Abstract
Background Epigenetics may predict treatment sensitivity and clinical course for patients with meningiomas more accurately than histopathology. Nonetheless, targeting epigenetic mechanisms is understudied for pharmacotherapeutic development for these tumors. The bio-molecular insights and potential therapeutic development of meningioma epigenetics led us to investigate epigenetic inhibition in meningiomas.
Methods We screened a 43-tumor cohort using a 139-compound epigenetic inhibitor library to assess sensitivity of relevant meningioma subgroups to epigenetic inhibition. The cohort was composed of 5 cell lines and 38 tumors cultured directly from surgery; mean patient age was 56.6 years ± 13.9 standard deviation. Tumor categories: 38 primary tumors, 5 recurrent; 33 from females, 10 from males; 32 = grade 1; 10 = grade 2; 1 = grade 3.
Results Consistent with our previous results, histone deacetylase inhibitors (HDACi) were the most efficacious class. Panobinostat significantly reduced cell viability in 36 of 43 tumors; 41 tumors had significant sensitivity to some HDACi. G9a inhibition and Jumonji-domain inhibition also significantly reduced cell viability across the cohort; tumors that lost sensitivity to panobinostat maintained sensitivity to either G9a or Jumonji-domain inhibition. Sensitivity to G9a and HDAC inhibition increased with tumor grade; tumor responses did not separate by gender. Few differences were found between recurrent and primary tumors, or between those with prior radiation versus those without.
Conclusions Few efforts have investigated the efficacy of targeting epigenetic mechanisms to treat meningiomas, making the clinical utility of epigenetic inhibition largely unknown. Our results suggest that epigenetic inhibition is a targetable area for meningioma pharmacotherapy.
Keywords: meningioma, screening, HDAC, G9a, Jumonji domain
Introduction
Meningiomas are a common brain tumor, accounting for approximately 33.8% of all primary central nervous system neoplasms. 1 2 3 Despite a clinically benign reputation, low-grade meningiomas cause long-term neurological deficits and decrease overall survival, 4 while higher grade meningiomas have worse outcomes and are responsible for 20 to 30% of all meningioma cases. 5 6 7 8 Given the limited role of pharmacotherapy in the treatment of meningioma and lack of a targeted Food and Drug Administration-approved compound for these tumors, there is a clear need for additional research to help treat this patient population. 9
The importance of epigenetics in meningioma biology, and for predicting tumor clinical behavior, has recently emerged. 10 11 12 13 14 15 16 17 18 Epigenetics refers to a group of processes that regulate the physical accessibility and structure of DNA, which subsequently regulate DNA replication, repair, and transcription. 19 20 21 Dysregulation of these processes can drive meningioma oncogenesis, and has been shown in previous studies to cause transcriptional changes secondary to AKT mutation and NF2 deletion. 10 14 15 16 17 Clinically speaking, CpG island and promoter methylation profiles are known to correlate with tumor recurrence and treatment response more accurately than histopathologic grading using the World Health Organization (WHO) criteria. 18 To date, few efforts have investigated the efficacy of targeting these mechanisms for the treatment of meningiomas, making the clinical utility of epigenetic inhibition largely unknown.
In a prior study, we investigated epigenetic inhibition in meningioma in vitro and found histone deacetylase inhibition (HDAC and HDACi, respectively) may have the potential to treat meningiomas broadly. 22 However, we were previously unable to shed light on important clinical analyses such as tumor grade, prior treatment, recurrence status, or patient gender due to an under-powered cohort. In this study, we present an expanded 43-tumor cohort, screened using a 139-compound epigenetic inhibitor library, and provide insight into the sensitivity of relevant meningioma subgroups to epigenetic inhibition.
Methods
Patient Enrollment, Tissue Collection, and Data
All procedures pertaining to patient consent, enrollment, tissue collection, data acquisition, and reporting were performed in strict accordance with COMIRB protocol #13–3007 at the University of Colorado, Anschutz Medical Campus, as well as IRB protocol #00010924 at the University of Utah. Tumors were collected from the operating rooms at their respective institution and were immediately transported to our laboratories for tissue culture. Patient age, gender, histopathology, and clinical characteristics were recorded; all patient identifiers were removed.
Ethics Approval/Consent to Participate/Consent to Publish
This study was approved by the Colorado Multiple Institutional Review Board (COMIRB Protocol 13–3007), with the following approval statement: “This study was reviewed and approved under the “2018 Requirements” of the Federal Policy for the Protection of Human Subjects.” Approval date: November 14, 2019; no expiration date.
The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of the University of Colorado Anschutz (COMIRB Protocol #13–3007; approval date, November 14, 2019).
Informed Consent Statement
Informed consent was obtained from all subjects involved in the study. Study subjects are de-identified. Consent to publish is acknowledged in the consent form and protocol.
Tissue Culture Methodology
Meningiomas were sectioned into 1 mm × 1 mm pieces in a sterile tissue culture hood in PBS and digested in dispase overnight to generate a single cell suspension. The resulting suspensions were cultured in flasks coated with a decellularized neural matrix thin film. 23 A 200 µL of the thin film solution was added to the bottom of a T175 flask, evenly applied with a cell scraper, and left to dry overnight. The thin film solution was prepared using a 3:1:1 ratio of bovine brain:bovine dura:type 1 collagen. Decellularized bovine brain and dura were prepared as previously described. 23 Briefly, bovine brain or bovine dura was sectioned into 1 cm × 1 cm pieces and suspended in 0.1% weight/vol sodium dodecyl sulfate, sterile PBS, and 1% pen/strep. Solution was changed every day for 20 days until tissue was translucent. Tissue was then washed with sterile PBS, pepsin digested, and the final solution was lyophilized. Meningiomas were cultured in high-glucose DMEM with 15% FBS, 1% pen/strep, and 1% Glutamax, passaged weekly, and media were changed twice per week. All tumors were screened within 3 weeks of the initial surgery, with the exception of the cell lines from the University of Utah.
Compound Screening
A 139-epigenetic compound library, acquired from Cayman Chemical, was used in a 96-well plate format to perform the high-throughput screen. Cells were seeded at a density of 2,500 cells per well into neural matrix coated 96-well plates and dosed with a single treatment of the compound library at a concentration of 5 mM for 72 hours. Each compound was tested in triplicate, and cell viability was assayed using MTS tetrazolium. The MTS reagents were prepared by combining PES powder (0.21 mg/mL) to MTS powder (2 mg/mL) in PBS. The final MTS concentration used was 0.33 mg/mL, and absorbance was read at 490 nm.
Statistics
The R statistical suit was used for all analysis (version 4.0.2, https://www.r-project.org/ ). Statistical significance was defined as p <0.05 for all tests. The Mann–Whitney U-test was used to identify individually significant compounds. For comparisons between two groups, a Mann–Whitney U-test was used, and to determine significance between three or more subgroups, a Kruskal–Wallis test was performed followed by a Wilcoxon rank-sum test to calculate the comparisons between group levels. Data are displayed as the mean ± standard deviation (SD).
Results
Demographics
The full cohort ( Table 1 ) was composed of 43 total tumors, 38 of which came from University of Colorado Anschutz operating rooms and were immediately screened, while 5 were from previously generated cell lines and are denoted as such by an asterisk in Table 1 . One was a grade 3, 10 were grade 2, and 32 were grade 1. Ten tumors came from male patients while 33 came from female patients. Three tumors had prior radiation and 35 had no prior history, while the 5 cell lines were unknown. Thirty-eight tumors were primary and five were recurrent. The mean patient age was 57.0 years ± 14.1 SD, with one cell line coming from a patient of unknown age.
Table 1. Patient demographics in meningioma cohort.
| No. | Sample | Age | Gender | Grade | Histologic subtype | Recurrence | Radiation | Chemotherapy |
|---|---|---|---|---|---|---|---|---|
| 1 | IOMM Lee* | 61 | Male | 3 | Anaplastic | Recurrent | Unknown | Unknown |
| 2 | CH157* | 59 | Female | 2 | Not specified | Primary | Unknown | Unknown |
| 3 | J13–2 | 54 | Female | 2 | Atypical | Primary | No | No |
| 4 | J17–2 | 68 | Female | 2 | Atypical | Primary | No | No |
| 5 | J19–3 | 95 | Female | 2 | Atypical | Primary | No | No |
| 6 | J19–4 | 49 | Female | 2 | Atypical | Recurrent | No | No |
| 7 | J24–5 | 77 | Female | 2 | Atypical | Recurrent | Yes | No |
| 8 | J8–4 | 48 | Female | 2 | Atypical | Primary | No | No |
| 9 | K3–1 | 68 | Female | 2 | Atypical | Primary | No | No |
| 10 | K7–4 | 26 | Male | 2 | Atypical | Primary | No | No |
| 11 | K7–7 | 46 | Male | 2 | Atypical | Recurrent | Yes | Yes |
| 12 | GAR* | 71 | Male | 1 | Not specified | Primary | Unknown | No |
| 13 | H23–3 | 83 | Female | 1 | Small cell with psammoma bodies | Primary | No | No |
| 14 | J10–6 | 54 | Male | 1 | Meningothelial | Primary | No | No |
| 15 | J11–2 | 57 | Female | 1 | Not specified | Primary | No | No |
| 16 | J11–6 | 56 | Female | 1 | Not specified | Primary | No | No |
| 17 | J11–7 | 58 | Male | 1 | Meningothelial | Primary | No | No |
| 18 | J15–4 | 40 | Female | 1 | Myxoid | Primary | No | No |
| 19 | J16–5 | 49 | Female | 1 | Psammomatous | Primary | No | No |
| 20 | J17–7 | 49 | Female | 1 | Not specified | Primary | No | No |
| 21 | J18–6 | 41 | Female | 1 | Meningothelial | Primary | Yes | Yes |
| 22 | J18–7 | 41 | Female | 1 | Meningothelial | Primary | No | No |
| 23 | J19–2 | 44 | Female | 1 | Psammomatous | Primary | No | No |
| 24 | J19–5 | 39 | Female | 1 | Meningothelial | Primary | No | No |
| 25 | J19–6 | 44 | Female | 1 | Transitional | Primary | No | No |
| 26 | J19–7 | 57 | Female | 1 | Meningothelial | Primary | No | No |
| 27 | J2–2 | 65 | Female | 1 | Meningothelial | Primary | No | No |
| 28 | J21–4 | 55 | Female | 1 | Rhabdoid morphology present | Primary | No | No |
| 29 | J21–5 | 65 | Male | 1 | Secretory | Primary | No | No |
| 30 | J21–7 | 51 | Female | 1 | Fibrous | Recurrent | No | No |
| 31 | J24–2 | 77 | Female | 1 | Not specified | Primary | No | No |
| 32 | J4–1 | 63 | Female | 1 | Fibrous | Primary | No | No |
| 33 | J4–4 | 67 | Male | 1 | Fibrous | Primary | No | No |
| 34 | J7–6 | 57 | Female | 1 | Transitional | Primary | No | No |
| 35 | J8–3 | 38 | Female | 1 | Meningothelial | Primary | No | No |
| 36 | J9–3 | 70 | Female | 1 | Psammomatous | Primary | No | No |
| 37 | JEN* | 55 | Male | 1 | Psammomatous | Primary | Unknown | No |
| 38 | K10–1 | 78 | Female | 1 | Meningothelial | Primary | No | No |
| 39 | K4–7 | 61 | Male | 1 | Meningothelial | Primary | No | No |
| 40 | K6–6 | 71 | Female | 1 | Meningothelial | Primary | No | No |
| 41 | K9–6 | 37 | Female | 1 | Meningothelial | Primary | No | No |
| 42 | K9–7 | 52 | Female | 1 | Not specified | Primary | No | No |
| 43 | SAM* | Female | 1 | Psammomatous | Primary | Unknown | No |
Note: Relevant clinical information for each tumor sample denoted by sample identifier code including patient age, tumor grade, histological subtype, recurrence status, history of radiation (yes/no), and prior history of chemotherapy (yes/no). Cell lines acquired from the University of Utah are denoted with an asterisk.
Screening Results
All 43 tumors in our cohort were successfully screened in triplicate across the entire 139-compound library ( Fig. 1a , Supplementary Tables S1 and S2 , available in the online version). These results show clear trends in both individual tumor drug sensitivity profiles, as well as highlight compounds with the ability to inhibit the whole cohort of tumors. To identify compounds that could inhibit the whole cohort of meningiomas, we filtered compounds for those that significantly reduced cell viability when calculated as an average across all tumors ( Fig. 1b , Supplementary Table S3 , available in the online version). We identified seven total compounds that met these criteria ( Fig. 1b , Supplementary Table S3 , available in the online version). Panobinostat was the broadly most effective compound with an average cell viability of 51.2% ± 22.9 SD ( p = 1.08 × 10 −66 ), and significantly reduced cell viability in 36 of 43 individual tumors ( Supplementary Tables S1 and S2 , available in the online version). The remainder of the broadly effective compounds in a descending order of cell viability following panobinostat included LAQ824 (57.7% ± 22.6 SD, p = 1.47 × 10 −68 ), HC-toxin (63.9% ± 26.2 SD, p = 2.5 × 10 −51 ), gemcitabine (72.3% ± 21.3 SD, p = 8.95 × 10 −54 ), JIB-04 (76.3% ± 29.8 SD, p = 5.06 × 10 −25 ), UNC0631(78.6% ± 29.9 SD, p = 4.14 × 10 −20 ), and apicidin (79.6% ± 22.2 SD, p = 1.69 × 10 −28 ). Next, we isolated the top five most effective compounds for individual tumors and plotted these as a heat map for the whole cohort ( Fig. 1c ). Panobinostat was the most effective compound in 19 tumors, JIB-04 was the most effective compound in 7 tumors, LAQ824 was most effective compound in 5 tumors, and UNC0631 was most effective in 5 tumors ( Fig. 1c , Supplementary Table S4 , available in the online version).
Fig. 1.
Full meningioma cohort sensitivity to epigenetic compound library. (a) Heat map of the full tumor cohort ( n = 43) screened with the 139-compound epigenetic library shown on a scale from 0 to 100% cell viability. (b) Broadly effective compounds were determined as compounds that significantly ( p < 0.05) reduce the average viability of the cohort to 80% or less and are displayed as a rank-ordered heat map by the most effective compound across the entire cohort. Each compound is displayed with a table showing the average cohort viability (Avg), standard deviation (SD), p -value ( p ), and number of samples the compound significantly reduced viability in (Num Sig). Compounds are colored by mechanism of action, including histone deacetylase (HDAC), Jumonji histone demethylase (Jumonji Domain), and G9a histone methyltransferase (G9a). (c) The top five most effective compounds for each tumor sample, determined as significantly reducing viability the most, are displayed as the filled tiles in the heat map. Compounds that are not in the top five are shown as white space.
Tumors that Lack Sensitivity to Panobinostat
Given that panobinostat significantly reduced cell viability in 36 of 43 tumors and was the most effective compound in 19 of these tumors, we wanted to examine tumors that lacked sensitivity to panobinostat to understand meningioma sensitivity to other compounds and mechanisms that may overcome resistance to HDACi. We used stricter criteria to define sensitivity to panobinostat by including a viability reduction to at least 80%, as well as statistical significance ( p < 0.05), giving us a smaller cohort of 35 tumors having sensitivity to panobinostat ( Fig. 2a ) and 8 tumors that did not ( Fig. 2b ).
Fig. 2.
Meningioma cohort separated by sensitivity to panobinostat.(a) Tumor samples sensitive to panobinostat ( n = 35) were determined as those in which panobinostat significantly ( p < 0.05) reduced individual cell viability to 80% or less. (b) Meningioma samples that did not meet these criteria were considered as panobinostat nonsensitive ( n = 8). Of the panobinostat nonsensitive group (b), there were two tumor samples that were only sensitive to other compounds (c) , displayed with the most effective compound for each tumor: OTX015 was the most effective compound for J4–1 and UNC0631 was the most effective compound for K9–7. Broadly effective compounds are colored by mechanism of action.
The cohort that lacked sensitivity to panobinostat still displayed sensitivity to HDACi in 4 of the 8 tumors and remained sensitive to Jumonji-domain inhibition by JIB-04 in 4 of 8 tumors ( p = 0.077, compared with the whole cohort) as well as G9a inhibition by UNC0646 in 4 of 8 tumors ( p = 0.43, compared with the whole cohort), which is a similar efficacy to the full 43-tumor cohort ( Fig. 2a , b and Supplementary Table S5 , available in the online version). Of the two tumors that appear to lack sensitivity to any inhibitor in Fig. 2b , one (K9–7) does have significant sensitivity to a different G9a inhibitor, UNC0631, while the other tumor (J4–1) was sensitive to bromo-domain inhibition by OTX-015 ( Fig. 2c ). Taken together, of the 8 tumors that were not sensitive to panobinostat, 4 were sensitive to other HDACi, while 5 of 8 were sensitive to G9a inhibition (UNC0646 and UNC0631), and 4 were sensitive to Jumonji-domain inhibition ( Supplementary Table S5A , B , available in the online version). Notably, the two tumors in Fig. 2c were both primary, grade 1 tumors from female patients.
Tumor Grade
Tumors were grouped by WHO grade and categorized as either grade 1, 2, or 3. The most effective compound across grade 1 tumors was panobinostat (52.1% ± 24.3 SD, p = 4.83 × 10 −47 ; Fig. 3a ). In total, 6 compounds were identified that significantly reduced viability in grade 1 tumors, 3 of which were HDACi, 1 was a Jumonji-domain inhibitor, 1 was a G9a inhibitor, and 1 was a nucleic acid synthesis inhibitor ( Fig. 3a , Supplementary Table S6A , available in the online version). The most effective compound across grade 2 tumors was LAQ824 (47.9% ± 15 SD, p = 1.94 × 10 −23 , Fig. 3b ). Seven compounds were identified that significantly inhibit grade 2 cell viability, five of which were HDACi, one was a Jumonji-domain inhibitor, and one was a nucleic acid synthesis inhibitor ( Fig. 3b , Supplementary Table S6B , available in the online version). Finally, for the single grade 3 tumor, UNC0631 was the most effective compound (22.1% ± 1.2 SD, p = 8.15 × 10 −4 , Fig. 2c ). Of the top 15 significant compounds against this tumor, 7 were HDACi, and 5 were G9a inhibitors ( Fig. 2c , Supplementary Table S6C , available in the online version).
Fig. 3.
Meningioma cohort compared by tumor grade. Heat maps displaying the broadly most effective compounds for grade 1 (a) , grade 2 (b) , and grade 3 (c) meningiomas. Broadly effective compounds for each tumor grade were determined as compounds that significantly ( p < 0.05) reduce the average viability to 80% or less and are limited to the top 15 most effective compounds, when applicable. Heat maps are displayed with a table showing the average cohort viability (Avg), standard deviation (SD), p -value ( p ), and number of samples each compound significantly reduced viability in (Num Sig). Compounds are colored by mechanism of action.
Comparing sensitivity to epigenetic inhibition across grade, HDACi significantly decreased cell viability to a greater degree as grade increased in 8 of 10 HDACi identified as the most effective compounds across grade ( Supplementary Fig. S1a , available in the online version). G9a inhibition also significantly decreased cell viability to a greater extent as grade increased in four of five compounds ( Supplementary Fig. S1b , available in the online version). Jumonji-domain inhibition, while effective in grade 1 and 2 tumors, was not significantly different across grade.
Recurrence
Tumors were classified either primary or recurrent to understand the differences in recurrence status in relation to epigenetic inhibition. Six total compounds were found to significantly reduce the average viability in primary tumors to 80% or less, three of which were HDACi, one was a nucleic acid synthesis inhibitor, one was G9a, and one was a Jumonji-domain inhibitor ( Fig. 4a , Supplementary Table S7A , available in the online version). The most effective compound was panobinostat (52.0% ± 23.5 SD, p = 1.79 × 10 −57 ). Eleven compounds were identified that significantly reduced viability in the recurrent tumor cohort ( Fig. 4b ); 6 were HDACi, 2 were G9a, 1 was a Jumonji-domain inhibitor, 1 was a nucleic acid syntheses inhibitor, and 1 was a menin inhibition ( Fig. 4b , Supplementary Table S7B , available in the online version). The most effective compound was panobinostat (44.8% ± 16.9 SD, p = 6.3 × 10 −11 ). No significant differences in the top compounds reported were found between recurrent and primary tumors ( Supplementary Fig. S2 , available in the online version).
Fig. 4.
Meningioma cohort compared across primary and recurrent tumors. Heat maps displaying the broadly most effective compounds for primary (a) and recurrent (b) meningiomas in our cohort. Broadly effective compounds for primary and recurrent tumors were determined as compounds that significantly ( p < 0.05) reduce the average viability to 80% or less and are limited to the top 15 most effective compounds, when applicable. Heat maps are displayed with a table showing the average cohort viability (Avg), standard deviation (SD), p -value ( p ), and number of samples each compound significantly reduced viability in (Num Sig). Compounds are colored by mechanism of action.
History of Radiation
To understand the differences in prior radiation on sensitivity to epigenetic inhibition, we grouped tumors by those that had prior radiation, those that had not received prior radiation, or those that were unknown. Panobinostat was the most effective compound in tumors without a history of radiation (52.5% ± 22.7 SD, p = 6.68 × 10 −57 , Fig. 5a ). In total, 5 compounds significantly reduced cell viability to 80% or less in tumors without prior radiation; 3 were HDACi, 1 was a Jumonji-domain inhibitor, and 1 was a nucleic acid synthesis inhibitor ( Fig. 5a , Supplementary Table S8A , available in the online version). For the cohort with prior radiation, panobinostat was also the most effective compound (35.8% ± 4.6 SD, p = 3.77 × 10 −7 ) ( Fig. 5b ). Nine compounds were identified that significantly reduced cell viability in tumors with a history of prior radiation; four were HDACi, one was a Jumonji-domain inhibitor, and one was a nucleic acid synthesis inhibitor ( Fig. 5b , Supplementary Table S8B , available in the online version). Few differences were found between tumors with prior radiation compared with those without a history of radiation; however, panobinostat was more effective in tumors with prior radiation ( p = 0.023, Supplementary Fig. S3A , available in the online version), while a G9a inhibitor was more effective in tumors without a history of radiation ( p = 0.045, Supplementary Fig. S3B , available in the online version).
Fig. 5.
Meningioma cohort compared by history of radiation. Heat maps displaying the broadly most effective compounds for meningiomas that have no history of radiation (a) and prior history of radiation (b) . Broadly effective compounds for no and prior radiation groups were determined as compounds that significantly ( p < 0.05) reduce the average viability to 80% or less and are limited to the top 15 most effective compounds, when applicable. Heat maps are displayed with a table showing the average cohort viability (Avg), standard deviation (SD), p -value ( p ), and number of samples each compound significantly reduced viability in (Num Sig). Compounds are colored by mechanism of action.
Gender
The final clinical variable we investigated was gender. The most effective compound in tumors from female patients was panobinostat (53.1% ± 23.0 SD, p = 1 × 10 −52 , Fig. 6a ). Five compounds were found to significantly reduce tumor viability to 80% or less in tumors from female patients, three of which were HDACi, one was a Jumonji-domain inhibitor, and one was a nucleic acid synthesis inhibitor ( Fig. 6a , Supplementary Table S9A , available in the online version). The most effective compound in male patients was also panobinostat (45.4% ± 21.8 SD, p = 5.78 × 10 −16 , Fig. 6b ). Of the 15 most effective compounds that significantly reduced cell viability in male patients, eight of which were HDACi, and four were G9a inhibitors, among other mechanisms ( Fig. 6b , Supplementary Table S9B , available in the online version). In general, tumors from male patients were more sensitive to epigenetic inhibition than tumors from female patients across many mechanisms, denoted by a greater reduction in viability in 8 of 10 HDACi, 3 of 5 G9a inhibitors, menin inhibition, and DNMT1 inhibition ( Supplementary Fig. S4 , available in the online version).
Fig. 6.
Meningioma cohort compared across patient gender. Heat maps displaying the broadly most effective compounds for meningiomas from male (a) and female (b) patients. Broadly effective compounds for tumors from male and female patients were determined as compounds that significantly ( p < 0.05) reduce the average viability to 80% or less and are limited to the top 15 most effective compounds, when applicable. Heat maps are displayed with a table showing the average cohort viability (Avg), standard deviation (SD), p -value ( p ), and number of samples each compound significantly reduced viability in (Num Sig). Compounds are colored by mechanism of action.
Discussion
In this study we present a 43-meningioma tumor cohort screened against a 139-compound epigenetic inhibitor library. This cohort represents a large increase in total tumors from our prior study, 22 allowing us to investigate clinically relevant subgroups. Overall, HDACi was the most effective class of inhibitor in the cohort, with panobinostat showing the greatest efficacy and significantly reducing cell viability in 36 of 43 tumors. In total, 41 tumors in the cohort were sensitive to HDACi, further supporting the therapeutic potential of this class of molecules for the treatment of meningioma. Jumonji-domain inhibition and G9a inhibition also significantly reduced cell viability across the cohort, indicating these mechanisms play an important role in meningioma biology and represent additional therapeutic targets.
Looking across tumor grade, HDACi decreased tumor viability with increasing grade, indicating HDAC activity is important not only for meningioma biology, but also for disease progression. In a similar pattern, we also identify that G9a inhibition decreases viability with increasing tumor grade, suggesting the combination of HDAC activity and G9a may play a role in tumor progression through histone deacetylation and subsequent di-methylation of H3K9me2, which is an epigenetic change required to coordinate cell cycle progression. 24 Patients with higher grade meningiomas are obviously in need of further treatment options, and our study suggests that emphasis on this cohort with HDAC inhibition may hold promise. Further, one-quarter to one-third of patients with meningioma will have recurrence within 15 years, so we may need to consider upfront therapies for these patients. Our data offer a rationale for potential therapeutics to avoid the multiple recurrences that these patients often experience.
Compared with our prior study, we currently present three meningiomas with prior radiation, allowing us to investigate differences more accurately in epigenetic sensitivity based on radiation history. Interestingly, only panobinostat and the JAK2 inhibitor, lestaurtinib, showed greater sensitivity in tumors with prior radiation, while all other compounds showed no difference. It is important to note that lestaurtinib is known to work upstream of KNT2A, a set-domain methyltransferase responsible for methylating H3K4me3, and has shown efficacy in KNT2A dysregulated cancers. 25 Surprisingly, male gender was the only other clinical variable that showed significantly different tumor sensitivity profiles and demonstrated greater reduction in tumor viability to epigenetic inhibition compared with tumors from females. Prior studies have noted a correlation between male gender and tumor grade and given the correlation to HDACi and G9a inhibition in both higher grade and male tumors in this study, epigenetic dysregulation may explain this observation.
The effect of epigenetic dysregulation in meningioma is an emerging field. Jumonji-domain deletions have been previously reported in meningioma cell lines, 26 though the effects of this are still unknown. In contrast, over-activity of Jumonji-domain proteins and subsequent demethylation of H3K27me3 has been implicated in meningioma recurrence. 27 28 29 30 Interestingly, little is known about G9a, and its target histone modification H3K9me2, in the biology of meningioma, making this a novel area of investigation. As the field of meningioma epigenetics continues to evolve, it is important to relate this body of work to its therapeutic potential as treatments for these tumors.
Conclusion
The pathobiology of meningiomas is closely linked to epigenetic mechanisms, suggesting that this may be a target-rich environment for known inhibitors. Our study here employing a large cohort of cultured tumors covering numerous clinical variables for treatment with this inhibitor class demonstrates the potential utility of epigenetic inhibition for meningioma pharmacotherapy.
Funding Statement
Funding Funding for this work was acquired through Meningioma Mommas and the Department of Neurosurgery at the University of Colorado, Anschutz Medical Campus. Tissue was acquired by the neurosurgeons and Department of Neurosurgery Team through the Neurosurgery Neural Tissue Bank at the University of Colorado Anschutz Medical Campus. Cell lines were donated by the University of Utah. The authors would like to thank the patients who selflessly consented to donate their tissue for this work.
Conflict of Interest None declared.
Co-first Authors.
Institutional Review Board Approval
The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of the University of Colorado Anschutz (COMIRB Protocol #13–3007; approval date: November 14, 2019).
Authors' Contributions
Philip D. Tatman: conceptualization, methodology, validation, formal analysis, investigation, data curation, writing—original draft, writing—review and editing, visualization. Tadeusz H. Wroblewski: conceptualization, methodology, validation, formal analysis, investigation, data curation, writing—original draft, writing—review and editing, visualization. Anthony R. Fringuello: methodology, data curation, investigation, writing—review and editing. Samuel R. Scherer: methodology, data curation, investigation, writing—review and editing. William B. Foreman: methodology, data curation, investigation, writing—review and editing. Denise M. Damek: conceptualization, formal analysis, resources, writing—review and editing. Kevin O. Lillehei: conceptualization, resources, writing—review and editing, funding acquisition. Randy L. Jensen: conceptualization, resources, writing—review and editing. A. Samy Youssef: conceptualization, resources, writing—review and editing. D. Ryan Ormond: conceptualization, resources, writing—original draft, writing—review and editing, supervision. Michael W. Graner: conceptualization, resources, writing—original draft, writing—review and editing, supervision, project administration, funding acquisition.
All authors have read and agreed to the published version of the manuscript.
Supplementary Material
References
- 1.Ostrom Q T, Gittleman H, Xu Jet al. CBTRUS Statistical Report: primary brain and other central nervous system tumors diagnosed in the United States in 2009-2013 Neuro-oncol 201618(suppl_5):v1–v75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ostrom Q T, Gittleman H, Liao Pet al. CBTRUS Statistical Report: primary brain and other central nervous system tumors diagnosed in the United States in 2010-2014 Neuro-oncol 201719(suppl_5):v1–v88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ostrom Q T, Gittleman H, Truitt G, Boscia A, Kruchko C, Barnholtz-Sloan J S.CBTRUS Statistical Report: primary brain and other central nervous system tumors diagnosed in the United States in 2011-2015 Neuro-oncol 201820(suppl_4):iv1–iv86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.van Alkemade H, de Leau M, Dieleman E M et al. Impaired survival and long-term neurological problems in benign meningioma. Neuro-oncol. 2012;14(05):658–666. doi: 10.1093/neuonc/nos013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Choy W, Kim W, Nagasawa D et al. The molecular genetics and tumor pathogenesis of meningiomas and the future directions of meningioma treatments. Neurosurg Focus. 2011;30(05):E6. doi: 10.3171/2011.2.FOCUS1116. [DOI] [PubMed] [Google Scholar]
- 6.Rogers L, Barani I, Chamberlain M et al. Meningiomas: knowledge base, treatment outcomes, and uncertainties. A RANO review. J Neurosurg. 2015;122(01):4–23. doi: 10.3171/2014.7.JNS131644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Aghi M K, Carter B S, Cosgrove G Ret al. Long-term recurrence rates of atypical meningiomas after gross total resection with or without postoperative adjuvant radiation Neurosurgery 2009640156–60., discussion 60 [DOI] [PubMed] [Google Scholar]
- 8.Jääskeläinen J. Seemingly complete removal of histologically benign intracranial meningioma: late recurrence rate and factors predicting recurrence in 657 patients. A multivariate analysis. Surg Neurol. 1986;26(05):461–469. doi: 10.1016/0090-3019(86)90259-4. [DOI] [PubMed] [Google Scholar]
- 9.Nigim F, Wakimoto H, Kasper E M, Ackermans L, Temel Y. Emerging medical treatments for meningioma in the molecular era. Biomedicines. 2018;6(03):E86. doi: 10.3390/biomedicines6030086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Petrilli A M, Fernández-Valle C. Role of Merlin/NF2 inactivation in tumor biology. Oncogene. 2016;35(05):537–548. doi: 10.1038/onc.2015.125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Al-Mefty O, Kadri P A, Pravdenkova S, Sawyer J R, Stangeby C, Husain M. Malignant progression in meningioma: documentation of a series and analysis of cytogenetic findings. J Neurosurg. 2004;101(02):210–218. doi: 10.3171/jns.2004.101.2.0210. [DOI] [PubMed] [Google Scholar]
- 12.Barski D, Wolter M, Reifenberger G, Riemenschneider M J. Hypermethylation and transcriptional downregulation of the TIMP3 gene is associated with allelic loss on 22q12.3 and malignancy in meningiomas. Brain Pathol. 2010;20(03):623–631. doi: 10.1111/j.1750-3639.2009.00340.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bello M J, Amiñoso C, Lopez-Marin I et al. DNA methylation of multiple promoter-associated CpG islands in meningiomas: relationship with the allelic status at 1p and 22q. Acta Neuropathol. 2004;108(05):413–421. doi: 10.1007/s00401-004-0911-6. [DOI] [PubMed] [Google Scholar]
- 14.He S, Pham M H, Pease M et al. A review of epigenetic and gene expression alterations associated with intracranial meningiomas. Neurosurg Focus. 2013;35(06):E5. doi: 10.3171/2013.10.FOCUS13360. [DOI] [PubMed] [Google Scholar]
- 15.Bush M L, Oblinger J, Brendel V et al. AR42, a novel histone deacetylase inhibitor, as a potential therapy for vestibular schwannomas and meningiomas. Neuro-oncol. 2011;13(09):983–999. doi: 10.1093/neuonc/nor072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chen C S, Weng S C, Tseng P H, Lin H P, Chen C S. Histone acetylation-independent effect of histone deacetylase inhibitors on Akt through the reshuffling of protein phosphatase 1 complexes. J Biol Chem. 2005;280(46):38879–38887. doi: 10.1074/jbc.M505733200. [DOI] [PubMed] [Google Scholar]
- 17.Rong R, Tang X, Gutmann D H, Ye K. Neurofibromatosis 2 (NF2) tumor suppressor merlin inhibits phosphatidylinositol 3-kinase through binding to PIKE-L. Proc Natl Acad Sci U S A. 2004;101(52):18200–18205. doi: 10.1073/pnas.0405971102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sahm F, Schrimpf D, Stichel D et al. DNA methylation-based classification and grading system for meningioma: a multicentre, retrospective analysis. Lancet Oncol. 2017;18(05):682–694. doi: 10.1016/S1470-2045(17)30155-9. [DOI] [PubMed] [Google Scholar]
- 19.Hatada I.The epigenomics of cancerIn: Cho WC, ed. An Omics Perspective on Cancer Research.New York, NY: Springer; 201051–67. [Google Scholar]
- 20.Perdigoto C N. Epigenetic cancer evolution, one cell at a time. Nat Rev Genet. 2019;20(08):434–435. doi: 10.1038/s41576-019-0143-1. [DOI] [PubMed] [Google Scholar]
- 21.Galani V, Lampri E, Varouktsi A, Alexiou G, Mitselou A, Kyritsis A P. Genetic and epigenetic alterations in meningiomas. Clin Neurol Neurosurg. 2017;158:119–125. doi: 10.1016/j.clineuro.2017.05.002. [DOI] [PubMed] [Google Scholar]
- 22.Tatman P D, Wroblewski T H, Fringuello A R et al. High-throughput mechanistic screening of epigenetic compounds for the potential treatment of meningiomas. J Clin Med. 2021;10(14):3150. doi: 10.3390/jcm10143150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.DeQuach J A, Yuan S H, Goldstein L S, Christman K L.Decellularized porcine brain matrix for cell culture and tissue engineering scaffolds Tissue Eng Part A 201117(21–22):2583–2592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Park J A, Kim A J, Kang Y, Jung Y J, Kim H K, Kim K C. Deacetylation and methylation at histone H3 lysine 9 (H3K9) coordinate chromosome condensation during cell cycle progression. Mol Cells. 2011;31(04):343–349. doi: 10.1007/s10059-011-0044-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Brown P A, Kairalla J A, Hilden J M et al. FLT3 inhibitor lestaurtinib plus chemotherapy for newly diagnosed KMT2A-rearranged infant acute lymphoblastic leukemia: Children's Oncology Group trial AALL0631. Leukemia. 2021;35(05):1279–1290. doi: 10.1038/s41375-021-01177-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mei Y, Bi W L, Greenwald N F et al. Genomic profile of human meningioma cell lines. PLoS One. 2017;12(05):e0178322. doi: 10.1371/journal.pone.0178322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Jung M, Kim S I, Lim K Y et al. The substantial loss of H3K27me3 can stratify risk in grade 2, but not in grade 3 meningioma. Hum Pathol. 2021;115:96–103. doi: 10.1016/j.humpath.2021.06.005. [DOI] [PubMed] [Google Scholar]
- 28.Katz L M, Hielscher T, Liechty B et al. Loss of histone H3K27me3 identifies a subset of meningiomas with increased risk of recurrence. Acta Neuropathol. 2018;135(06):955–963. doi: 10.1007/s00401-018-1844-9. [DOI] [PubMed] [Google Scholar]
- 29.International Consortium on Meningiomas . Behling F, Fodi C, Gepfner-Tuma I et al. H3K27me3 loss indicates an increased risk of recurrence in the Tübingen meningioma cohort. Neuro-oncol. 2021;23(08):1273–1281. doi: 10.1093/neuonc/noaa303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Gauchotte G, Peyre M, Pouget C et al. Prognostic value of histopathological features and loss of H3K27me3 immunolabeling in anaplastic meningioma: a multicenter retrospective study. J Neuropathol Exp Neurol. 2020;79(07):754–762. doi: 10.1093/jnen/nlaa038. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.