Curcumin potentiates the ErbB receptors inhibitor Afatinib for enhanced antitumor activity in malignant mesothelioma

Abstract Several attempts have been made to develop targeted therapies for malignant mesothelioma (MM), an aggressive tumour with a poor prognosis. In this study we evaluated whether Curcumin (CUR) potentiated the antitumor activity of the ErbB receptors inhibitor Afatinib (AFA) on MM, employing cell lines cultured in vitro and mice bearing intraperitoneally transplanted, syngeneic MM cells. The rationale behind this hypothesis was that CUR could counteract mechanisms of acquired resistance to AFA. We analysed CUR and AFA effects on MM cell growth, cell cycle, autophagy, and on the modulation of tumour-supporting signalling pathways. This study demonstrated that, as compared to the individual compounds, the combination of AFA + CUR had a stronger effect on MM progression which can be ascribed either to increased tumour cell growth inhibition or to an enhanced pro-apoptotic effect. These results warrant future studies aimed at further exploring the therapeutic potential of AFA + CUR-based combination regimens for MM treatment.


Introduction
Malignant mesothelioma (MM) is a rare and aggressive tumour arising from the mesothelial cell lining of serous membranes and mainly associated with asbestos exposure (Rossini et al. 2018;Asciak et al. 2021).MM can also arise following exposure to asbestiform mineral fibres and ionising radiation as well as after chronic serous inflammation (Visci et al. 2022).
Indeed, the prognosis of MM patients is presently poor, since the tumour is seldom amenable to surgical eradication and resistant to conventional multimodal therapies.The current standard of care is based on the combination of cisplatin and pemetrexed, and the 5-year overall survival (OS) is only of 5% for all tumour stages (Asciak et al. 2021;Davis et al. 2022).
Although several attempts have been made to develop targeted therapies for MM (Palumbo et al. 2008;Davis et al. 2022), studies conducted in the past years have shown the overall limited clinical activity of agents targeting different signal transduction pathways on this tumour (Palumbo et al. 2008;Jakobsen and Sørensen 2011).In this context, the evidence that the Epidermal Growth Factor Receptor (EGFR) is frequently overexpressed in MM, has provided the rationale for testing EGFR-tyrosine kinase inhibitors (TKIs) in MM patients.Indeed, compared to normal pleura, EGFR overexpression has been reported in a percentage of MMs ranging from 44% to 97% (Destro et al. 2006;Chia et al. 2019).On the other hand, EGFR is rarely mutated in this tumour (Destro et al. 2006;Velcheti et al. 2009;Chia et al. 2019;Chia et al. 2021), which is one of the possible explanations for the poor clinical impact of first-generation, reversible EGFR TKIs such as Gefitinib and Erlotinib on MM, since these inhibitors are known to require sensitising EGFR mutations to exhibit significant potency (Marquez-Medina and Popat 2015; Duggirala et al. 2022).
Afatinib (AFA) is a second-generation, irreversible TKI first approved for the treatment of advanced non-small-cell lung cancer (NSCLC) harbouring EGFR mutations (Marquez-Medina and Popat 2015).AFA properties may renew the interest in the therapeutic potential of EGFR inhibitors for treating MM.Indeed, as compared to first-generation EGFR TKIs, AFA displays greater activity against wild-type as well as mutated EGFR, both in vitro and in clinical studies (Marquez-Medina and Popat 2015;Vavalà 2017;Ezeife et al. 2018).Moreover, beside EGFR, AFA is able to inhibit also ErbB2 and ErbB4 receptors, which are also expressed in MM (Marquez-Medina and Popat 2015;Vavalà 2017).In particular, ErbB2 expression has been reported in 6-97% of MM cases, and ErbB4 expression in about 50% of MM cases (Thirkettle et al. 2000;Klampatsa et al. 2017).Accordingly, AFA may have therapeutic activity in EGFR wild-type MM patients due to its ability to inhibit signalling from all ErbB homo-and heterodimers (Marquez-Medina and Popat 2015;Vavalà 2017).By the way, a partial response has been obtained with AFA in a patient with EGFR-mutated MM, unresponsive to cytotoxic therapy (Agatsuma et al. 2017).
Beside the EGFR mutational status, mechanisms of MM cell resistance to EGFR TKIs appear to involve the concurrent activation of alternative "bypass" signalling pathways (Huang and Fu 2015;Chia et al. 2019).In this respect, there is a growing interest in natural polyphenolic compounds, such as Curcumin (CUR), able to target multiple signalling pathways and overcome tumour cell resistance to chemotherapy and targeted therapy agents (Fantini et al. 2015;Chmielewska-Kassassir and Wozniak 2021).CUR has been shown to be able to suppress tumour cells proliferation, to induce apoptosis, to inhibit epithelial-mesenchymal transition (EMT), neoangiogenesis, invasion and metastasis in different types of cancer (Benvenuto et al. 2020;Arena et al. 2021;Karaboga Arslan et al. 2022).
We and others have previously demonstrated the preclinical efficacy of CUR on MM, both in vitro and in vivo (Masuelli et al. 2017;Sayeed et al. 2017;Baldi et al. 2020).Interestingly, it has been reported that CUR can enhance the efficacy of EGFR TKIs in NSCLC with wild-type EGFR (Chen et al. 2019).
In the present study we investigated whether CUR could potentiate AFA antitumor activity on human and mouse MM cells cultured in vitro.Further, we addressed the effect of the AFA-CUR combination in a mouse model, in which the transplantation of syngeneic MM cells induces ascites in the peritoneal space.

Cell lines and treatments
Human pleural MM cell lines (H-Meso-1, MM-F1, MM-B1) were kindly provided by Prof. Antonio Procopio (Università Politecnica delle Marche, Ancona, Italy) (Reale et al. 1987;Catalano et al. 2001).The murine MM cell line #40a, kindly provided by Dr. Agnes Kane (Department of Pathology and Laboratory Medicine, Brown University, Providence, RI, USA), was derived from the 40-cell line after two passages in the peritoneal cavity of syngeneic C57BL/6 mice following pristane administration one week before cells transplant (Goodglick et al. 1997).These passages allowed the selection of cells, which reproducibly form ascites when intraperitoneally injected in mice (Benvenuto et al. 2023).#40a and H-Meso-1 cells have an epithelial morphology, while MM-F1 and MM-B1 cells have sarcomatous and biphasic features, respectively.Cells were grown as previously described (Benvenuto et al. 2023).AFA and CUR were dissolved in DMSO.For treatments, cells were incubated for the indicated times in the presence of CUR and AFA, alone or in combination, at different concentrations (CUR: range 2.5-20 µM; AFA: range 2.5-10 µM; AFA + CUR: range 2.5 + 2.5 to 10 + 10 µM) or vehicle control (DMSO ≤ 0.1%).

SRB assay
Cells were seeded at 5 × 10 3 /well in 96-well plates.After 24 h, the medium was changed and cells were incubated for 24, 48 and 72 h with CUR and/or AFA, or with DMSO.The SRB assay was then performed as previously described (Benvenuto et al. 2018).The percentage survival of the cultures treated with CUR and/or AFA was calculated by normalisation of their optical density (O.D.) values to those of the control cultures treated with DMSO (Benvenuto et al. 2018).
The combined effects of CUR and AFA were analysed according to the method of Kern.Cell survival data were processed to obtain an index (R) defined as follows: R = S exp /S obs , where S exp , the expected survival, is the product of the percentage survival observed with AFA alone and the percentage survival observed with CUR alone, and S obs , the observed survival, is the actual percentage survival observed with the AFA + CUR combination.An R index lower than 1 indicates that the combination exerts a less than additive effect; an R index of 1 indicates that the effect is additive, and any value of R greater than unity indicates a synergistic interaction (Bei et al. 2022).

FACS analysis of DNA content
Asynchronized log-phase growing cells (60% confluent, approximately 2.5 × 10 5 cells/well in 6-well plates), either pre-treated or not with ZVAD-FMK (final concentration 40 μM) for 2 h, were treated with CUR and/or AFA (final concentration 10 μM) or DMSO.After 48 h, adherent and suspended cells were harvested, stained with propidium iodide and analysed by flow cytometry as previously described (Benvenuto et al. 2023), using a FACS-Calibur cytometer running CellQuest Pro 5.2 software (BD Biosciences, San Jose, CA, USA).

Western blotting
About 1.5x10 6 cells were seeded in 100 mm tissue culture dishes 24 h prior to the addition of CUR and/or AFA (final concentration 10 μM) or DMSO.After 24 h of treatment, cells were lysed, and lysates (80 μg) resolved in 10-15% SDS-PAGE and transferred to nitrocellulose membranes as previously described (Santarelli et al. 2008;Alesiani et al. 2009;Masuelli et al. 2012).The enhanced chemiluminescence system ECL LiteAblot (Euroclone, Milan, Italy) was used to detect protein expression.The O.D. of each band was analysed with the ImageJ 1.53e software (NIH, Bethesda, MD, USA) after blot scanning and expressed as bar graphs in the figures.
Treatments were started simultaneously with the inoculation of MM cells.The experiments were conducted in accordance with the ethical standards and according to the Declaration of Helsinki and the ARRIVE guidelines.A veterinary surgeon was present during the experiments.The animal care, both before and after the experiments, was performed only by trained personnel.Mice were bred under pathogen-free conditions in the animal facilities of the University of Rome "Tor Vergata" and handled in compliance with European Union (EU Directive 2010/63/EU) and institutional standards for animal research.The investigation was conducted with the formal approval of the local animal care committees (institutional and national), and animal experiments were registered as legislation requires (Authorization from Ministry of Health no.179_2020-PR).

Analysis of antitumor activity in vivo
Growth of #40a cells in C57BL/6 mice induces ascites.Accordingly, the abdominal circumference of mice was monitored before the inoculation of cells and then every week.Tumour-bearing mice were euthanized at the first signs of distress or when their abdominal circumference exceeded 12 cm (Benvenuto et al. 2018).

Statistical analysis
Data distribution of cell proliferation, cell survival and FACS analysis was preliminarily verified using the Kolmogorov-Smirnov test, and the data sets were analysed by one-way analysis of variance (ANOVA) followed by the Newman-Keuls test (GraphPad Prism 5.0 software).Differences in the intensity of immunoreactive bands were evaluated by one-way ANOVA followed by the Newman-Keuls test.Survival curves and tumour volumes were analysed using the Kaplan-Meier method and compared with a log-rank test (Mantel-Cox).The threshold for statistical significance was set at p < .05.

CUR potentiates AFA in inhibiting the growth of MM cell lines
The effect of CUR and AFA on the growth of human (H-Meso-1, MM-F1, MM-B1) and murine (#40a) MM cell lines was investigated by SRB assay after 24, 48 and 72 h of treatment with increasing concentrations of CUR (range 2.5-20 µM) or AFA (2.5-10 µM) alone, or after treatment with equimolar concentrations of AFA and CUR (2.5-10 µM).Cells treated with the DMSO vehicle were used as controls.
CUR and AFA, alone and in combination, significantly inhibited MM cell growth in a dose-and time-dependent manner (Figure 1 and Supplementary Figure 1).Indeed, a significant cell growth inhibition was observed in all cell lines already after 24 h of treatment with the highest concentrations of the single compounds.Further, both CUR and AFA inhibited the growth of H-Meso-1, MM-B1 and #40a cell lines even at the lowest doses after a longer incubation when used alone.Among the two compounds, AFA was more potent at all concentrations and times tested (Figure 1).In particular, after 72 h of treatment, the CUR half-maximal inhibitory concentrations (IC 50 s) were comprised between 15 and 40 µM, depending on the cell line, whereas AFA IC 50 s were in the 1-12 µM concentration range (Supplementary Table 2).Still, when both compounds were used at 10 µM, the growth inhibition obtained with AFA was potentiated by CUR at all the time points investigated in MM-B1 and MM-F1 cells, and at early time points in H-Meso-1 and #40a cells (Figure 1 and Supplementary Figure 1).
The mode of interaction between CUR and AFA 10 µM was also evaluated using the Kern index (R), which was calculated from cell growth inhibition data as reported in the Materials and methods section.R values <1 indicate that the combination exerts a less than additive effect, R = 1 indicates that the effect is additive, R values >1 indicate a synergistic interaction (Bei et al. 2022).As shown in Figure 2, the AFA + CUR combination had synergistic effects in MM-F1 cells over the whole time period investigated and in MM-B1 cells after 48-72 h of treatment.Moreover, in these two cell lines the synergism on cell growth inhibition showed a growing trend with longer incubation times.On the other hand, in the two epithelioid MM cell lines, H-Meso-1 and #40a, the growth inhibitory effect of the two drugs appeared, respectively, slightly synergistic and synergistic at early time points and less than additive at later time points.

CUR potentiates apoptosis induction by AFA in epithelioid MM cell lines
The SRB assay is based on the measurement of cellular protein content but does not distinguish viable from non-viable cells (Vichai and Kirtikara 2006).Therefore, in order to discriminate between viable and apoptotic cells, FACS analysis of DNA content was performed after 48 h of treatment with the compounds, alone or in combination, at 10 µM.The universal caspase inhibitor Z-VAD-FMK was also used to confirm cell death induction by apoptosis.
As shown in Figure 3a, the percentage of cells in the subG1 phase of the cell cycle (i.e.apoptotic cells with a hypodiploid DNA content) was increased by AFA in each cell line, whereas CUR treatment significantly increased the subG1 fraction in H-Meso-1 only.As for the pro-apoptotic effect of the drug combination, AFA + CUR was significantly more effective than the single treatments in increasing the subG1 fraction in the epithelioid H-Meso-1 and #40a, but not in the other two cell lines (Figure 3a).This increased effectiveness was particularly evident in the H-Meso-1 cell line, where the percentage of cells in subG1 was nearly doubled by the combined treatment compared to AFA alone.
These findings were corroborated by Western blot studies: as compared to the treatment with AFA alone, the combined treatment further increased the ratio between the pro-apoptotic Bax and the anti-apoptotic Bcl-2 proteins as well as the levels of the of apoptosis-related DNA fragmentation marker γH2AX in both epithelioid MM cell lines (Figure 3b,c).Moreover, in H-Meso-1 cells the ratio between the cleaved and uncleaved PARP-1 protein was also slightly increased by AFA + CUR as compared to AFA alone (Figure 3b,c) (Benvenuto et al. 2023).
Interestingly, as compared to the effect of AFA alone (i.e. the more potent among the two compounds), the subG1 increase induced by AFA + CUR was accompanied by a significant decrease of the percentage of cells in all the other phases of the cell cycle in H-Meso-1 and of the G0/G1 fraction in #40a cells (Supplementary Table 3).In contrast, as compared to AFA, the combined treatment caused minor or not significant changes in cell cycle distribution in the non-epithelioid MM-F1 and MM-B1 cell lines (Supplementary Table 3).These findings, coupled with those of SRB assays, indicate that CUR potentiates the effect of AFA through different mechanisms: by reducing the proliferation rate in the absence of an evident arrest of the cell cycle, as observed in the non-epithelioid cell lines, or by increasing the induction of apoptosis, as observed in the cell lines with an epithelioid phenotype.

Effects of CUR and AFA, alone or in combination, on the expression of markers of autophagy
Autophagy can serve as a rescue pathway able to increase cancer cell survival in stress conditions, including those caused by antineoplastic agents (Amaravadi et al. 2019).Therefore, the expression levels of autophagy markers including Beclin-1, SQSTM-1/p62 and LC3-I/II, were investigated by Western blotting in MM cell lines treated with CUR and AFA, alone or in combination, at 10 µM for 24 h.Beclin-1 acts as a positive regulator in many steps of autophagy, LC3-II is a marker of autophagosomes and SQSTM-1/p62 is a selective autophagy substrate.Beclin-1 or LC3-II accumulation in the absence of a significant reduction of p62 levels indicates a late-stage inhibition of the autophagic process (Palumbo et al. 2022;Benvenuto et al. 2023).
In CUR-, AFA-and AFA + CUR-treated MM-F1 cells, despite the reduction of Beclin-1, no significant changes on autophagy could be detected on the basis of LC3-II and p62 levels.Conversely, in the other three cell lines the changes in the individual markers investigated were consistent with a variable inhibition of the autophagic process (Figure 4).On the whole, minor or no significant differences in the extent of such inhibition were observed when comparing the effect of AFA + CUR with those of AFA alone on H-Meso-1 and MM-B1 cells.On the other hand, in #40a cells the treatment with AFA increased both Beclin-1 and p62, and a further increase of p62 was induced by AFA + CUR, suggesting a mild potentiation of autophagy inhibition by the drug combination vs. AFA alone (Figure 4).

Effects of CUR and AFA, alone or in combination, on expression of EGFR and ErbB2 and activation of downstream signalling effectors
Expression of EGFR and ErbB2 and activation of the downstream pro-survival signalling effectors ERK1/2, p38 and AKT (Benvenuto et al. 2023) were next analysed by Western blotting after 24 h of treatment with AFA, CUR or AFA + CUR (10 μM) (Figure 5).AFA affected EGFR levels in only two of the four MM cell lines.The effects were cell line-dependent since EGFR was increased in MM-B1 and decreased in #40a cells, and slightly but significantly upregulated by the simultaneous administration of AFA + CUR (Figure 5).By comparison, AFA had a striking effect on ErbB2, leading to the nearly complete downregulation of this receptor in all four cell lines (Figure 5).Further, in all the cell lines tested the downregulation ErbB2 was paralleled by a marked reduction of phospho-ERK1 and phospho-ERK2 levels.However, as regards ErbB2 Western blotting performed on H-Meso-1 and #40a cell lines treated for 24 h with afa, cur, afa + cur (10 μM), or dMSo (images have been cropped).(c) densitometric ratios between Bax and Bcl-2, γH2aX and actin, cleaved ParP-1 and ParP-1.the intensity of the bands from two different experiments was quantified using the ImageJ 1.53e software.actin was used as an internal control.(a, c) data are expressed as the mean ± Sd of two independent experiments.Statistical significance of the effects obtained with cur and afa, alone or in combination, was calculated vs those obtained in (1) dMSo-, (2) cur-and (3) afa-treated cells with one-way anoVa.the statistical significance of the effect obtained with cur, afa or afa + cur in the presence of the Z-Vad-fMK inhibitor (Z-Vad) was calculated with respect to that obtained in cells treated with cur, afa or afa + cur, respectively.×p < .05;*p < .01;#p < .001.
and phospho-ERK1/2 the effect of the combined AFA + CUR treatment was not significantly different from that of AFA, with one notable exception.Indeed, the AFA + CUR treatment was more potent than the treatment with AFA alone in decreasing phospho-ERK2 levels in H-Meso-1 cells, consistent with the strong pro-apoptotic effect that the combined treatment exerted in this cell line (Shukla et al. 2011).Furthermore, this latter finding was mirrored by the increased efficacy of the drug combination vs. AFA alone in reducing the levels of the phospho-active pro-survival kinase AKT in H-Meso-1 cells, but not in the other cell lines.As for the p38 kinase, its phosphorylation levels were reduced to a variable extent by AFA in two out of the four cell lines investigated, and not further modified by the combination of AFA with CUR.

CUR potentiates AFA in inhibiting the growth of #40a MM cells in syngeneic C57BL/6 mice
To evaluate the in vivo antitumor efficacy of AFA and CUR, alone or in combination, groups of C57BL/6 mice (n = 9/group) were i.p. inoculated with 1x10 6 syngeneic #40a MM cells and treated via the same route with AFA, CUR, AFA + CUR or with the drug vehicles alone (PBS + DMSO and corn oil).Treatments began simultaneously with the inoculation of cells and were then repeated twice a week.The i.p. inoculation of #40a cells induces ascites, so that to monitor cancer cell growth the mice abdominal circumference was measured before cell inoculation and weekly thereafter.
After 21 days of treatment, mice treated with AFA and AFA + CUR showed a significantly lower increase in the abdominal circumference as compared to both control mice (AFA 7.3 vs. CTR 8 cm, p < .01;AFA + CUR 7.1.vs. CTR 8 cm, p < .001)and CUR-treated mice (AFA 7.3 vs. CUR 7.9 cm, p < .01;AFA + CUR 7.1.vs. CUR 7.9 cm, p < .001)(Figure 6). .effects of cur and afa, alone or in combination, on expression of eGfr and erbB2 and activation of downstream signalling effectors.(a) Western blot analysis on cells treated for 24 h with afa, cur, afa + cur (10 μM), or dMSo (images have been cropped).(b-g) densitometric ratios between eGfr and tubulin (b), erbB2 and tubulin (c), perK1 and erK (d), perK2 and erK (e), pp38 and p38 (f), paKt and aKt (g). the intensity of the bands from two independent experiments was quantified using the ImageJ 1.53e software.tubulin was used as an internal control.data are expressed as the mean ± Sd of two independent experiments.Statistical significance of the effects obtained with cur and afa, alone or in combination, was calculated vs those obtained with (1) dMSo, (2) cur and (3) afa with one-way anoVa (×p ≤ .05,*p ≤ .01,#p ≤ .001).
After 28 d, the gain in the abdominal circumference of AFA-and AFA + CUR-treated mice was further reduced in comparison to that of control and CUR-treated mice (AFA 7.7 vs CTR 9.0 cm, p < .001;AFA + CUR 7.3 vs. CTR 9.0 cm, p < .001;AFA 7.7 vs. CUR 8.6 cm, p < .01;AFA + CUR 7.3 vs. CUR 8.6 cm, p < .001).After 35 d it became evident the potentiating effect of CUR on AFA antitumor activity, as showed by the reduced gain in the mean abdominal circumference of AFA + CUR-treated mice as compared to those treated with AFA alone (AFA + CUR 7.9 vs. AFA 8.6 cm, p < .05)(Figure 6).The increased antitumor activity of the AFA + CUR combination was also evident in terms of extended survival.Indeed, the combined treatment significantly increased mice mean survival as compared to CUR alone (42 v.s 29 d, p = .0015)and AFA alone (42 vs. 38 days, p = .047).According to the analysis of mice survival by the log-rank test (Mantel-Cox), the Hazard Ratio for AFA-treated mice was 3.3 times greater than that of AFA + CUR-treated mice (Figure 6 and Table 1).
These findings demonstrate that, even though AFA has a strong antitumor activity on its own, a more potent antitumor efficacy on MM cells can be achieved via its combination with CUR.

Discussion
CUR, a non-flavonoid polyphenol found in the plant Curcuma longa, is widely used in traditional Asian medicine and indicated as "the golden spice" due to its multiple anti-inflammatory, antimicrobial and anticancer properties (Gupta et al. 2013;Patel et al. 2020).CUR potential as an antitumor agent relies on its ability to target a variety of signalling pathways involved in cell proliferation, cell death, autophagy, angiogenesis, invasion, migration, metastasis and chemoresistance (Giordano and Tommonaro 2019;Arena et al. 2021;Kubczak et al. 2021).While the preclinical efficacy of CUR on MM, as well as that of AFA, has been previously demonstrated (Okita et al. 2015;Huang et al. 2017;Masuelli et al. 2017;Sayeed et al. 2017;Baldi et al. 2020;Bei et al. 2022), in the present study we tested the hypothesis that CUR could potentiate AFA antitumor activity on MM, using cell lines cultured in vitro and mice bearing intraperitoneally-transplanted, syngeneic MM cells.The rationale behind this hypothesis was that CUR could counteract the activation of alternative "bypass" pathways responsible for decreased sensitivity to the ErbB TKIs AFA (Fantini et al. 2015;Chmielewska-Kassassir and Wozniak 2021).
The complementary information provided by SRB assays and FACS analysis demonstrates that CUR can actually potentiate the effect of AFA, and that the stronger antitumor effect of the drugs combination can be ascribed to a reduced proliferation rate, in the absence of an evident arrest of the cell cycle, or to an  increased apoptosis.Remarkably, the stronger pro-apoptotic effect of the drugs combination was observed in the two cell lines with an epithelioid phenotype, H-Meso-1 and #40a, but not in the sarcomatous MM-F1 and biphasic MM-B1 cell lines.In light of these findings, it is tempting to speculate that the mechanisms underlying the interaction between CUR and AFA could be histotype-dependent.Autophagy exerts well known tumour-supporting functions, since it allows tumour cells to cope with endogenous and environmental stress conditions (Masuelli et al. 2017;Amaravadi et al. 2019;Benvenuto et al. 2020).Consistent with the results of our previous study (Benvenuto et al. 2023), AFA inhibited autophagy in H-Meso-1, MM-B1 and #40a cells.However, minor or no significant differences in the extent of such inhibition were observed when comparing the effect of AFA + CUR with those of AFA alone, except for a mild potentiation of autophagy inhibition by the drug combination vs. AFA in #40a cells.These results indicate that, on the whole, the improved antitumor efficacy of the AFA + CUR combination should rely on other mechanisms than the modulation of the autophagic process.
Our analysis of the effects of the single and combined treatments on EGFR and ErbB2 revealed that the more consistent changes involved the latter receptor, which was dramatically downregulated by AFA in all the cell lines tested, in agreement with previous findings obtained in MM and other tumour cell types (Janjigian et al. 2013;Yonesaka et al. 2015;Benvenuto et al. 2023).In this regard, it has been recently demonstrated that, unlike the reversible TKIs Gefitinib and Erlotinib, the irreversible inhibitors AFA and Dacomitinib, which also targets both EGFR and ErbB2, selectively downregulate ErbB2 and it has been proposed that such downregulation is mediated by the endocytosis of the receptor from the plasma membrane, its accumulation at endosomes and subsequent degradation (Ghosh et al. 2023).
Notably, the AFA-induced downregulation of ErbB2 we observed in MM cells was nearly complete and it was not affected by the simultaneous administration of CUR, indicating that the mechanisms by which CUR potentiates the effects of AFA should act downstream of ErbB receptors.The ERK1/2 kinases are the main signalling effectors activated by EGFR and ErbB2 and indeed the phospho-active levels of these proteins were decreased by AFA in all four MM cell lines.Also in this respect, however, the effect of the combined treatment was not significantly different from that of AFA alone in MM-F1, MM-B1 and #40a cells.Instead, in H-Meso-1 cells, i.e. the cell line where the AFA + CUR combination induced the stronger increase in apoptosis nearly doubling the percentage of cells in subG1 as compared to AFA alone, the combined treatment was also more effective than AFA in decreasing phospho-ERK2 levels as well as in decreasing the phosphorylated levels of the pro-survival kinase AKT.Moreover, as far as the reduction of phospho-ERK2 is concerned, the effect of CUR and AFA in H-Meso-1 cells was actually synergistic since when CUR was administered alone to the cells, at the opposite, it slightly increased the levels of the phosphorylated kinase.Altogether these results indicate that a strong potentiation of AFA pro-apoptotic effects by CUR may involve a synergistic inhibition of ERK2.Still, further studies will be required to define the mechanisms responsible for the synergistic inhibition of cell proliferation observed in the cell lines with non-epithelioid histotype.
Beside the findings of in vitro studies, the results obtained in mice transplanted with the syngeneic #40a MM cells demonstrate that, even though AFA has a strong antitumor activity on its own, a more potent antitumor efficacy on MM can be achieved also in vivo via its combination with CUR.Indeed, the combined treatment significantly increased the mice mean survival as compared to the treatment with the individual compounds.
While this study was focused on the effects of the AFA + CUR combination on tumour cells, an interesting issue for future investigations is their combined effect on the tumour immune microenvironment.In this respect, abnormal activation of ErbB receptors signalling is known to concur to the evasion of antitumor immunity (Kumagai et al. 2021) and AFA has been shown to modulate the immune function in cancer patients (Tu et al. 2021;Cao et al. 2022).CUR, on the other side, has been traditionally regarded as an immunosuppressive compound, but increasing evidence indicates that it can support anti-tumour immunity by regulating the tumour immunosuppressive microenvironment through multiple mechanisms (Wang et al. 2020;Paul and Sa 2021).Accordingly, the combined activity of AFA and CUR may strengthen the efficacy of immunotherapy approaches in MM patients.
In conclusion, the results of the present study demonstrate that, as compared to the effect of the individual compounds administered alone, the combination of AFA and CUR can have a stronger impact on MM progression which can be ascribed either to increased growth inhibition or to an enhanced pro-apoptotic effect.
Despite the clinical safety, tolerability, and therapeutic efficacy of CUR on various types of cancer (Karaboga Arslan et al. 2022), its clinical applications are known to be limited by its poor stability, solubility and bioavailability in vivo (Dei Cas and Ghidoni 2019;Karaboga Arslan et al. 2022).Still, due to the specific features of MM, which is characterised by high local invasiveness and low metastatic efficiency (Benvenuto et al. 2023), clinically relevant concentrations of the compound could be achieved directly at the tumour site in MM patients via CUR intracavitary administration (Pouliquen et al. 2017;Hocking et al. 2021).These considerations and the results presented herein warrant future studies aimed at further exploring the therapeutic potential of AFA + CUR-based combination regimens for the treatment of MM.

Figure 1 .
Figure1.effect of cur and afa, alone or in combination, on MM cell growth.the growth of human (H-Meso-1, MM-f1, MM-B1) and murine (#40a) MM cell lines was assessed by the SrB assay after 24, 48, and 72 h of treatment with dMSo, cur or afa, alone or in combination.the percentage of surviving cells treated with the compounds was calculated by normalising the o.d.value to that of the control cultures (dMSo).the results are expressed as the mean ± Sd of two independent experiments performed in triplicate.Statistical significance of the effects obtained with cur and afa, alone or in combination, was calculated vs those obtained in (1) dMSo-, (2) cur-, and (3) afa-treated cells with one-way anoVa.×p < .05;*p < .01;#p < .001.

Figure 2 .
Figure 2. Interaction between cur and afa in growth inhibition of MM cell lines.reported are Kern index values (r) at 24, 48 and 72 h of treatment with cur and afa 10 µM.R > 1 indicates that the effect of the two compounds is synergistic, R = 1 indicates that the effect is additive, R < 1 indicates that the effect is less than additive.

Figure 3 .
Figure 3. effect of cur and afa, alone or in combination, on apoptosis induction.(a) effect of cur, afa and afa + cur, in the presence or absence of the Z-Vad-fMK inhibitor, on the percentage of cells in subG1 phase, as evaluated by facS analysis.(b)Western blotting performed on H-Meso-1 and #40a cell lines treated for 24 h with afa, cur, afa + cur (10 μM), or dMSo (images have been cropped).(c) densitometric ratios between Bax and Bcl-2, γH2aX and actin, cleaved ParP-1 and ParP-1.the intensity of the bands from two different experiments was quantified using the ImageJ 1.53e software.actin was used as an internal control.(a, c) data are expressed as the mean ± Sd of two independent experiments.Statistical significance of the effects obtained with cur and afa, alone or in combination, was calculated vs those obtained in (1) dMSo-, (2) cur-and (3) afa-treated cells with one-way anoVa.the statistical significance of the effect obtained with cur, afa or afa + cur in the presence of the Z-Vad-fMK inhibitor (Z-Vad) was calculated with respect to that obtained in cells treated with cur, afa or afa + cur, respectively.×p < .05;*p < .01;#p < .001.

Figure 4 .
Figure 4. effect of cur and afa, alone or in combination, on autophagy.(a) Western blotting analysis on cells treated for 24 h with afa, cur, afa + cur (10 μM), or dMSo (images have been cropped).(b-d) densitometric ratios between Beclin-1 and actin (b), SQStM-1/p62 and actin (c), lc3-II and lc3-I (d). the intensity of the bands from two different experiments was quantified using the image J 1.53e software.actin was used as an internal control.data are expressed as the mean ± Sd of two independent experiments.Statistical significance of the effects obtained with cur and afa, alone or in combination, was calculated vs those obtained in (1) dMSo, (2) cur-and (3) afa-treated cells with one-way anoVa (×p ≤.05, *p ≤ .01,#p ≤ .001).

Figure 5
Figure5.effects of cur and afa, alone or in combination, on expression of eGfr and erbB2 and activation of downstream signalling effectors.(a) Western blot analysis on cells treated for 24 h with afa, cur, afa + cur (10 μM), or dMSo (images have been cropped).(b-g) densitometric ratios between eGfr and tubulin (b), erbB2 and tubulin (c), perK1 and erK (d), perK2 and erK (e), pp38 and p38 (f), paKt and aKt (g). the intensity of the bands from two independent experiments was quantified using the ImageJ 1.53e software.tubulin was used as an internal control.data are expressed as the mean ± Sd of two independent experiments.Statistical significance of the effects obtained with cur and afa, alone or in combination, was calculated vs those obtained with (1) dMSo, (2) cur and (3) afa with one-way anoVa (×p ≤ .05,*p ≤ .01,#p ≤ .001).

Figure 6 .
Figure 6.antitumor efficacy of afa and cur, alone or in combination, in c57Bl/6 mice i.p. transplanted with syngeneic MM #40a cells.Shown are the differences in (a) mean abdominal circumferences and (b) survival time of mice treated with cur, afa, afa + cur (a + c) or drug vehicles (ctr).

Table 1 .
analysis of mice survival after treatment with cur, afa, afa + cur or drug vehicles (ctr) by the log-rank test (Mantel-cox).