Multi-omics from a Single Tumor Sample: Discovering Epigenetic Mechanisms for Malignant Peripheral Nerve Tumors Using Covaris’ Ultrasonication Technology
At ASMS 2023, Joanna Lempiäinen, PhD, a Postdoctoral Research Scholar at Ben Garcia’s lab at Washington University in St. Louis (WUSTL), showcased how they are using Covaris ultrasonication and sample prep workflows to simultaneously characterize the total proteome, histone PTMs, and RNA modifications from one FFPE sample of malignant peripheral nerve tumor tissue. They then used the same setup to perform ChIP-seq analysis.
Introduction
Acquiring reliable proteomics and epigenetics data from limited volumes of tissue is essential for understanding dysregulated signaling in diseases like malignant peripheral nerve sheath tumor (MPNST).
MPNST, a form of sarcoma, develops in the connective tissues that surround nerves and is often linked to a genetic condition known as neurofibromatosis type 1 (NF-1). This rare yet aggressive cancer has a tendency to metastasize, resulting in a grim prognosis: most patients die within five years of diagnosis.
Loss of PRC2 Predicts Poor Survival
Secondary mutations, in addition to NF1 mutations, are what drive the development of MPNST. These mutations commonly occur in CDKN2A and p53 genes, as well as subunits of PRC2; a protein complex that binds to chromatin to inhibit gene expression through H3K27me3 deposition. In MPNST, SUZ12 and EED subunits are often mutated and these mutations can co-occur with the CDKN2A and p53 mutations.
However, in some cases, the functionality of the PRC2 complex is retained, offering a glimmer of hope. PRC2 loss, on the other hand, predicts poor survival and the loss of this Histone PTM can occasionally be used to help diagnose or distinguish MPNST from other types of cancer.
Previous research by John Wojcik et al., found that antigen presentation proteins are significantly downregulated in PRC2 loss cases, which may explain why PRC2 loss cases are worse than those where PRC2 is retained. In addition, there’s a noted loss of K27 trimethylation along with an upregulation of several histone PTMs.
While studies have focused on SUZ12 mutants in MPNST, EED subunit mutations are found in about 30% of MPNSTs. Dr. Lempiäinen’s team aimed to compare these two mutant types to pinpoint common pathways that could be targeted in mutant tumors. The goal was to uncover combinations of epigenetic drugs that might effectively treat the more aggressive PRC2 loss MPNST.
In a separate paper performed in collaboration with the Garcia Lab, different methods of protein extraction from FFPE samples were evaluated. The paper’s authors concluded that ultrasonication with Covaris Adaptive Focused Acoustics® technology combined with S-trap was the most efficient for FFPE protein extraction (Figure 1). This method was named HYPERsol, or High Yield Protein Extraction and Recovery by direct solubilization.
Figure 1. Ultrasonication with Covaris Adaptive Focused Acoustics® technology combined with S-trap is most efficient for FFPE protein extraction.
Leveraging Covaris Technology in the Analysis of SUZ12 and EED Role in MPNST
Dr. Lempiäinen’s team, in collaboration with Dr. Angela Hirbe’s lab at WUSTL’s Oncology Division, planned to conduct an immune checkpoint inhibition combined with different epigenetic drugs. Upon completion, tumor samples would be sent to Dr. Lempiäinen’s lab for analysis of the tumor proteome and histone PTMs.
Test samples consisted of FFPE and flash frozen MPNST tumors from humans and PDX mouse sources. The FFPE samples were processed using the HYPERsol published protocol. The protein extraction was followed by characterization of histone PTMs and additional RNA modifications. Frozen samples were simultaneously processed using a bead beater and probe sonicator, and RNA binding columns were applied for extraction of RNA or proteins from a single sample. Final steps were similar to the FFPE workflow (Figure 2).
The team found that the convenience of using Covaris technology was hugely beneficial. During implementation of the HYPERsol protocol, it was discovered that FFPE punches could be directly added into SDS buffer and sonicated for homogenization, eliminating the need for other equipment. The sonication process was so efficient that samples could simply be placed into Covaris tubes, and the instrument would take care of the rest.
Figure 2. FFPE and frozen MPNST tumor tissue samples, from human or PDX mouse, were processed using two different workflows.
Sonication Optimization – Longer Sonication Not Required
The team investigated whether increasing sonication time would enhance protein extraction. Findings revealed that both five-minute and ten-minute sonication times yielded over 3,500 proteins each, with comparable proteome correlation. Therefore, the existing published protocol already maximizes extraction from FFPE, negating the need for longer sonication (Figure 3).
Figure 3. Sonication optimization process and results.
FFPE Proteomes Correlate with Frozen Proteomes
In addition, they compared protein numbers between FFPE and frozen tissue from human and PDX tumors. Results showed a correlation and similar averages for protein group quantity, indicating consistent processing of FFPE samples. The team concluded that FFPE sample preparation was as effective as the frozen method and could be used in future studies (Figure 4).
Figure 4. Comparison of proteins extracted from FFPE and frozen tissue from human and PDX tumors showed correlation between the sample types.
100+ Histone PTMs Identified in a Single Sample
The team carried out an analysis of selected histone PTMs by using total cell lysates and their published protocol, identifying all PTMs from one peptide at the H3-tail. From one sample, they identified more than 100 different histone PTMs or combinations, demonstrating the vast amount of data that can be obtained (Figure 5).
Figure 5. More than 100 histone PTMs/combinations identified in one sample.
The researchers were also successful in identifying various RNA modifications from both the FFPE samples processed with Covaris and the frozen tissue samples, some of which may play a role in cancer progression (Figure 6).
Figure 6. Workflows successfully identify RNA modifications from FFPE and frozen samples.
Easy Integration of Covaris Ultrasonication into ChIP-seq Workflow
Given the interest in epigenetics of MPNST and different histonePTMs, the team performed ChIP-seq on these samples, using Covaris’ E220 Focused-ultrasonicator (Figure 7). Findings revealed that the same consumables used for FFPE extraction could be utilized to perform ChIP-seq, eliminating the need for extra tubes. The settings on the E220 remained the same except for a decrease in temperature due to these cell lines being in a different buffer.
Figure 7. ChIP-seq of histone PTMs in MPNST cell lines using the Covaris E220 Focused-ultrasonicator with Covaris microTUBEs
Conclusion
Most MPNSTs are driven by SUZ12 and EED mutations, resulting in PRC2 loss and major epigenetic changes. Findings from this study indicate that the FFPE proteome prepped with Covaris’ E220 Focused-ultrasonicator correlates well with frozen tissue, validating its continued use. Interestingly, the team was able to obtain histone PTM and RNA modification data from the same sample, enabling them to obtain substantial data from just one source.
Interested in discussing how AFA technology can benefit your protein extraction workflow? Click here to reach out to one of our specialists.