P-MAPA immunotherapy potentiates the effect of cisplatin on serous ovarian carcinoma through targeting TLR4 signaling

Forty adult female Fischer 344 rats (60-day-old, weighing ±200 g), were obtained from the Multidisciplinary Center for Biological Investigation (CEMIB) at University of Campinas (UNICAMP). Animals were individually housed in polypropylene cages with laboratory-grade pine shavings as bedding and kept in a climatized room under controlled temperature at 22 ± 1 °C with a 12/12 h light/dark cycle, lights switched on at 6:00 a.m. During the procedures, standard rodent food (3074 SIF, Purina Ltda., Campinas, SP, Brazil) and filtered tap water were provided ad libitum. The animals were divided into four groups (n = 10/group): OC: composed of rats that were induced with 7,12-dimethylbenz(a)anthracene (DMBA) and received vehicle only; OC + P-MAPA: composed of rats that were induced with DMBA and received P-MAPA as therapy; OC + CIS: composed of rats that were induced with DMBA and received cisplatin as therapy; OC + P-MAPA+CIS: composed of rats that were induced with DMBA and received simultaneous doses of P-MAPA and cisplatin as treatment. This chemically-induced rat model of OC is appropriate for studying the structural and functional components of the tumor microenvironment, including the immune system [21, 22].

After OC was developed (200-day-old), the animals designated to receive P-MAPA were administered i.p. doses of 5 mg/kg b.w. dissolved in 0.20 mL of 0.9% (v/v) saline solution (vehicle) at a final concentration of 5 mg/mL [13]. The animals treated with cisplatin received two doses of 5 mg/kg, via i.p., twice a week (on Monday and Thursday), dissolved in 0.20 mL of the vehicle. The combination therapy was administered twice a week via i.p. (5 mg/kg P-MAPA and 5 mg/kg CIS). Control animals received injections of the vehicle in the same procedures of treated animals. The injections were administered in the morning (between 09:00 and 10:00 h, Fig. 1a and b), twice a week for 8 consecutive weeks. After the experimental period, all animals were euthanized by decapitation (during early morning at 8:00 a.m.) for sample collection.

To minimize any kind of discomfort and pain during experimentation, the rats were anesthetized with ketamine and xylazine so that they were rendered unconscious. Euthanasia was performed in accordance with the Canadian Council on Animal Experimentation. This experimental protocol was previously approved by the Ethical Committee of the Institute of Bioscience/UNESP, Campus of Botucatu, SP, Brazil (Permit Number: 950).

After defining the experimental groups, all the animals (n = 40) were anesthetized using 10% ketamine (60 mg/kg, i.p., Vibra® Roseira, SP, Brazil) and 2% xylazine (5 mg/kg, i.p., Vibra® Roseira, SP, Brazil) during the estrous stage. The left flank of the animals was incised through the skin and abdominal wall to access the ovaries. The left ovaries were injected direct under the ovarian bursa with a dose of 100 μg of DMBA (Sigma Chemical Co, St. Louis, MO) dissolved in 10 μL of sesame oil (vehicle) [23‐25]. Right ovaries were injected with the vehicle only (sham-surgery). After the ovaries were moved back to the abdominal cavity, muscle layers and skin were tightly closed using a 3–0 silk suture, and prophylactic treatment with antibiotic benzylpenicillin potassium was given to the animals. Over the next 120 days of development, tumors were monitored by ultrasound to predict the volume and total size.

After euthanasia, the animals were necropsied and studied for general systemic abnormalities. During necropsy, the internal reproductive organs, and those related to body metabolism were carefully dissected. Animals displaying evident OC had a portion of a large tumor removed, rapidly frozen, and stored at − 80°C. For the histopathological analyses, serous OC was fixed in 10% buffered formalin followed by dehydration, clearing, and paraffin embedding. The tissue was serially sectioned at 5-μm thickness, and every 10th tissue section was stained with hematoxylin and eosin. Histopathological data were analyzed and reported by a pathologist.

Tissue sections of papillary OCs were deparaffinized and microwaved at 700 W using 0.01-M sodium citrate buffer (pH 6.0). After antigen retrieval, the peroxidase activity was blocked, and OC tissues were incubated with 3% BSA for 1 h. Then, OC sections were incubated overnight with primary antibodies (Abcam, Cambridge, UK): rabbit polyclonal anti-TLR2 (1:100) and mouse monoclonal anti-TLR4 (1:100). After the reactions, tissues were washed 3× and incubated with polymer Anti-Mouse IgG or Anti-Rabbit (DAKO ® CYT) at room temperature (RT) for 1 h. Finally, the slides were reacted with diaminobenzidine (DAB; Sigma, St. Louis, MO), and OC sections counterstained with hematoxylin for 1 min. To eliminate false positive reactions, control slides were obtained without the primary antibody. Acquisition of all images was performed under a microscope (Zeiss Axiophot II, Carl Zeiss, Oberkochen, Germany), considering the intensity of immunoreactivity as absent (0), low (+), moderate (++), or high (+++).

The OC tissues were isolated, washed in PBS and fixed in 4% paraformaldehyde for 10 min. For tissue permeabilization, proteinase K was used and nonspecific bindings were blocked with 1% BSA for 1 h. Then OCs were incubated overnight with rabbit polyclonal NF-kB p65 antibody diluted 1/100 (ab7970, Abcam, Cambridge, MA), followed by another incubation with secondary rabbit polyclonal anti-IgG conjugated to FITC (1:200 dilution, Santa Cruz Biotechnology, Inc., CA) for 1 h. For nuclei staining, 4,6-diamidino-2-phenylindole (DAPI; Sigma, St Louis, MO) was added (5 min). Positive staining in OC cells was analyzed using a fluorescence microscope (Zeiss Axiophot II, Oberkochen, Germany) at 40× magnification (excitation filter 590 nm, emission filter 650 nm) and for DAPI staining (excitation filter 365 nm, emission filter 485 nm). Relative fluorescence in merged images was calculated using Image J software.

For the quantitative determination of proinflammatory cytokines, levels of IFN-γ (cat. R1F00), IL-6 (cat. R6000B), and TNF-α (cat. RTA00) were achieved using similar amounts of proteins extracted from the OC samples. The commercial pre-coated rat Quantikine® ELISA kits were provided by R&D Systems (Minneapolis, MN).

All data were performed using analysis of variance (ANOVA) and presented as the mean ± standard deviation (SD). Significant results were subjected to post-hoc analysis with Tukey tests, and statistical significance level was set to P < 0.05 for all analyses. Data were analyzed using GraphPad Prism 5.0 scientific graphing software (GraphPad Software, San Diego, CA).

The OC incidence rate was increased from 4 to 8 months after DMBA injection reaching almost 80% of the total developed neoplasms (Fig. 2a). We initially inoculated 13 animals by group, and at least 10 animals exhibited aggressive carcinoma; no death occurred during tumor development. To confirm the effect of P-MAPA and CIS on this rat model of OC, tumor volume and masses were investigated. Animals treated with CIS alone and the combination of P-MAPA with CIS showed an effective reduction in OC volume after 30 days of treatment (24.4% and 20.3%, respectively, compared with the OC group; Fig. 2b). Most notably, P-MAPA and CIS alone or in combination significantly reduced the tumor volume after 60 days of treatment (16.3%, 41% and 32.2%, respectively, compared with the OC group; Fig. 2b). The combination of P-MAPA with CIS was still more efficient in promoting a reduction in OC volume compared to P-MAPA alone (20% reduced after 60-day treatment). At the end of treatments, CIS- and P-MAPA+CIS-treated groups exhibited a significant decrease in OC masses (30% and 26%, respectively, lower than the OC group; Fig. 2c).

After 180 days of OC induction, the tumors were carefully assessed. Most of the left ovaries from the OC group showed solid mass with scattered areas of necrosis (Fig. 3a, e, i). Furthermore, these animals showed many peritoneal implants followed by abdominal adhesions composed of fibrous bands (Fig. 3j); when untreated, 100% of the control females died at 280 days after DMBA induction (Fig. 3l). In addition, the majority of P-MAPA-treated animals exhibited OC with serous fluid-filled cysts (Fig. 3b, f). Supported by reduced tumor volume and masses, the left ovaries of the animals in the CIS and P-MAPA+CIS groups had great morphological differences evidenced by mobile and soft pale tissue with few adhesions (Fig. 3c, d, g, h). Conversely, the right ovaries (sham-operated animals) appeared largely intact with no macroscopic lesion. Figure 3k shows a non-expansive and controlled OC in an animal that received the combination of P-MAPA and CIS as therapy. To evaluate the prognostic value of the different treatments, we first investigated non-treated animals with OC over several weeks after tumor induction (Fig. 3l). As expected, animals from the OC group revealed an impressive shorter survival rate (280 days after OC acquisition). Surprisingly, treatment with P-MAPA showed the longer survival rate by about 65% (~ 462 days), whereas CIS and P-MAPA+CIS treatments only had an increase in animal survival by about 25% (~ 350 days) and 35% (~ 380 days), respectively, from 32 to 50 weeks after tumor induction (Fig. 3l). After the period of OC development, the majority of tumors were malignant carcinomas classified as high-grade serous papillary (Fig. 4). In general, OCs showed a mixture of papillary and cystic pattern with extensive projections, cellular atypia, and stromal invasion. Furthermore, the presence of scattered mitosis and tumor-infiltrating lymphocytes was obvious, thus, reflecting a potential compartment to be challenged with immunostimulants; although the tumor size was significantly reduced, no histopathological changes were observed after P-MAPA and CIS treatments (Fig. 4a-e).

In response to P-MAPA immunotherapy, the expressions and immunostaining of TLR2 were enhanced (1.46-fold increase vs. OC) in the cells of serous papillary OCs (Table 1, Fig. 5a, b, i, j). CIS therapy and the combination of P-MAPA with CIS showed no significant effect on the TLR2 levels (p > 0.05; Table 1, Fig. 5c, d, i, j). Notably, P-MAPA therapy positively regulated TLR4 in these cells (1.50-fold increase vs. OC; Table 1, Fig. 5e, f, i, k), and CIS treatment also upregulated TLR4 but not in the same intensity as P-MAPA did (1.37-fold increase vs. OC; Table 1, Fig. 5e, g, i, k). Unexpectedly, the combination of P-MAPA with CIS promoted a higher overexpression of TLR4 compared to the therapies alone (1.02- and 1.36-fold increase vs. P-MAPA and CIS groups, respectively; Table 1, Fig. 5h, i, k).

To investigate the effects of P-MAPA and CIS upon MyD88- or TRIF-dependent signaling pathways, these adaptor molecules were measured in the OC tissues. Interestingly, P-MAPA and CIS alone or even the combination of P-MAPA and CIS promoted an increase in MyD88 levels (1.56-, 1.48-, and 1.58-fold increases, respectively vs. OC; Fig. 6a, b), thereby providing evidence for a MyD88-dependent TLR4-mediated signaling. In addition, the combination of P-MAPA with CIS resulted in the elevation of TRIF levels (1.45-fold increase vs. OC; Fig. 6a, b), being able to further activate downstream molecules via the TRIF-dependent signaling pathway.

The characterization of NF-kB signaling pathway was also evaluated in both cytoplasmic and nuclear extracts of OC following the expressions of IKK-α, p-IkBα, and NF-kB p65 subunit. While IKK-α was only upregulated by CIS treatment (1.57-fold increase vs. OC; Fig. 6a, b), the phosphorylation of IkBα was significantly increased by P-MAPA, CIS, and the combination P-MAPA+CIS (1.45-, 1.55-, and 1.60-fold higher, respectively vs. OC; Fig. 6a, b), thus proving that IKK complex is active in mediating IkB phosphorylation, and consequently, its degradation. Surprisingly, P-MAPA and the combination of P-MAPA with CIS induced the upregulation of cytosolic NF-kB p65 (1.61- and 1.69-fold increases, respectively vs. OC; Fig. 6a, b), and more effectively, P-MAPA, CIS, and P-MAPA+CIS treatments dramatically enhanced nuclear NF-kB p65 (1.49-, 1.50-, and 1.58-fold increases, respectively vs. OC; Fig. 6b, c); particularly, a considerable amount of the NF-kB p65 complex was translocated into the nucleus. Immunofluorescence assay further indicated the localization and expression level of cytoplasmic and nuclear NF-kB p65 in OC cells. P-MAPA immunotherapy resulted in high p65 expression (Fig. 6d I and II; the level of fluorescence increased from 43% ± 0.67 (OC) to 86% ± 11.3 (P-MAPA) as well as CIS therapy (Fig. 6d I and III; the level of fluorescence increased from 43% ± 0.67 (OC) to 102% ± 14.1 (CIS). Lastly, the combination of P-MAPA and CIS promoted the highest p65 immunofluorescence intensity (134% ± 17.1; Fig. 6d IV).

To verify the effects of therapies on pro-inflammatory cytokines, the representative levels of IFN-γ, TNF-α, and IL-6 were measured in serous OC samples of animals and were reported to be 33 ± 6.1 pg/mL, 5.68 ± 1.88 pg/mL and 23.3 ± 2.38 pg/mL, respectively (Fig. 7a–c). Treatment with P-MAPA and the combination P-MAPA + CIS promoted an increase in the concentration of IFN-γ in OC tissues (1.56- and 1.67-fold increases, respectively vs. the OC group; Fig. 7a). The levels of TNF-α and IL-6 were unchanged in serous OC after P-MAPA and CIS treatments (Fig. 7b and c).