Expression of FSHR and HK2 in A2780 and A2780CP ovarian cancer cells
To screen the appropriate cell lines, we detected the expression of FSHR and HK2 in A2780 ovarian cancer and A2780CP cisplatin-resistant derivative cells. As shown in Figure 1A, both A2780 and A2780CP cells expressed FSHR and HK2 protein. FSHR and HK2 expression were also detected in the tumor xenografts of nude mice by immunohistochemistry. The A2780 and A2780CP tumor xenografts both expressed FSHR and HK2 protein (Figure 1B).
Preparation and characterization of nanoparticle complexes
FSH-derived peptides have been used for targeted therapy of ovarian cancer in terms of the limited expression of FSHR [24, 29]. The synthetic peptide of the FSH β chain sequence (amino acids 33-53) can bind FSHR by mimicking FSH . We previously revealed the high selectivity of drug delivery mediated by FSH β 33-53 peptides . However, natural L-amino acids are not stable enough in the bloodstream. Studies have shown that D-amino acids based on a retro-inverso design are resistant to endogenous proteolysis and have long-lasting effects [31-32]. The retro-inverso FSH β 33-53 peptides exhibited similar advantages in our previous study . To enhance the selectivity of PEG-PEI (PP) as an HK2 shRNA carrier, we synthesized FSH β 33-53 peptide-conjugated PEG-PEI (FPP) and retro-inverso FSH β 33-53 peptide-conjugated PEG-PEI (DFPP). The copolymers were characterized by 1H-NMR. The spectra of FPP and DFPP simultaneously showed the peaks of the peptides, PEG and PEI (Figure 2A), which indicated the successful synthesis of the copolymers.
Then, the copolymers and HK2 shRNA plasmids were combined via electrostatic interactions at different N/P ratios and named PP-shHK2, FPP-shHK2 and DFPP-shHK2. Agarose gel electrophoresis was used to investigate the combination of plasmids and copolymer (Figure 2B). The loading ability of the copolymers increased as the N/P ratio increased. When the N/P ratio was higher than or equal to 10, the plasmids could be completely loaded by the copolymers. Nanoparticle complexes with an N/P ratio of at least 25 had higher transfection efficiency and were more stable in our previous studies. Thus, an N/P ratio of 25 was used in subsequent experiments. The average particle size of PP-shHK2 was 125.7 ± 42.35 nm, and the average particle size of FPP-shHK2 and DFPP-shHK2 were 152.4 ± 20.15 nm and 163.4 ± 26.45 nm, respectively. The zeta potential values of PP-shHK2, FPP-shHK2 and DFPP-shHK2 were 42.30 ± 2.27 mV, 26.70 ± 2.61 mV and 29.40 ± 1.67 mV, respectively. All of the complexes exhibited relatively homogeneous spherical shapes under transmission electron microscopy (Figure 2C).
Retro-inverso FSH β 33-53 peptide enhanced the gene knockdown effects of HK2 shRNA-loaded nanoparticles in vitro
To screen for the most effective HK2 shRNA sequence, we detected the knockdown effects of four sequences in ovarian cancer cells by means of transfection reagents. As shown in Figure 3A and 3B, shHK2-1 most effectively reduced the expression levels of HK2 mRNA and protein compared with the control treatment and the other three shRNA sequences in A2780 and A2780CP cells. Thus, the shHK2-1 plasmid was loaded into PEG-PEI copolymers (named PP-shHK2, FPP-shHK2 and DFPP-shHK2) and used in subsequent experiments.
Then, the knockdown effects of nanoparticle complexes on HK2 were confirmed in ovarian cancer cells (Figure 3C and 3D). Both A2780 and A2780CP cells exhibited a decrease in HK2 levels after treatment with nanoparticle complexes containing HK2 shRNA (shHK2-1 sequence). Treatment with the naked shHK2 plasmid did not reduce HK2 expression when Lipofectamine 2000 was not added to the culture medium. Even PP-shHK2, which was without FSH peptide modification, could reduce the level of HK2 to some extent, which was consistent with the high transfection efficiency of PEI as a polymeric vector . The attachment of targeting ligands promotes tumor specificity and transfection efficiency and reduces the toxicity of PEI polymers [33-34]. In this study, the FSH β 33-53 peptide and retro-inverso FSH β 33-53 peptide were used to target FSHR-positive ovarian cancer. Compared with PP-shHK2, the HK2 shRNA-loaded nanoparticles with FSH peptide modification, FPP-shHK2 and DFPP-shHK2, greatly reduced HK2 expression. DFPP-shHK2 showed the greatest knockdown effect. Accordingly, cellular HK2 enzyme activity was significantly decreased in both A2780 and A2780CP cells after FPP-shHK2 and DFPP-shHK2 treatment (Figure 3E), which could be due to the decreased levels of HK2 expression. These data indicated that HK2 shRNA-loaded nanoparticles with FSH peptides or retro-inverso FSH peptide modification effectively attenuated HK2 expression even without the help of transfection reagents, which further enhanced the feasibility of the in vivo administration of RNAi drugs and nanoparticles.
HK2 knockdown by retro-inverso FSH β 33-53 peptide-conjugated nanoparticles facilitated a shift in the glucose metabolism pattern of ovarian cancer cells
In contrast to normal cells, cancer cells prefer aerobic glycolysis even at normal oxygen levels. One possible explanation is that cancer cells have an intense metabolic demand without efficient ATP production . Cancer cells generate glycolytic intermediates that are available for the synthesis of amino acids, nucleotides, and fatty acids to meet the demands of fast proliferation. Another possible explanation is that the switch to aerobic glycolysis is linked to therapeutic resistance . Highly invasive ovarian cancer cells present a more intensive glycolytic phenotype than less invasive ovarian cancer cells [37-38]. HK2 is crucial for aerobic glycolysis and is overexpressed in ovarian cancer [15-17].
To determine whether knockdown of HK2 reverses the glucose metabolism pattern, OCR and ECAR were evaluated in cisplatin-sensitive and cisplatin-resistant ovarian cancer cells. A2780 and A2780CP cells treated with DFPP-shHK2 showed significantly enhanced FCCP-stimulated oxygen consumption, suggesting that HK2 silencing upregulated mitochondrial respiration (Figure 4A-4C). Oligomycin was used to inhibit ATP production by the electron transport chain and to evaluate the glycolytic reserves of cells. A2780 and A2780CP cells treated with DFPP-shHK2 exhibited a significant decrease in oligomycin-stimulated ECAR, indicating the reversal of the Warburg effect in cells (Figure 4D-4F). Glucose consumption, lactate production and cellular ATP levels were further measured to determine the glycolytic shift (Figure 4G-4I). After treatment with DFPP-shHK2 for 24 h, lactate levels in the medium and cellular ATP levels significantly decreased, while glucose levels in the medium were increased in both A2780 and A2780CP cells. The effects of cisplatin alone or in combination with DFPP-shHK2 on glucose metabolism were also investigated. As shown in Figure 4, both A2780 and A2780CP cells exhibited a significant shift towards OXPHOS-dependent glucose metabolism after treatment with a combination of cisplatin and DFPP-shHK2 in comparison with that in cells treated with cisplatin alone. Taken together, these data suggested that knockdown of HK2 facilitated a shift towards normal oxidative glucose metabolism even in cisplatin-resistant ovarian cancer cells. The underlying mechanism could be associated with changes in mitochondrial function and biogenesis. After depletion of HK2, there was an increase in the activity of the electron transport chain complexes I/II/III/IV/V, peroxisome-proliferator-activated receptor-γ coactivator α (PGC1α) and mitochondrial transcription factor A (mTFA) .
Antitumor effects of HK2 shRNA-loaded nanoparticles on chemosensitive ovarian cancer in vitro and in vivo
To determine the effects of HK2 shRNA-loaded nanoparticles on the viability, migration, invasion and apoptosis of chemosensitive ovarian cancer cells, cisplatin-sensitive A2780 cells were used. As shown in Figure 5, FPP-shHK2 and DFPP-shHK2 significantly inhibited the proliferation, invasion and migration and induced apoptosis of A2780 cells, especially DFPP-shHK2, which had the greatest effect on the knockdown of HK2 expression. In contrast, overexpression of HK2 could promote cell proliferation, migration, and invasion and the stemness of ovarian cancer cells via the FAK/ERK1/2 signaling pathway .
The antitumor effects of HK2 shRNA-loaded nanoparticles were further investigated using nude mice bearing cisplatin-sensitive A2780 tumor xenografts. PP-shHK2, FPP-shHK2 or DFPP-shHK2 nanoparticles were intravenously administered. As shown in Figure 6, FPP-shHK2 and DFPP-shHK2 significantly suppressed tumor growth compared with the control treatment. DFPP-shHK2 treatment exhibited the highest antitumor effect, with an inhibition rate of 70.1%. There was no difference between the PP-shHK2 and control groups, which might be due to the weak knockdown effect of PP-shHK2 on HK2 expression.
These data were consistent with the data from the in vitro experiments and further indicated that FPP-shHK2 and DFPP-shHK2 effectively inhibited tumor growth in cisplatin-sensitive ovarian cancer. The possible reasons for this could involve the increased accumulation of HK2 shRNA in the tumor sites, which was mediated by FSH peptides, and the subsequent suppression of the glycolysis of cancer cells. Although the related downstream molecular mechanisms remain unclear, the effects of metabolites on tumors have been partially elucidated. Lactate has been identified as a primary fuel for the TCA cycle, and lung cancer cells prefer to use lactate as an energy source for tumor metabolism [40-41]. It has been indicated that the purpose of the Warburg effect is to generate lactate for carcinogenesis . Lactate is involved in the development of cancer, including angiogenesis, immune escape, and cell migration . The enhancement by lactate of the migratory potential of cancer cells occurs in a dose-dependent manner . Here, knockdown of HK2 greatly reduced extracellular lactate levels, which could contribute to the antitumor effects in vitro and in vivo.
HK2 knockdown by retro-inverso FSH β 33-53 peptide-conjugated nanoparticles improved cisplatin sensitivity in vitro
An increasing number of studies have revealed that the enhancement of aerobic glycolysis contributes to chemoresistance [6, 45-46]. An increase in glycolysis has been observed in chemoresistant cancer cells, and increased intracellular ATP levels can induce a drug-resistant phenotype . Signaling pathways, such as the HIF-1α signaling pathway, that are activated by dysregulated metabolism may also contribute to chemoresistance. HK2 overexpression is related to the chemoresistance of glioblastoma, acute myeloid leukemia, colorectal cancer and ovarian cancer [15, 48-50]. To further investigate the effect of HK2 knockdown on ovarian cancer chemoresistance, cisplatin-resistant A2780CP cells were treated with HK2 shRNA-loaded nanoparticles. As shown in Figure 7A, A2780CP cells were not effectively killed by DFPP-shHK2 alone, unlike cisplatin-sensitive A2780 cells. Next, the effects of DFPP-shHK2 combined with cisplatin were detected (Figure 7B and 7C). Both A2780 and A2780CP cells showed an increased sensitivity to cisplatin when treated with the DFPP-shHK2 combination. Proliferation was inhibited, and apoptosis was enhanced even in cisplatin-resistant A2780CP cells. The 24-h half maximal inhibitory concentrations (IC50) of cisplatin for the A2780CP and DFPP-shHK2-treated A2780CP cells were 16.5 and 10 μg/mL, respectively. The inhibitory effects could be linked to the shift of glycolysis to OXPHOS-dependent glucose metabolism after depletion of HK2. The reduced levels of glycolytic intermediates such as lactate and reduced ATP levels could not sustain the demands of ovarian cancer cells to resist the killing power of cisplatin.
Furthermore, we investigated the changes in cisplatin resistance-related proteins after HK2 shRNA-loaded nanoparticle treatment. The expression level of ATP7b, which is an ATP-related transport protein that promotes cisplatin resistance by pumping cisplatin out of cancer cells, was significantly decreased by DFPP-shHK2-induced HK2 knockdown (Figure 7D). However, other cisplatin transport proteins, such as ATP7a, PGP and MRP2, showed no changes. The expression levels of the mitochondrial apoptosis-associated proteins Bax and cytochrome c were increased after DFPP-shHK2 treatment. These data suggested that HK2 knockdown induced by DFPP-shHK2 effectively improved the cisplatin sensitivity of ovarian cancer cells by regulating cisplatin transport proteins and increasing apoptosis through the mitochondrial pathway.
Antitumor effects of retro-inverso FSH β 33-53 peptide-conjugated nanoparticles on cisplatin-resistant tumor xenografts
To evaluate the in vitro effects of HK2 depletion, nude mice bearing cisplatin-resistant A2780CP tumor xenografts were intravenously administered DFPP-shHK2 and cisplatin alone or in combination. As shown in Figure 8A-8C, DFPP-shHK2 significantly suppressed tumor growth, while the inhibitory effect in cisplatin-resistant A2780CP tumors was obviously lower than that in cisplatin-sensitive A2780 tumors. When the mice were treated with DFPP-shHK2 in combination with cisplatin, both the tumor volumes and tumor weights were dramatically reduced, with an inhibition rate of 78.1%. Interestingly, the expression level of the cisplatin resistance-associated protein ATP7b in A2780CP tumor xenografts showed no changes in the cisplatin alone group but showed a decrease in the DFPP-shHK2 combination group. Bax and cytochrome c expression were increased in the cisplatin alone and combination groups compared with those in the control group (Figure 8D). Together, the results show that targeting HK2 could be an effective therapeutic strategy for cisplatin-resistant ovarian cancer, although the molecular mechanisms of drug resistance that are affected by aerobic glycolysis need further elucidation.
In vivo toxicity of HK2 shRNA-loaded nanoparticles
The systemic toxicity of small molecular inhibitors of HK2 has limited their clinical use. High levels of HK2 inhibitors that sufficiently suppress tumor growth might simultaneously lead to toxicity in normal cells. Additionally, the polymers used in the delivery system easily accumulate in lung and liver tissues, which may cause toxic effects. It has been proven that the modification of targeting peptides on polymers helps to achieve tumor-specific delivery and overcome natural accumulation . Since FSHR is overexpressed in different histologic types of ovarian cancer but not in non-ovarian healthy tissues , we designed FSHR-mediated nanocarriers for HK2 shRNA to increase tumor uptake and decrease toxicity by using the binding peptides of FSHR. As shown in Figure 9, the in vivo toxicity of HK2 shRNA-loaded nanoparticles was evaluated in nude mice bearing A2780 tumor xenografts. The body weights were not significantly reduced in mice treated with HK2 shRNA-loaded nanoparticles during the study period. The blood levels of ALT, Cr and BUN exhibited no significant differences among all groups. HE staining of the liver, kidney, spleen and lung did not show any obvious damage.
Although the relationship between FSH, FSHR and ovarian cancer remains controversial, the safety and effectiveness of FSHR- and FSH-derived peptides has been examined in studies of the targeted immunotherapy of ovarian cancer. Targeting chimeric receptors for T cell activation using full-length FSH produces strong therapeutic effects against ovarian cancer in vivo without any measurable toxicity, even in the presence of tumor-free ovaries . Similar results have been observed in a study of FSHR-redirected T-cells using FSH fragments, including the FSH β 33-53 peptide .