scoring was straightforward with unambiguously positive or negative results. AKR1B10 identified most tumors accurately with 100% sensitivity and 99% specificity. FH was 100% specific but showed slightly reduced 91% sensitivity. Both FH and AKR1B10 displayed also intermediate staining intensities. We suggest that when patient's medical history and/or histopathologic tumor characteristics indicate potential FH-deficiency, the tumor's FH status is determined by 2SC staining. When aberrant staining is observed, the patient can be directed to genetic counseling and mutation screening.

Mechanical forces in a constrained cellular environment were recently established as a facilitator of chromosomal damage. Whether this could contribute to tumorigenesis is not known. Uterine leiomyomas are common neoplasms that display relatively few chromosomal aberrations. We hypothesized that if mechanical forces contribute to chromosomal damage, signs of this could be seen in uterine leiomyomas from parous women. We examined the karyotypes of 1946 tumors, and found a striking overrepresentation of chromosomal damage associated with parity. We then subjected myometrial cells to physiological forces similar to those encountered during pregnancy, and found this to cause DNA breaks and a DNA repair response. While mechanical forces acting in constrained cellular environments may thus contribute to neoplastic degeneration, and genesis of uterine leiomyoma, further studies are needed to prove possible causality of the observed association. No evidence for progression to malignancy was found.

Precision medicine carries great potential for management of all tumor types. The aim of this retrospective study was to investigate if the two most common genetically distinct uterine fibroid subclasses, driven by aberrations in MED12 and HMGA2 genes, respectively, influence response to treatment with the progesterone receptor modulator ulipristal acetate. Changes in diameter and mutation status were derived for 101 uterine fibroids surgically removed after ulipristal acetate treatment. A significant difference in treatment response between the two major subclasses was detected. MED12 mutant fibroids had 4.4 times higher odds of shrinking in response to ulipristal acetate treatment as compared to HMGA2 driven fibroids (95% confidence interval 1.37–13.9; P = 0.013), and in a multivariate analysis molecular subclassification was an independent predictive factor. Compatible with this finding, gene expression and DNA methylation analyses revealed subclass specific differences in progesterone receptor signaling. The work provides a proof-of-principle that uterine fibroid treatment response is influenced by molecular subclass and that the genetic subclasses should be taken into account when evaluating current and future uterine fibroid therapies.