2.1 USP19 stabilizes MGMT

Following alkylation stress, MGMT functions as a DNA repair enzyme that specifically transfers alkyl adducts from the O6 position of guanine to the cysteine residue in its active site. Their reversible binding of the alkyl group to MGMT protein functionally inhibits its enzyme activity and leads to protein degradation. Previous studies have reported that MGMT is ubiquitinated and degraded through the ubiquitin–proteasome pathway after it repairs the O6-methylguanine-DNA lesions. However, how the deubiquitinating process regulates MGMT is unclear. To identify potential DUBs that can deubiquitinate and stabilize MGMT, we overexpressed a panel of deubiquitinases in cells individually and examined the MGMT protein level. As shown in Figure 1A, overexpression of ubiquitin-specific protease USP19 could dramatically increase MGMT protein level. To further confirm the relationship between USP19 and MGMT, we used two different specific short hairpin RNAs (shRNAs) to deplete endogenous USP19 in T98G cells and then detected MGMT protein level. We found that knockdown of USP19 dramatically decreased MGMT protein level, which can be rescued by proteasome inhibitor MG132 (Figure 1B). As shown in Figure 1C,D, depletion of endogenous USP19 by two USP19-specific shRNAs markedly decreased protein level, but not the MGMT mRNA level, suggesting that USP19 regulates MGMT through post-translational modification. To further establish whether USP19 regulates MGMT protein stability, we treated cells with cycloheximide (CHX) and determined the half-life of MGMT. As shown in Figure 1E,F, MGMT stability was dramatically decreased in USP19-depleted cells, while reconstitution of USP19 could rescue MGMT protein stability in USP19-depleted cells. Taken together, these results suggest that USP19 regulates MGMT protein level and stability in a proteasome-dependent manner and USP19 is a potential DUB for MGMT.

2.2 USP19 interacts with MGMT

To further investigate the mechanism by which USP19 regulates MGMT protein level, we tested whether the USP19 interacts with MGMT. As shown in Figure 2A,B, immunoprecipitating FLAG-tagged MGMT in 293T cells was able to pull down USP19. Additionally, endogenous immunoprecipitation further confirmed that MGMT binds to USP19 (Figure 2B). To further investigate the mechanism of binding between USP19 and MGMT, we found that the N terminal (1-596aa) and UI domain of USP19 were responsible for binding with MGMT (Figure 2C).

2.3 USP19 deubiquitinates MGMT in vivo and in vitro

To further clarify the effect of USP19-mediated MGMT stabilization, we assessed whether USP19 deubiquitinated MGMT in cells. As shown in Figure 3A, knockdown of USP19 resulted in a dramatic increase in MGMT polyubiquitination in cell. Conversely, overexpressing USP19 WT in cell resulted in a significant decrease in polyubiquitination of MGMT, while overexpressing USP19 CS mutant failed to alter the level of MGMT polyubiquitination (Figure 3B). To further investigate whether USP19 directly deubiquitinated MGMT, we performed an in vitro deubiquitination assay using 293T cell expressed FLAG-USP19, S-MGMT independently and purified by FLAG-tag, S-tag affinity agrose beads, respectively. While WT USP19 removed ubiquitin chain from MGMT, the catalytic inactive CS mutant could not (Figure 3C). These results indicate that USP19 is a deubiquitinase for MGMT.

2.4 USP19 regulates the sensitivity of MGMT-positive GBM cell lines to TMZ through MGMT stabilization

It is also known that MGMT plays an important role in resistance of GBM to TMZ. Thus, our hypothesis is that USP19 regulates GBM response to TMZ through MGMT stabilization. To further confirm whether USP19 affects the GBM sensitivity to TMZ, the colony formation assay was performed by T98G, LN18, U251, and U87 cell lines, and these cell lines were separated into two groups depending on the promoter of MGMT gene status, and we checked the MGMT expression status by western blot (Figure 4A). T98G and LN18 are MGMT-positive cell lines, which had high levels of MGMT activity, and the gene promoter is unmethylated, while U251 and U87 are MGMT-negative cell lines, which had low levels of MGMT activity and the gene promoter is methylated. All four different cell lines with USP19 knockdown and MGMT-positive cell lines with MGMT overexpression after USP19 depletion were employed to detect the response to TMZ. As shown in Figure 4B–E, depletion of USP19 sensitized T98G and LN18 cells but not U251 and U87 cells to TMZ. While restoring MGMT expression in USP19 knockdown, T98G and LN18 cells reversed the sensitivity to TMZ (Figure 4B,C). Taken together, these results indicated that USP19 could regulate the sensitivity of MGMT-positive GBM cell lines to TMZ through MGMT stabilization.

2.5 USP19 regulates response of GBM cells to chemotherapy through MGMT stabilization in xenograft

Previous research has shown that MGMT is linked to the therapeutic success of alkylating agent chemotherapy, specifically TMZ treatment. Furthermore, our results show that USP19-MGMT axis related to the MGMT-positive GBM cell response to TMZ treatment. To gain more proof to certify the biological function of USP19-MGMT axis in GBM cells chemo-resistance in vivo, we employed shRNA to specifically knockdown USP19 in T98G cells, and nude mice were subcutaneously implanted with T98G USP19 depletion cells and control (Ctrl) cells. As shown in Figure 4F–H, depletion of USP19 dramatically sensitized T98G cells to TMZ in flank tumor models. Taken together, our results suggest that USP19 deficiency causes sensitivity of GBM cells to TMZ.

2.6 USP19 is positively correlated to MGMT in clinical GBM cancer samples

The oral alkylating agent, TMZ, is currently used as first-line therapy for GBM patients due to its being bioavailable when taken orally, and it is able to cross the blood–brain barrier. After the small molecular TMZ through in the cerebral, it plays the antitumor growth function due to inducing methylation damage in cancer cell DNA. However, MGMT induces drug resistance by removing DNA methylation, which is a major bottleneck for clinical treatment. Previously, some studies showed that expression of MGMT is high level in TMZ resistance cells, and the MGMT gene promoter methylated is the main reason for expression silence. Apart from this, the mechanism of regulating response of unmethylated GBM cells is still unclear. We were interested in investigating whether USP19 relative MGMT expression in clinic GBM samples. As shown in Figure 5A–C, the expressions of USP19 and MGMT in GBM samples were positively correlated. Since USP19-MGMT axis regulates the sensitivity of GBM to TMZ, we analyzed the overall survival of GBM patients in different USP19 expression backgrounds. As shown in Figure 5D, the data suggest that patients with USP19 alteration have better clinical outcomes. These results suggested that USP19 expression enhanced TMZ stability and induced resistance in GBM, which led to patients having poor treatment outcomes.

4.1 Cell culture, constructs, and antibodies

We purchased glioblastoma cell lines T98G, LN18, and neuroblastoma cell line SK-N-SH from ATCC. All the cell lines were tested and confirmed by the Mayo Clinic Medical Genome Facility Center. FLAG-USP19 WT and C506S were purchased from Addgene, and we subcloned it to pCMV-HA and pLV.3-FLAG vectors. FLAG-MGMT WT was purchased from Genescript, and we subcloned it to pLV.3-FLAG vector. The USP19 constructs were kindly provided by Yihong Ye. U87 and U251 glioblastoma cell lines were kindly provided by Jann N. Sarkaria.

Anti-USP19 (A301-586A) antibody was purchased from Bethyl. Anti-MGMT (E6M7V) antibody was purchased from cell signaling. Anti-Ub (sc-8017) antibody was purchased from Santa Cruz. Antibodies against MGMT (MAB16200), HA (H9658), FLAG (F1804), β-actin (A1978), and β-Tubulin (T8328) were purchased from Sigma.

4.2 RNA interference

USP19 shRNAs were purchased from Sigma: USP19 sh#1, CCGGGCGTGATTTGATTCTGTTGTACTCGAGTACAACAGAATCAAATCACGCTTTTTG; USP19 sh#2, CCGGGCTGATGAACAGCTTTGCATACTCGAGTATGCAAAGCTGTTCATCAGCTTTTTG. We generated the lentiviruses for USP19 shRNAs according to the standard protocol.

4.3 Deubiquitination assay in vivo

HEK293T cells transfected with FLAG-MGMT, and His-Ub as indicated were incubated with 25 μM MG132. After 4 h incubation, the cells were carefully washed with pre-chilled PBS for two times and harvested. Cells were lysed in 8 M Urea, 0.1 M NaH₂PO₄, 300 mM NaCl, and 0.01 M Tris (pH 8.0). Lysates were briefly sonicated to shear DNA and incubated with Ni-NTA agarose beads (QIAGEN) for 1–2 h at Room Temperature. Beads were washed five times with 8 M Urea, 0.1 M NaH₂PO₄, 300 mM NaCl, and 0.01 M Tris (pH 8.0). Input and beads were boiled in loading buffer and subjected to SDS-PAGE and immunoblotting.

4.4 Deubiquitination assay in vitro

To prepare ubiquitinated MGMT as the substrate and USP19 WT/CS for the in vitro deubiquitination assay, HEK293T cells were transfected with USP19 WT, USP19 CS or His-Ub and S-MGMT independently. After 36 h, ubiquitinated MGMT, USP19 WT, and USP19 CS were purified from the cell extracts with anti-S-affinity or anti-FLAG-affinity column in lysis buffer (50 mM Tris–HCl pH 7.8, 137 mM NaCl, 10 mM NaF, 1 mM EDTA, 1% Triton X-100, 0.2% Sarcosyl, 1 mM DTT, 10% glycerol, and fresh proteinase inhibitors). Following extensive washing with the FLAG-lysis buffer, the USP19 WT and USP19 CS proteins were eluted with FLAG-peptides (Sigma). For the in vitro deubiquitination assay, ubiquitinated S-MGMT protein was incubated with USP19 in a deubiquitination buffer (50 mM Tris–HCl pH 8.0, 50 mM NaCl, 1 mM EDTA, 10 mM DTT, 5% glycerol) overnight at 16°C.

4.5 Cell survival assay

The U87, U251, T98G, and LN18 glioblastoma cell lines are stably expressed with the indicated constructs. For clonogenic assay, 500 cells (T98G and LN18), 3000 cells (U87), and 1000 cells (U251) were seeded in 6-well plate for 24 h, and then treated with indicated concentration of temozolomide. After 2 weeks, cells were washed with PBS, fixed with methanol, and stained using 0.1% crystal violet, and the colony numbers were counted.

4.6 Tumor growth study in xenograft

T98G cells stably expressed the control, and USP19 shRNAs were injected subcutaneously into female athymic nu/nu mice (NCI) flank at 1 × 10⁶ cells/mouse by oral gavage at 50 mg/kg in oral plus, while the control group was treated with the same vehicle. The tumor volumes were measured using digital calipers every 3 days. The experiment protocol was approved by the IACUC at Mayo Clinic.

4.7 Tissue microarray

The immunostaining was blindly scored by pathologists. The IHC score is calculated by combining the quantity score (percentage of positive stained cells) with the staining intensity score. The quantity ranges from 0 to 4, that is 0, no immunostaining; 1, 1%–10% of cells are stained; 2, 11%–50% are positive; 3, 51%–80% are positive; and 4, ≥81% of cells are positive. The staining intensity was scored as 0 (negative), 1 (weak), 2 (moderate), and 3 (strong). The score for each tissue was calculated by multiplying the intensity with the quantity score (the range of this calculation was therefore 0–12). An IHC score of 9–12 was considered a strong immunoreactivity; 5–8, moderate; 1–4, weak; and 0, negative. Samples with an IHC score of more than 4 were considered to be high, and less than 4 were considered to be low. The score of tumor tissue was determined by comparing it with the staining intensity of normal tubules on the same slide.

4.8 Statistical analysis

Data are presented as the mean ± s.e.m. of three independent experiments for cell proliferation. Data are presented as the mean ± s.d. (n = 6) for cell survival assay. Data are presented as the mean ± s.d. of five mice for the tumor growth study in xenograft. The Student's t-test and Chi-squared test were utilized for the statistical analyses (*p < 0.05; **p < 0.01; ***p < 0.001).