RU2575828C2 - Method for identifying compounds for treating cancer - Google Patents

Abstract

FIELD: medicine, pharmaceutics.

SUBSTANCE: invention refers to oncology and concerns a complex containing a combination of polyinosinic-polycytidylic acid (pIC) and polyethylenimine (PEI), and using it as an agent for treating cancer, particularly melanoma.

EFFECT: invention contains a pharmaceutical composition containing the complex applicable in treating melanoma, wherein the above complex induces autophagia in melanoma cells or in a cell line produced from the melanoma cells, and the above complex contains the combination of pIC and linear PEI in N/P ratio 1:5.

10 cl, 1 tbl, 13 ex, 15 dwg

Description

Technical field

The present invention relates to the field of oncology, and it is mainly directed to the identification of compounds that can be used to treat various types of cancer, for example: melanoma, cancer of the pancreas, colon, bladder, glioma, breast, prostate, lung and ovary .

In addition, the present invention also encompasses compounds defined in this manner, for example, compound BO-110 (see below), which is capable of causing distinct tumor cell death in all of the above types of cancer.

State of the art

Melanoma remains the prototype of solid cancers with an increasing number of diseases and an extremely poor prognosis in the later stages (Jemal et al., 2008). Significant efforts have been devoted to identifying the molecular determinants that underlie the chemical and immune resistance of melanoma. The only products approved by the US Food and Drug Administration (FDA) for the treatment of metastatic melanoma are the dacarbazine alkylating agent (DTIC) and the immunomodulator IL-2 (Tawbi and Kirkwood, 2007). However, a reliable and complete response to treatment with metastatic melanoma is rarely found in more than 5% of patients, and secondary toxicity can be very serious. As a result, currently the average life expectancy of patients with metastatic melanoma is from 6 to 10 months and, therefore, the development of more advanced treatment methods is a priority for this disease (Jemal et al., 2007).

Initially, a synthetic viral dsRNA analogue called pIC (polyinosine-polycytidylic acid), a compound that has been used for more than four decades to stimulate the immune system independently of interferon (IFN) (Field et al., 1967), was considered a promising therapeutic agent against melanoma. Unfortunately, clinical studies with naked pIC revealed its low stability, low degree of induction of interferon, and the absence of an antitumor effect in melanoma (Robinson et al., 1976). Thus, as a monotherapy, pIC is considered an ineffective anti-melanoma drug.

High-performance histogenetic analysis and systematic functional studies have greatly expanded our understanding of the occurrence and progression of melanoma and the complex mechanisms associated with treatment failure (Fecher et al., 2007; Gray-Schopfer et al., 2007). Corresponding defects and changes in the signal cascades of BRAF / MAPK reactions were identified; PI3K / AKT, NF-κB, or NOTCH, constituting a promising platform for rational drug development (Gray-Schopfer et al., 2007). However, targeted therapy has not yet been proven effective in trials of melanoma (Flaherty, 2006). Cell death programs controlled by mitochondria and / or the endoplasmic reticulum also undergo testing, although they are invariably ineffective in vivo (Hersey and Zhang, 2008). Therefore, the currently used anticancer drugs either do not achieve their goal (s) productively, or must be administered in dosages that lead to intolerant toxicity for normal cell compartments (Tawbi and Kirkwood, 2007). It is important to note that during treatment, compensatory mechanisms can be activated, leading to the selection of cell populations with even greater resistance to chemicals (Lev et al., 2004; Shatton et al., 2008; Wolter et al., 2007).

In fact, it is believed that melanoma has a strong ability to avoid apoptosis through various pathways, which gives the melanoma the ability to progress, form metastases and not respond to treatment with various therapeutic methods (review by Ivanov et al., 2003).

Unlike standard chemotherapy aimed at destroying tumor cells mainly “from the inside” (that is, by activating the internal program of cell death), indirect cascades of intercellular interactions are traditionally involved in immunotherapy. In melanoma, the main efforts were focused on increasing the content or functional effectiveness of two components: antigen-presenting cells and cytotoxic T cells (Wilcox and Markovic, 2007). Vaccines, as well as antibodies directed against inhibitory immunomodulators (e.g. CTL4), are also being tested, albeit with disappointing results in phase IV clinical trials (Kirkwood et al., 2008). Recently, studies have been conducted on stimulation of the immune system by activating Toll-like receptors (TLRs) -3, -4, -7 and -9 to provide cytotoxic destruction of melanoma cells using natural killer cells (NK), dendritic cells (DC) and T- cells (Kirkwood et al., 2008, and Tormo et al., 2007). However, numerous studies, including those of the inventors, have demonstrated the inherent ability of cells to bypass immunological therapy by downregulating (adjusting) immunoreactive surface markers. Melanomas can also have an inhibitory effect on the host (e.g., inhibiting the maturation of antigen presenting cells or blocking the complete activation of T cells) (Tormo et al., 2006; Ilkovitch and Lopez, 2008; Verma et al., 2008). Thus, melanomas demonstrate their inherent ability to avoid the antitumor effect of immunomodulators.

In the field of immunotherapy, one of the molecules whose increase is being studied as a potential positive factor for treating melanoma is MDA-5 (gene 5 associated with differentiation of melanoma), a product originally described as a gene associated with melanoma differentiation (Kang et al., 2002 ) MDA-5 is a helicase that recognizes and is activated by long double-stranded RNA (dsRNA) (Yoneyama et al., 2005). Other RNA helicases are RIG-1 (retinoic acid-inducible protein 1, also called Dsx58), which recognizes short-dsRNA naked 3-phosphates, and LGP2 (also called Dhx58), which is a negative regulator in the recognition of dsRNA.

As long as dsRNA can be created as a result of and during viral infections, MDA-5 acts as the first line of innate immunity against viral pathogens (Akira et al. , 2006). In addition, MDA-5 has caspase activation recruitment domains (CARDs). Helicase and CARD domains co-activate NF-κB and other transcription factors involved in the regulation of cytokines (Kawai et al., 2005). Thus, the most famous function of MDA-5 is immune stimulation.

From a therapeutic point of view, it is known that both IFN-β and dsRNA induce transcription of the Mda-5 gene. As a result, it has been suggested that dsRNA plays a role in increasing the expression of Mda-5 in IFN-induced growth inhibition. In addition, it was shown (Kang et al., 2002) that the induction of endogenous MDA-5 using IFN-β is cytostatic (in other words, it stops the cell cycle). Thus, to activate the death of tumor cells, MDA-5 must be overexpressed ectopically at high levels (Kovacsovics et al., 2002). Moreover, this pro-apoptotic activity of ectopic expression of MDA-5 is not effective in tumor cells with the hyperactive RAS / MEK / ERK pathway (Lin et al., 2006), as is the case with melanomas (Chin et al., 2006). Thus, an unresolved issue in this area was how to activate endogenous MDA-5 with chemotherapeutic agents exclusively for the tumor area (without causing secondary toxicity in normal cells).

US 2007/0259372 proposes the identification of IFN-β, IFN-α or IFN-γ agonists or antagonists using compounds capable of enhancing MDA-5 expression. This patent also suggested that Mda-5 gene expression inducers (using its promoter) can be considered candidate compounds for inducing terminal differentiation of tumor cells. It also suggested that MDA-5 could play a role in generating apoptotic signals through its CARD domain. However, until now it was not known which MDA-5 targets are capable of triggering apoptosis, and how to activate it in a manner that is monitored and selective for tumor cells. In addition, since melanoma cells have an active RAS / MEK / ERK pathway (which inhibits MDA-5), as well as a pronounced ability to avoid apoptotic cell death, it was not clear that apoptotic signals mediated by MDA-5 may be an effective mechanism for therapy against melanoma. Consequently, based on previous information on the regulation and function of MDA-5, this protein was not considered a convincing target for methods to identify candidates for therapeutic agents against melanoma.

Autophagy is becoming an alternative strategy to activate the endogenous mechanism of cancer cell death.

This process involves a complex cascade of events that ultimately lead to sequestration of cytosolic components for subsequent degradation of lysosomes (Xie and Klionsky, 2007). Many mechanisms of autophagy are described depending on the mechanism of absorption and the nature of the cargo delivered to autolysosomes. In the context of anti-cancer treatment, macroautophagy or mass degradation of cellular organelles and protein aggregates is of interest because of its ability to jeopardize cell viability due to dysfunction or excessive depletion of major organelles (e.g., endoplasmic reticulum or mitochondria) (Hoyer-Hansen, 2008) .

However, it is not known how autophagy is regulated, and its therapeutic potential is not clear and simple (Hippert et al., 2006). On the other hand, macroautophagy (which will be referred to simply as “autophagy” later in this document) has demonstrated significant potential for protecting cells from various aggressive intracellular and extracellular signals, including anti-cancer drugs. Due to this activity, autophagy can stimulate tumor development (Mizushima et al., 2008; Kroemer and Levine, 2008).

Paradoxically, autophagy is also associated with cell death (Kromer et al., 2009). Thus, excessive or persistent autophagy can contribute to cell death due to depletion of major organelles (i.e., endoplasmic reticulum or mitochondria), changes in survival signals, deregulation of lysosomal enzymes, and / or activation of caspase-dependent apoptosis programs (Xie and Klionsky, 2007).

Therefore, it is unclear whether autophagy will enhance the chemical and immune resistance of melanoma instead of improving treatment outcomes. In addition, none of the more than 20 autophagy genes currently described in mammalian cells has been characterized in detail in the case of melanoma. As a result of this, it is not known whether autophagy is regulated in various ways in melanoma cells and normal cells, which could provide an opportunity for therapeutic intervention. A similar situation occurs in the case of aggressive cancers, such as those affecting the pancreas, bladder, prostate gland and brain.

In this situation, there remains a need to identify therapeutic agents for the treatment of cancer, alternative to those that are already officially used, and in particular those that are suitable for the treatment of immunocompromised patients. There remains also the need to discover possible new therapeutic targets for the development of methods for identifying candidate therapeutic agents for cancer among compounds that can target these targets.

Thus, the present invention provides a solution to both problems.

Description of the invention

As indicated above, the present invention is primarily directed to the identification of therapeutic targets, markers or parameters that provide the basis for developing a method useful for identifying compounds (among those that are capable of acting on these therapeutic targets, markers or parameters) capable of treating cancer.

One of the markers defined in the present invention, which is useful for identifying compounds that can treat cancer, is the level of activation of the helicase of the MDA-5 family. This parameter can be determined by checking the presence of proteolytic cleavage of the protein, which leads to the separation of helicase and caspase domains: in order to be a therapeutic agent for cancer, the candidate compound must lead to proteolytic cleavage, which indicates that it can cause activation autophagy and apoptosis mechanisms leading to the death of cancer cells. A possible methodology for this test is immunoblotting (Western blotting) of protein extracts of cell culture and verification of signal bands corresponding to the whole protein and fragments corresponding to helicase domains and caspase domains. In particular, as shown in Example 3, the appearance of a 30 kDa band indicates the presence of proteolytic cleavage. Alternatively, this may also indicate activation of other helicases of the MDA-5 family, such as RIG-I or LGP2.

Another marker defined in the present invention also useful for identifying compounds capable of treating cancer is the level of NOXA expression. The rationale is to increase the expression levels of the corresponding genes when the mechanisms of autophagy and apoptosis are activated. Determination of the expression levels of these proteins can be carried out, for example, by determining the concentration of the corresponding messenger RNA (this can be done, for example, by Northern blotting or RT-PCR) or the concentration of the protein itself in the protein extract of the corresponding cell culture (for example, by transfer, similar to Western blotting). In addition, NOXA can be determined in situ (in tissue samples) by immunohistochemical methods.

In a preferred embodiment of the present invention, both MDA-5 and NOXA are determined as confirmation that the mechanisms of autophagy and apoptosis are activated, since MDA-5 is considered to be a link between them.

In addition, the invention may include the step of determining the induction of autophagy by a candidate compound for use against cancer. Autophagy induction can be determined by several methods, which include:

• Monitoring post-translational modification of a protein expressed by autophagy gene 8 (called ATG8 or LC3). The LC3 protein is processed and lipidated upon incorporation into autophagosomes, which are membrane structures into which cellular components are captured during autophagy. Due to the fact that the conformation and electrophoretic mobility of this protein changes upon lipidation, one of the possible methods for checking the induction of autophagy is the use of immunoblotting with antibodies directed against this protein. This method makes it possible to verify that after electrophoresis, the band corresponding to the protein differs in the control samples compared to samples in which, presumably, the compound induced autophagy. Alternatively, if the antibody specifically recognizes the shape of the autophagosomes, binding of the antibody will confirm the induction of autophagy in the sample treated with the candidate compound.

• Monitoring changes in the intracellular distribution of the LC3 protein, since another distinguishing feature of autophagy is the movement of LC3 from the cytosol to newly formed autophagosomes (Xie and Klionsky, 2007). Thus, the formation of protein foci can be considered an indicator of the formation of autophagosomes, especially in the early stages. Detection can be carried out by monitoring endogenous LC3 by immunofluorescence or immunochemistry methods on fixed cells or fixed tissues. Alternatively, autophagy can be visualized in living cells by determining the cellular localization of the fluorescent derivative of LC3. As a fluorescent protein, it is customary to use GFP (green fluorescent protein) or RFP (red fluorescent protein), which can monitor autophagy using fluorescence microscopy: changes in the cell distribution of fluorescence from diffuse to focal staining indicate the induction of autophagy. In the present invention, this technique involves the use of either cells that have been transfected with a vector that allows the transient expression of a fusion protein (such as a plasmid or recombinant virus), or cells into which a DNA fragment capable of expressing a fusion protein formed by a reporter protein and LC3, integrated into the genome stably. An example of such a strategy is shown below in example 1, when the cells were pre-transfected with recombinant retrovirus. This led to the insertion into the cellular genome of a DNA fragment containing together the coding parts of the GFP and LC3 proteins, so that this leads to the formation of a fusion protein in the cellular genomic DNA. Thus, it is possible to conduct tests for changes in the intracellular distribution with stably transfected cells.

• Use of transmission electron microscopy to detect penetration of a candidate compound into a cell. This situation controls the formation of autophagosomes in the later stages of the autophagy process. Visualization of the accumulation of dense structures (associated with the membrane) is considered a characteristic feature of the formation of autophagosomes. The process of autophagy can be confirmed at later stages of the process (that is, 24 or 30 hours after treatment with the test compound), during which the formation of large phagocytic vacuoles, indicating cell death, can be seen with an electron microscope.

In view of the foregoing, a first embodiment of the present invention relates to a method (hereinafter referred to as the method of the invention) for identifying compounds used to treat cancer, comprising the steps of:

a) bringing the candidate compound into contact with a cancer cell culture or a cancer cell line derived from cancer cells;

b) determining at least one of the following parameters:

i. the level of activation of the helicase family MDA-5;

ii. NOXA expression level;

iii. or a combination thereof;

c) comparing the data obtained in step b) with the data observed in the control from the same cells treated in the same way, but in the absence of the candidate compound;

d) the choice of compounds that led to a significant increase in the parameter or parameters defined in paragraph b), compared with the control.

It should be noted that the difference between the data obtained from the cell culture treated with the candidate compound and from the untreated control will be considered statistically significant if p <0.05 is obtained in the analysis.

In a preferred embodiment, the method of the invention also determines whether the candidate compound induces autophagy in cancer cells, in a cell line derived from cancer cells, or in a murine model of cancer. As explained above, the determination of autophagy induction can be carried out by checking the expression level, the presence of a post-translational modification, or the intracellular localization of the autophagy protein. More specifically, autophagy induction is determined by a method selected from: changing the electrophoretic mobility of the LC3 protein or detecting the formation of foci of the LC3 protein. Alternatively, the induction of autophagy is determined by checking the presence of autophagosomes using microscopic examination, for example, using transmission electron microscopy.

In a further preferred embodiment, the method of the invention described above comprises three steps: determining the level of activation of MDA-5, the level of expression of NOXA, and the induction of autophagy.

The method according to the invention can be used to identify compounds used as therapeutic agents for treating various types of cancer, for example: melanoma, cancer of the pancreas, colon, bladder, breast, prostate, lung and ovary.

Therefore, if the present invention is directed to the identification of compounds for use as therapeutic agents for the treatment of melanoma, then it will include the following steps:

a) bringing the candidate compound into contact with a melanoma cell culture or a cell line derived from melanoma cells;

b) determining at least one of the following parameters:

i. MDA-5 family helicase activation level;

ii. NOXA expression level;

iii. or a combination thereof;

c) comparing the data obtained in step b) with the data observed in the control from the same cells treated in the same way, but in the absence of the candidate compound;

d) the choice of compounds that led to a significant increase in the parameter or parameters defined in paragraph b), compared with the control.

As for suitable cell lines, any melanoma cell line, preferably human, can be used. Examples of suitable cell lines used in the examples of the invention are human cell lines SK-Mel-19, SK-Mel-28, SK-Mel-103 and SK-Mel-147, as well as murine B16 cells. Normal control cells are melanocytes or other skin cells, as well as cells of the immune system, which, as a rule, are centers of secondary toxicity in the treatment of cancer.

Alternatively, the method of the invention can be directed to identifying compounds for use as therapeutic agents for treating at least one of the following cancers: carcinomas of the pancreas, colon, bladder, breast, prostate, lung, and ovary. In this case, the method according to the invention will include the following steps:

a) bringing the candidate compound into contact with a cancer cell culture of at least one of the above types of cancer, or a cell line derived from cells of at least one of the above types of cancer;

b) determining at least one of the following parameters:

i. MDA-5 family helicase activation level;

ii. NOXA expression level;

iii. or a combination thereof;

c) comparing the data obtained in step b) with the data observed in the control from the same cells treated in the same way, but in the absence of the candidate compound;

d) the choice of compounds that led to a significant increase in the parameter or parameters defined in paragraph b), compared with the control.

In this case, the cell line will be selected from the group of pancreatic cancer cell lines: IMIMPC2, MiaPaCa2, Aspc1, A6L, SKPC1 and Panc-1; or from the group of colon cancer cell lines: CACO, SW480 and SW1222; or from a group of bladder cancer cell lines: RT112, MGHu4, 639V, 253J, MGHu3 and SW1170; or from a group of glioma cell lines: U87MG, U251 and T98G; or from a group of breast cancer cell lines: MDA231, MCF7, and T47D; or from a group of prostate cancer cell lines: LNCaP, PC3, and DU145; or from a group of lung cancer cell lines: H1299 and NCI H460; or from the group of ovarian cancer cell lines: NCI H23, CHQK1 and SK-OV-3.

A preferred embodiment of the method of the invention is described below in the examples of the invention. In this case, the method according to the invention is carried out using a combination of determining the activation of MDA-5, analysis of gene expression (observing an increase in NOXA expression) and confirming the activation of autophagy by the three possible methods described previously: monitoring post-translational modifications of the LC3 protein by immunoblotting, tracking changes in cell distribution LC3 by fluorescence detection by fluorescent GFP protein (with cells previously transfected with a recombinant retrovirus containing the structure, capable of expressing the GFP-LC3 fusion protein) by confirming the formation of autophagosomes by electron transmission microscopy after 5 hours of treatment with the candidate compound and confirming the presence of phagocytic vacuoles after 30 hours of treatment.

It should be noted that the above method according to the invention allows to identify a new compound, which is a combination of double-stranded RNA (dsRNA) or its analogue and polycation. In a preferred embodiment of the invention, said compound is BO-110 (pIC ^PEI ), which is a combination of polyinosine-polycytidylic acid (pIC) and polyethyleneimine (PEI).

As shown in the examples and figures of the present invention below, BO-110 can be effectively used to treat various types of cancer, for example: melanoma, cancer of the pancreas, colon, bladder, breast, prostate, lung and ovary.

Therefore, the present invention also relates to a pharmaceutical composition containing BO-110, for use in the treatment of cancer, for example: melanoma, cancer of the pancreas, colon, bladder, breast, prostate, lung and ovary. This pharmaceutical composition is also useful for treating immunocompromised patients.

Surprisingly, with the functional interaction of pIC and PEI, a synergistic technical effect is achieved, as a result of which the sum of the technical effects of the individual components improves and changes. Thus, BO-110 penetrates the cells and acts in a different way than its components PEI and pIC. In particular, while PEI has no noticeable cellular effect, and isolated pIC signals temporarily elicit immune responses that ultimately have no biological effect in vivo , BO-110 has the ability to selectively kill tumor cells. As a result, BO-110 illustrates the concept of synthetic lethality described for genes or compounds that, as separate substances, have no activity, but when combined, have a different and therapeutically applicable anti-cancer effect.

Therefore, one of the most important points of the present invention is the unexpected discovery that the polyinosine-polycytidylic acid (pIC) viral dsRNA mimetic changes its route of entry and delivery to tumor cells. From standard recognition of TLR-3 (Toll-like receptor 3), pIC can be targeted to a family of dsRNA sensors (other than TLR3) if this dsRNA is combined with a family of carriers that specifically allow delivery to the cytosol. Such activity changes the mechanism of action of dsRNA from a non-essential immunomodulator to a large-scale killer of tumor cells. Anticancer activity has been demonstrated with polyethyleneimine (PEI), as well as with lipofectamine, polyfect, or superfect. Although these carriers themselves were not biologically active as therapeutic agents, the present invention shows that they are able to protect the pIC molecule by maintaining it in a stable form, which allows activation of autophagy. Thus, the dsRNA / polycation combination, exemplified by BO-110, represents a new molecular formation with anticancer activity. Moreover, the mechanism of action of BO-110 was unexpected. This compound promotes the double induction of apoptosis and autophagy, leading to the coordinated and selective destruction of tumor cells, in particular, but not exclusively, melanoma cells, without affecting the viability of normal compartments. The apoptosis mechanism was triggered by the NOXA protein. Unlike other NOXA-inducing chemotherapeutic agents, BO-110 does not require a p53 tumor suppressor protein. This is an important advantage, since p53 is mutated, absent, or inactivated in the vast majority of human cancers. The effect is clearly superior to classical responses to naked viral RNA, which explains the poor results of clinical studies of naked pIC in the treatment of melanoma.

Thus, BO-110, but not a single dsRNA, was able to stimulate self-destruction of cancer cells and block cancer metastasis in vivo , even in mice with weakened immune systems. The genetic, functional and ultrastructural analysis described later in this invention shows that the induction of autophagy is triggered by pIC not to protect cells, but to selectively kill tumor cells. As further confirmation of these results, although pIC was considered as an inducer of immunity controlled by IFN, the observed effect does not depend on the activation of the pathway for production and secretion of IFN-α. According to this observation, the activation of the autophagy pathway occurs even in animals with weakened immunity. Together, these data show that BO-110 targets and identifies new points of application of effort for the clinical use of its own pathogen recognition programs, autophagy and tumor cell death.

Genetic and functional studies have identified endogenous MDA-5 as a link between autophagy and apoptosis pathways due to BO-110. This also differs from the previous disclosure regarding MDA-5, which was limited to apoptotic cell death caused by exogenous components. Also new was the interaction of MDA-5 / NOXA.

Thus, MDA-5 is presented as a suitable therapeutic target for screening substances for treating cancer with the peculiarity of triggering self-destruction of a tumor using auto / lysosomal and apoptotic proteases. As mentioned earlier, this strategy has advantages over standard therapeutic agents that use any of these mechanisms independently.

Similarly, the formation that allowed this pathway to be opened, since it is capable of activating it, is a BO-110 complex or other combinations of a dsRNA analog and a cationic carrier. As a result, these substances are good candidates for use in the manufacture of drugs for the treatment of cancer.

In the present invention, the term “long double-stranded RNA fragment” is used in contrast to RNA fragments known as short RNAs or interfering RNAs (siRNAs). Because of this, double-stranded RNA (duplex) is considered to be “long” if the double-stranded RNA fragment contains at least 25 nucleotides in the chain. Preferably, if the fragment used is equal in length to the intermediate products of double-stranded RNA appearing in the cells during the cell cycle of most RNA viruses, which are obviously a natural substrate of the MDA-5 helicase family, so the double-stranded RNA according to the invention is considered to be "long", especially if it contains at least 100 nucleotides in a chain and, more specifically, if it contains at least 1000 nucleotides in a chain.

Fragments of double-stranded RNA that are naturally occurring and may be useful for use according to the invention can be intermediate products of double-stranded RNA formed during the cell cycle of paramyxovirus (such as Newcastle disease virus, NDV, Sendai virus (SDV)), rhabdovirus ( vesicular stomatitis virus (VSV)); flavovirus (hepatitis C virus (HCV)), orthomyxovirus (influenza virus) and picornavirus (brain myocarditis virus (EMCV).

As for the double-stranded RNA analogues, besides polyinosine-polycytidylic acid (pIC), other dsRNA mimics may be useful for the invention: a) those whose skeleton is formed by a ribose-like compound, for example, those based on LNA (closed nucleic acid) : resistance to hydrolysis), morpholino and PNA (peptide nucleic acid), b) those in which at least one of the typical nitrogen bases of RNA nucleotides is replaced by analogues, which can also lead to different pairing variants, m like diaminopurin (which forms a pair with uracil through three hydrogen bonds), a xanthine / diaminpyrimidine pair (in which the form of keto / keto purine, xanthine, forms three hydrogen bonds with the form of amine / amine pyrimidine), or an isoguanine / isocytosine pair (in which the amine / keto purine form, isoguanine, forms three hydrogen bonds with the keto / amine pyrimidine form, isocytosine).

As for the polycationic carrier, all those that are capable of modifying the permeability of the plasma membrane and / or inducing endocytosis, stimulating the penetration of double-stranded RNA or its analogues into the cell and releasing them into the cytosol, thereby increasing the activation of cytosolic sensors of double-stranded RNA, are suitable like helicase MDA-5. In addition to polyethyleneimine (PEI) and lipofectamine, poly-L-lysine, polysilazane, polydihydroimidazolinium, polyalylamine and ethoxylated polyethyleneimine (ePEI) fall under this definition.

The invention is further illustrated in more detail by way of examples and figures.

Brief Description of the Figures

Figure 1. Induction of macroautophagy using BO-110 leads to the death of melanoma cells

Panel A presents visualization using epifluorescence of autophagosome-like focal staining of eGFP-LC3 in SK-Mel-103 melanoma cells treated for 12 h with 1 μg / ml pIC in complex with PEI (BO-110). Cells treated only with PEI are shown as controls.

Panel B depicts time-dependent accumulation of SK-Mel-103 cells demonstrating spot fluorescence emission of eGFP-LC3 when treated with BO-110 or a control placebo. Rapamycin was used as a classic autophagy inducer.

Panel C shows immunoblots of the total cell extracts isolated from SK-Mel-103 cells treated as indicated. Processing methods are indicated above each lane: control (without treatment: cell populations were incubated only in the presence of a solvent), pIC or BO-110 treated cells. The analyzed proteins are indicated on the left side of the panel: unmodified ATG8 (LC3 I), lipidated ATG8 (LC3 II) and load control (β-actin). On the right side of the panel are molecular weights, in kDa.

Panel D shows electron micrographs of SK-Mel-103 cells treated with BO-110 or PEI control. Arrows indicate autophagosomes and autolysosomes associated with membranes.

Panel E shows micrographs with strong (top) and weak (bottom) magnification of SK-Mel-103 cells 5 hours after incubation with a solvent (left column) or BO-110 (right column).

Panel F presents typical micrographs in a bright field (left and center) and under an electron microscope (right) of cell colonies after 30 hours of processing, as shown in the pictures.

Figure 2. Microscopy at time intervals of autophagy induction using BO-110 in melanoma cells

The sequence of micrographs obtained at the indicated time intervals after processing PEI (control) or BO-110 expressing eGFP-LC3 melanoma cells SK-Mel-103. Focal aggregates indicate the formation of autophagosomes. Arrows indicate destructive and detaching cells (dying cells) during processing.

Figure 3. Delivery to the cytosol pIC due to the presence of PEI selectively causes death of melanoma cells

Panel A shows a graph in which cell death as a percentage determined in the trypan blue exclusion test is presented after 24 and 48 hours of treatment, as indicated (NT: white bars, PEI processing: dark gray bars, pIC processing: light gray bars, BO-110: black bars). Data are presented as mean values ± SEM from three independent experiments with different cell lines indicated above the diagram. As can be seen, only melanoma cell populations treated with BO-110 die efficiently.

Panel B: electron micrographs of SK-Mel-28 and SK-Mel-103 cells treated with a solvent (control), PEI, pIC or BO-110 and visualized 12 hours after treatment. For each cell line, photographs are taken at two different magnifications. Arrows indicate autolysosomes observed only in cells treated with BO-110.

Figure 4. Selective cytotoxic activity of BO-110 against tumor cells

Panel A shows typical images in the bright field of melanocytes isolated from the human foreskin (upper row), human melanoma cells SK-Mel-103 (middle row), and murine melanoma B16 line (lower row) after treatment with solvent, PEI, pIC or BO-110 as indicated.

Panel B presents dose-response curves for the PEI and pIC treatments individually or in combination (BO-110) (group of columns on the right of each graph) of FM cells (foreskin melanocytes) and SK-Mel-103. The effect is expressed as the percentage of dead cells 24 hours after treatment.

The graph in panel C represents the cell death in percentage terms determined in the trypan blue exclusion test 24 hours after the treatments (control: no treatment; PEI, pIC and BO-110). Data are presented as mean values ± SEM from three independent experiments conducted with the indicated cell lines (SK-Mel-103 or foreskin fibroblasts). As in panel A, only melanoma cell populations treated with BO-110 die efficiently.

Figure 5. MDA-5 is the sensor and driving force of the cytotoxicity of BO-110 in melanoma cells.

Panel A shows immunoblots of total cell extracts isolated from SK-Mel-28 (top photo), SK-Mel-147 (bottom), untreated (NT) or processed with PEI, pIC, BO-110 or bortezomib (Bor), as indicated for different lines. The asterisk corresponds to a non-specific strip. The arrow indicates the place where the 30 kDa band should be observed, indicating the cleavage of MDA-5.

Panel B: Processing of MDA-5 in SK-Mel-147 cells not infected or infected with lentivirus expressing randomized or corresponding MDA-5 kshRNA visualized by immunoblotting of cell extracts after treatment with pIC, BO-110 or solvent, as indicated. As shown in panel A, an asterisk indicates a non-specific band and the arrows indicate the position of the 30 kDa band, indicative of MDA-5 cleavage.

Panel C shows the inhibitory effect of MDA-5 cshRNA on target toxicity induced by BO-110, as determined in the test with the exception of trypan blue 24 hours after treatment with only pIC (gray bars, ▒), complex BO-110 (black bars, █ ) or without treatment of cells (unshaded bars). The results of infection of the control kshRNA ("ksh control") are also presented. Data are presented as mean ± SEM from three independent experiments.

Figure 6. Effect of pharmacological autophagy inhibitors on the cytotoxic activity of BO-110

Panel A shows the effect of 3-methyladenine (3-MA) and chloroquine (Chlor) on the movement of EGFP-LC3 into autophagosomes, estimated based on the percentage of SK-Mel-103 cells having foci of fluorescence due to GFP-LC3 after 12 hours after processing BO-110 (black hatched bars) or buffer control (unshaded bars).

Panel B shows the inhibitory effect of chloroquine (Chlor), pepstatin A (PEP), or E64d on cell death, assessed by exclusion of trypan blue, 20 hours after treatment with solvent (white bars) or BO-110 (black bars). Data are presented as mean ± SEM from three independent experiments.

Panel C shows fluorescence confocal micrographs of SK-Mel-103 cells transfected with Cherry-GFP-LC3 to detect the formation of autophagosomes (35 red and green lesions) and autolysosomes (simply red lesions) after treatment with BO-110 or 25 nM rapamycin.

Panel D shows the inhibitory effect of 100 μM 100 bafilomycin (Bafil), 20 μM chloroquine (Chlor) or 10 μg / ml pepstatin (PEP) on cell death, assessed by exclusion of trypan blue, 20 hours after solvent treatment (white bars) or BO-110 (black bars). Data are presented as mean ± SEM from three independent experiments.

Panel E shows fluorescence confocal images of SK-Mel-103 cells treated with BO-110 (top photo) or BO-110 in the presence of chloroquine (bottom row of photos) and stably transfected with wild-type EGFP-Rab5 (left column in the figure), or incubated with BO-110 labeled with a red fluorescent agent (middle column photographs). The internalization of BO-110 in melanoma cells can be observed in the absence or presence of chloroquine.

Panel F presents fluorescence confocal images for viewing lysosome-dependent proteolysis due to the cleavage and release of fluorescence DQ-BSA (resulting in green fluorescence in SK-Mel-103 cells treated with both solvent (control) and BO-110) . In the presence of chloroquine (Chlor: middle row of photographs), no fluorescence signal was observed in the DQ-BSA column (middle column). Simultaneous image of cells in the presence of Lysotracker-Red (LTR-Red: left column of photographs) to demonstrate the lysosomal compartment, which creates a signal in all three rows of photographs.

Panel G shows a bar graph representing the joint localization of the corresponding DQ-BSA and Red-Lysotracker signals in the test cells of panel F (control: control, black shaded bars, █; Chl: chloroquine, unshaded bars; BO-110, gray shaded bars , ▒). Joint localization was evaluated in at least 150 cells in two independent experiments and expressed relative to the value obtained in control cells (cu: conventional units of fluorescence).

Panel H shows immunofluorescence confocal images of the foci of fluorescence due to GFP-LC3 (green in the original signal, which indicates the location of autophagosomes) and Lysotracker (red in the original signal, indicating the presence of lysosomes) in SK-Mel-103 cells after BO treatment -110 or buffer control as indicated in the images. The cores were stained with Hoechst dye (top photo, with a blue signal in the original). The bottom row shows the combination of the three previous images, which gives a yellow or orange color in areas that had green and red image signals corresponding to GFP-LC3 and Lysotracker, respectively, which indicates that the signals corresponding to GFP-LC3 and Lysotracker, located in the same areas.

Panel I presents confocal microscope images obtained in real time from SK-Mel-103 cells expressing EGFP-LC3 (green signal in the original), processed either with buffer control with pIC (“control”) or BO-110, and incubated in the presence of a Lysotracker (red signal in the original) or dye hechst (blue signal in the original), as shown in the images. The combination of three images (bottom right for each processing method, control or BO-110) revealed a stable joint localization (yellow signal) of GFP-LC3 and lysosomes (as expected due to the formation of autolysosomes) after processing BO-110, but not after processing " naked "pIC. In the third row of the panel, under the results of overlapping images, diagrams are shown obtained by quantifying the total cellular fluorescence intensity in the green channel in this plane (diagram labeled “GFP” for EGFP-LC3) and red (diagram labeled “LYSO” for Lysotracker). In the case of diagrams obtained with cells treated with BO-110, a similar distribution of both signals, eGFPLC3 and Lysotracker, is indicative of joint localization and, therefore, fusion of autophagosomes and lysosomes.

Panel J shows a population analysis of SK-Mel-103 cells treated with pIC (control) or BO-110, which represents the signal strength of EGFP-LC3 (green fluorescence, X axis) and Lysotracker (red signal, Y axis). Cells with double staining of both markers are enclosed in squares.

Figure 7. Endosomal traffic and the formation and decay of amphisomes during processing of BO-110

Panel A shows successive images of SK-Mel-103 melanoma cells expressing EGFP-Rab7 obtained by real-time fluorescence microscopy at indicated times after treatment with BO-110 or a control solvent. It should be noted that BO-110 caused the appearance of a large number of vesicles. Asterisks indicate fusion of the endosome-endosome (for clarity, only a few examples are shown).

Panel B presents photographs from a confocal microscope of SK-Mel-103 cells stably transfected with retrovirus, which led to the expression of green fluorescence due to the EGFP-Rab7 wt fusion protein (wild-type Rab7) (the first and third column of photographs on the left, in the second case, photographs were obtained in the presence of Lysotracker-Red, as indicated on the column) or the EGFP-Rab7 T22N fusion product by incubation in the presence of Lysotracker (right column of photographs). Cells were further treated with BO-110 (bottom row of the panel) or control solvent (top panel). Images were obtained 10 hours after treatment with BO-110. The two columns of photographs on the right contain values corresponding to the average area contained in Rab7-stained vesicles.

Panel C shows the sequence of confocal micrographs taken at the indicated time intervals (in seconds) shown in the photographs, which illustrate the fusion and incorporation of lysosomes into Rab7-positive vesicles after treatment with BO-110.

Panel D presents real-time images from a fluorescence microscope of SK-Mel-103 cells treated with BO-110 and stably transfected with retrovirus, which causes green or red fluorescence due to GFP-Rab7wt (GFP-Rab7 in photographs), Cherry-LC3 (Ch-LC3 in the photographs), or blue fluorescence due to Lysotracker-Blue (LTR-Blue in the figure). Images were obtained at the indicated time points (minutes) 1 hour after processing. The arrows indicate the first sequence in which each indicated marker could be visualized.

Panel E shows the incorporation of LC3 on the surface of vesicles with endosomal Rab7 prior to its internalization and subsequent degradation. These hybrid endosome / LC3 structures (amphisomes) were visualized by real-time fluorescence microscopy of SK-Mel-103 cells expressing EGFP-Rab7 or Cherry-LC3 (Ch-LC3 in photographs).

Figure 8. The cytotoxicity of BO-110 depends on the activation of effector and regulatory caspases.

Panel A shows the percentage of cell death caused by treating SK-Mel-147 cells with PEI as a buffer control (control, represented by open hatches), pIC (gray hatched columns), or the BO-110 complex (black hatched columns) in the presence of compounds listed under each chart: solvent (control buffer without PEI) (left chart), vitamin E (+ Vite, middle chart) or caspase inhibitor ZVAD-fmk (+ ZVAD, right chart). Percentage was measured in all cases by trypan blue exclusion.

Panel B presents the results of immunoblotting of extracts of metastatic melanoma cell lines (indicated on the left). They were obtained by collecting cells at the indicated time points after treatment (in hours, on each lane). Processing methods are indicated above time points: NT (without treatment: control cell populations incubated in the presence of buffer), PEI, pIC, BO-110 complex or Bor (bortezomib 25 mM). Next to each photo is a protein analysis: casp-9 (caspase 9), casp-8 (caspase 8) or tubulin (load control). The numbers on the right show the relative mass in kDa corresponding to the protein band present at this height.

Panel C presents the results of immunoblotting of SK-Mel-103 cell line extracts obtained by collecting the cells after the treatment indicated in hours (24 and 48 hours). Processing methods are indicated under time points separated by a horizontal line: NT (without treatment: control cell populations incubated in the presence of only buffer), PEI, pIC, and BO-110 complex. The analyzed proteins are shown next to each photograph: casp-9 (caspase 9), casp-8 (caspase 8), casp-7 (caspase 7), Casp-3 (caspase 3), or tubulin (load control).

Figure 9. BO-110 induces apoptosis through NOXA regardless of p53 status and without induction of compensatory activation of MCL-1

Panel A presents photographs of immunoblots of SK-Mel-28 cells (top) or SK-Mel-147 (bottom) obtained by collecting cells at time points indicated after treatment (in hours). Processing methods are indicated above time points: NT (without treatment: control cell populations incubated in the presence of buffer), PEI, pIC, BO-110 complex or Bor (bortezomib 25 mM). Next to each photograph is a protein whose level is analyzed: NOXA, Mcl-1 or tubulin (load control).

Panel B shows two separate graphs representing the relative levels of Mcl-1 (upper) and NOXA (lower graph) calculated by densitometry after immunoblotting performed on SK-Mel-28 cells, depending on the indicated time after treatment, expressed in as a percentage relative to the corresponding value obtained for untreated control cells. Next to each curve is the corresponding processing method.

Panel C presents photographs obtained after immunoblotting of the total SK-Mel-103 cell extracts carried out by collecting the cells after treatment (time is indicated in hours) for each treatment group. Processing methods are indicated above each lane: NT (without treatment: control cell populations incubated in the presence of buffer), PEI, pIC, BO-110 complex or Bor (bortezomib 25 mM). Next to each photograph is a protein whose level is analyzed: NOXA, Bcl-xL, Bcl-2 or tubulin (load control). At the bottom of the panel is the percentage of cell death observed after treatment.

Panel D shows photographs of immunoblots obtained to evaluate NOXA protein expression. Extracts were obtained from SK-Mel-103 melanoma cells treated with pIC or BO-110 for 24 hours two days after infection with a lentiviral vector expressing inactive kshRNA (ksh control) or kshRNA directed against NOXA.

Panel E shows a graph representing the degree of cell death (expressed as the percentage of dead cells) of SK-Mel-103 melanoma cells transduced with either control kshRNA (gray shaded bars, ▒) or kshRNA directed against NOXA (black shaded bars, █ ), and incubated with pIC or BO-110 for 24 hours (NT: without treatment, cells were incubated with a solvent for administration.)

Panel F shows the inhibitory effect of downregulation of MDA-5 on NOXA induction induced by BO-110, represented by NOXA levels (expressed in units of units, cu) in SK-Mel-103 cells transduced against control kshRNA or kshRNA directed against MDA-5. NOXA levels were measured densitometrically and presented relative to the untreated controls (non-infected: no interference, levels that correspond to a value of 1 in the case of naked pIC treatment and 100 in the case of BO-110 treatment).

Figure 10. Anti-melanoma activity of BO-110 in immunocompetent mice

Panel A.

Top panel. Scheme of an experimental approach for obtaining sc xenografts of B16 melanoma cells in syngenic C57BL / 6 mice. Also indicated is the treatment time for peritumoral injections of 50 μg (in 100 μl) (2 ng / kg) “naked” and in combination with PEI pIC. Control groups received 100 μl of 5% glucose or only PEI.

The bottom panel. Tumor growth, estimated by measuring with a caliper at the indicated times, is presented. 10 tumors were analyzed per experimental group. Typically, mice were euthanized when the tumor volume exceeded 1000 mm ³ . The experiment was repeated twice with similar results.

Panel B shows the intravenous implantation of B16-EGFP melanoma cells in C57BL / 6 syngeneic mice for subsequent intravenous treatment with 10 μg (per 100 μl) (1 ng / kg body weight) pIC or BO-110 at indicated times. Control groups received PEI in 5% glucose. 14 days after the introduction of the cells, the mice were euthanized and the lungs were treated to obtain fluorescence images.

Panel C shows a bar graph representing the number of pulmonary metastases observed in each treatment group, obtained by manually counting the external metastases (C). ( P * <0.01 between NT / PEI and BO-110; n = 5; generalized Mann-Whitney test).

Figure 11. IFN-α is induced due to BO-110, but is not sufficient to promote melanoma cell death

Panel A shows B16 melanoma cells and bone marrow macrophages treated with pIC or BO-110; RNA was isolated and quantified by PCR for the IFN IFIT-1 target. Relative IFIT-1 mRNA levels evaluated relative to untreated control cells are shown.

Panel B shows SK-Mel-103 melanoma cells treated with human recombinant IFN-α at indicated concentrations starting from 10 pg / ml, which is already higher than the secreted amount of IFN-α in BO-110 treated cells according to ELISA). Cell death was determined 24 hours after treatment. As a control, cells were simultaneously treated with BO-110 (24 h). Please note that high levels of IFN-α are not cytotoxic for melanoma cells and cannot reproduce effective lysis as a result of exposure to BO-110.

Figure 12. Immunosuppression does not impair the ability of BO-110 to block the spread of melanoma metastases.

Panel A shows the preparation and treatment of B16 melanoma-induced pulmonary metastases in SCID Beige mice (severe immunodeficiency against NK, B, and T cells). Images correspond to photographs under visible or fluorescent light of typical lungs of mice inoculated with iv B16 melanoma cells and treated with PEI, pIC or BO-110. Images were obtained 14 after cell injection.

Panel B presents the average number of metastases caused by B16 as indicated in panel A ( P * <0.01 between treatment groups PEI, pIC and BO-110; n = 5; generalized Mann-Whitney test).

Panel C shows a histological analysis of the lungs after administration of B16 cells in mice treated with PEI, pIC or BO-110. Typical H&E stained sections of the lungs from these treatment groups are shown, visualized at two different magnifications (10x and 40x).

Panel D shows the preparation and treatment of SK-Mel-103 melanoma-induced pulmonary metastases in SCID Beige mice (severe immunodeficiency against NK, B, and T cells). Images correspond to photographs under fluorescent light or visible light stained with H&E (lower line) typical lungs of mice inoculated with SK-Mel-103 melanoma cells and treated with PEI, pIC or BO-110.

Panel E presents the average number of metastases caused by SK-Mel-103, as indicated in panel D. ( P * <0.01; n = 5; generalized Mann-Whitney test).

Figure 13. Inhibition of BO-110 spread of metastases in Tyr mice :: NRAS ^Q61K ×INK4a / ARF ^{- / -}

Panel A presents a Kaplan-Meyer graph for progression-free metastases for Tyr :: Tyr :: Ras ^Q16K × INK4a / ARF ^{- / -} mice that were exposed to DMBA to cause pigmented lesions and then treated with PEI in 5% glucose (control: control), pIC or BO-110.

Panel B presents bar graphs for the cumulative average number of cutaneous melanocytic neoplasms appearing in each of the experimental groups of panel A. Counting was performed every 5 days and the tumors were grouped according to size ranges, as indicated in the diagrams.

Panel C shows typical images of cross sections (left column) and coronary sections (right column) obtained with PET / CT to test metabolic activity (inclusion of 18F-FDG), typical examples of mice that received PEI in 5% glucose (control ), Naked pIC or BO-110. Tumors are surrounded by dashed white lines. Asterisks indicate the position of the animal’s heart.

Panel D shows hematoxylin-eosin staining of samples of melanocytic lesions obtained from each treatment group described in panel A.

Panel E shows hematoxylin-eosin staining of skin, heart, liver, or lung tissue samples (as indicated on the left side of the photograph) in mice treated with 5% glucose diluent (control) or BO-110, which demonstrates the absence of toxicity associated with BO treatment -110, in the compartments of normal cells.

Figure 14. Cytotoxic activity of BO-110 against various tumor cells

Panel A shows the percentage of cell death in various tumor cell lines: pancreatic cancer (Pa), colon (C), bladder (Bl), glioma (G), breast cancer (Br), melanoma (M), cancer prostate gland (Pr), lung (L), and ovary (O), as evaluated in the test with the exception of trypan blue, performed 18 hours (left column) and 30 hours (right column) after exposure to BO-110. Data are presented as mean values ± SEM from three independent experiments conducted on these lines.

Panel B shows typical bright field images of the following tumor cell lines: MiaPaCa (pancreas), BT549 (mammary gland), 639V (bladder) and T98G (glioblastoma) after 24 hours with solvent or BO-110, as indicated. Although the BT549 line was initially resistant to BO-110, it eventually died in long-term viability tests (see panel B). Lines 639V and T98G were very sensitive to the cytotoxic effect of BO-110.

Panel C shows the viability tests of these tumor cell lines after 24 hours of treatment with solvent or BO-110. For a short-term viability test, the treated cells were fixed 24 hours after treatment and stained with crystal violet to visualize the formation of cell colonies. For a long-term viability test, one tenth of the cells treated for 24 hours were seeded, fixed and stained with crystal violet after 48 hours.

Figure 15. Induced BO-110 cell death depends on activation of MDA-5, Noxa, and autophagy in tumor cell lines.

This figure shows the immunoblots of the total cell extracts isolated from the following tumor cell lines: BT549 (mammary gland), 639V (bladder) and T98G (glioblastoma) after 24 hours with solvent treatment (designation “-”) or BO-110 (designation “ + "). Proteins analyzed were: MDA-5 _FL (full-length MDA-5), MDA-5 _C (a cleavage product of MDA-5), NOXA, caspase-9, or tubulin (load control). Note that higher induction of NOXA and MDA-5 _FL , as well as cleavage of caspase-9 and MDA-5 _c, occurs in more susceptible tumor cells to BO-110.

Examples

Analyzes from the examples described below were performed using the following materials and methods:

Cells and cell cultures

The human metastatic melanoma cell lines SK-Mel-19, SK-Mel-28, SK-Mel-103 and SK-Mel-147 and murine B16 cells have been described previously (Soengas et al., 2001). These cells were cultured in Dulbecco's modified Eagle medium (Life Technologies, Rockville, MD, USA) supplemented with 10% fetal calf serum (Nova-Tech Inc., Grand Island, NY, USA).

Human melanocytes were isolated from the foreskin of a newborn as described (Fernandez et al., 2005) and cultured in 254 medium supplemented with melanocyte growth factors (HMG-1) containing 10 ng / ml phorbol-12-myristate-13-acetate ( Cascade Biologies, Portland, OR, USA).

Human fibroblasts were isolated from the foreskin of the newborn and cultured in DMEM supplemented with 10% fetal calf serum.

In addition, cells of other types of tumors were obtained from a panel of 60 human tumor cell lines representing nine types of tumor tissue used in the National Cancer Institute (NCI) anticancer drug screening program. For a pancreatic tumor, the selected cell lines were: IMIMPC2, MiaPaCa, Aspc1, A6L, SKPC-1 and Panc-1; for colon cancer: CACO, SW480 and SW1222; for bladder cancer: RT112, MGHU4, 639V, 253J, MGHu3 and SW1170; for gliomas and glioblastomas: U87MG, U251 and T98G; for breast cancer: MDA-231, MCF7 and T47D; for prostate cancer: LNCaP, PC3 and DU145; for lung cancer: H1299 and NCIH460; and for ovarian cancer: NCI H23 and SK-OV-3.

All cells were cultured in Dulbecco's modified Eagle medium (Life Technologies, Rockville, MD, USA) supplemented with 10% fetal calf serum (Nova-Tech Inc., Grand Island, NY, USA).

Obtaining pIC in complex with PEI (BO-110)

A synthetic dsRNA analog, pIC, was purchased from InvivoGen (San Diego, CA). The reagents jetPEI ™, jetPEI-Fluor ™ and invivo-jetPEI ™ were purchased from Polyplus-transfection (Ikirch, Francia). These products containing a linear derivative of polyethyleneimine were used to complex with a pIC with an N / P ratio (JetPEI nitrogen residues per RNA phosphate) of 1 to 5 in vitro and in vivo , according to the manufacturer's protocol.

Unless otherwise indicated, pIC concentrations used were 1 μg / ml in cultured cells and 1-2 ng / kg body weight in mice.

Drug treatments and cell death analyzes

Bortezomib (Velcade, formerly PS-341) was obtained from Millenium Pharmaceuticals Inc. (Cambridge, MA); adriamycin (doxorubicin) from Sigma Chemical (St. Louis, MO); and etoposide from Bristol-Myers Squibb (New York, NY). The antioxidant tyrone and vitamin E were purchased from Sigma (St. Louis, MO), and the pan-caspase ZVAD inhibitor was purchased from the R&D System (Minneapolis, MN). 3-methyladenine (3-MA) was obtained from Sigma Chemical (St. Louis, MO). Chloroquine was obtained from Sigma Chemical (St Louis, MO, USA).

Cell viability assays in response to drug treatment were performed after plating melanocytes and melanoma cells at least 12 hours before drug treatment. The percentage of cell death at the indicated time points and concentrations used for treatment were evaluated by standard trypan blue exclusion tests as previously described (Wolter et al., 2007; Fernandez et al., 2005).

Cell proliferation assays in response to drug treatment were performed after tumor cells were seeded at least 12 hours before drug treatment. Cell growth at the indicated times and concentrations used for processing were evaluated in a crystal violet stain test.

Protein immunoblotting

To determine changes in protein levels, 2 × 10 ⁶ cells were treated as indicated and harvested at different times after treatment. Total cell lysates were subjected to electrophoresis in 10, 12 or 4-15% gradient gels with SDS under reducing conditions, and then transferred to Immobilon-P membranes (Millipore, Bedford, MA, USA).

Protein bands were detected using the ECL system (GE Healthcare, Buckighamshire, UK).

Primary antibodies included antibodies to: casp-9 and -3 from Novus Biological (Littleton, CO, USA); casp-8 (Ab-3) from Oncogene Research Products (San Diego, CA, USA); casp-7 from Cell Signaling Technology (Beverly, MA, USA); Bcl-xL from BD Transduction Laboratories (Franklin Lakes, NJ, USA); Blc-2 from Dako Diagnostics (Glostrup, Denmark); NOXA from Calbiochem (San Diego, CA, USA); MDAp53 from Novocastra Laboratories (Newcastle, UK); and tubulin (clone AC-74) from Sigma Chemical (St Louis, MO, USA). Antibody to MDA-5 described previously.

The secondary antibodies were anti-mouse or rabbit immunoglobulin antibodies from GE Healthcare. Image J was used to quantify changes in protein levels caused by various processing methods, treating untreated controls as a baseline expression standard.

RNA interference

The ksRNA lentiviral vector used to downregulate NOXA has been reported previously (Fernandez et al., 2005). The lentiviral vector plko used to downregulate MDA-5 (target sequence, SEQ ID NO: 1) was purchased from OpenBiosystems (Huntsville, AL). Randomized oligonucleotides have also been developed to create control kshRNA. Viruses were obtained from 293FT cells as described and used under conditions providing> 80% infection rate (Denoyelle et al., 2006). The downregulation of MDA-5 was confirmed by immunoblotting and RT-PCR (forward primer with SEQ ID No: 2 and reverse primer with SEQ ID No: 3). If indicated, treatment using pIC or BO-110 was started 3 days after infection with the corresponding cshRNA expressing viruses.

Microchips for determining the expression profile

Total RNA was isolated in at least two independent experiments and purified using the RNeasy kit (Quiagen). Samples treated with BO-110 were labeled with 2.5 mg of Cy5-UTP and used as a reference in hybridization reactions with 2.5 g of RNA labeled with Cy3-dUTP resulting from incubation with PIC or PEI. Labeled RNA was hybridized with two-color oligonucleotides from a microchip of the complete human genome (4 × 44K) from Agilent (Santa Clara, CA, USA), following the manufacturer's instructions. After washing, the slides were scanned using Scanarray 105000 XL (GSI Lumonics Kanata, Ontario, Canada) and images were analyzed using GenePix 4.0 software (Axon Instruments Inc., Union City, CA) as previously described (Alonso et al., 2007). When measuring the fluorescence intensity, the background values were automatically subtracted. The Cy3: Cy5 ratios were normalized to the value of the median ratio of all points. After normalization, points with intensity values for both channels (the sum of the averages) lower than the local background were discarded. The relations of the remaining points were subjected to a logarithmic transformation (base 2), and the arrays of repeating points were adjusted to the average. Grouping of non-essential pairs (UPGMA: a group of non-essential pairs) of genes expressed differently in control and test samples was performed using the Gene Expression Pattern Analysis Suite (GEPAS).

In vivo treatment response

Female C57BL / 6 mice were purchased from NIH (Bethesda, MA). Female SCID Beige mice with impaired NK, T, and B lymphocyte function were purchased from Charles Rivers (Wilmington, MA). All animals at the beginning of the experiments were 6-12 weeks old. Animal care was carried out in accordance with the approved methods of the University of Michigan Cancer Center.

Skin grafts were implanted by intradermal injection of 10 ⁵ melanoma B16 cells. 2 ng / kg pIC body weight alone or in combination with invivoJetPEI was administered by peritumoral injection on days 7, 11, 15 and 21 after tumor implantation. Additional treatment groups included the administration of JetPEI as a separate substance and placebo controls. Tumor volume was assessed by caliper measurement and calculated as V = L × W ^2/2 where L and W correspond to the length and width of the tumor, respectively.

Surrogate models of lung metastases were obtained by iv injection of 4 × 10 ⁵ B16-eGFP or 5 × 10 ⁵ SK-Mel-103-eGFP melanoma cells. Treatment was carried out on days 3, 6 and 9 by iv injection of 1 ng / kg pIC body weight alone or in combination with invivoJetPEI. Lungs were removed 14 days after infection, and external metastases were counted manually and classified by number and size. Alternatively, the Illumatool TLS LT-9500 fluorescent light system (Lightools Research, Encinitas, CA, USA) was used and the fluorescence emanating from the tumor cells was recorded on a Hamamatsu Orca 100 CCD camera. Metastatic lesion was monitored independently by analysis of hematoxylin-eosin stained paraffin sections. The experiments were carried out on groups of five mice and were repeated two to four times. Mice were euthanized when animals in the control groups showed signs of discomfort and breathing problems.

Autochthonous melanomas were obtained by crossing Tyr :: N-RasQ61K mice with C57BL / 6 genetic knockout mice Ink4a / Arf (Ackermann et al., 2005). To induce melanoma in the skin, 220 mg of 7,12-dimethylbenz [a] anthracene (DMBA) was applied to mice once per day at the age of 8-10 weeks. After the development of early melanocytic neoplasms (foci having at least 1 mm in diameter), mice were injected twice weekly with intraperitoneal injections of 1 ng / kg pIC body weight as a separate substance or in combination with invivoJetPEI. Melanoma and moles (nevi) were counted, and its dimensions were measured in two diameters using a caliper, expressing the average tumor volume in mm ³ .

Tumor size was also evaluated using PET-CT (Positron Emission Tomography - Computed Tomography). Inspection and acquisition of PET-CT images was performed using the PET-CT system for small animals View Explore General Electrics (Fairfield, CT, USA). 15 mBq of 18F-FDG (2-fluoro-2-deoxy-D-glucose) was injected to visualize and obtain PET images and reconstructed using the 3DOSEM algorithm. CT images were acquired in 16 frames with energies of 35 keV and 200 μA, and images were reconstructed using the FDK algorithm. Melanomas, metastases, and other organs were independently checked by analyzing paraffin sections stained with hematoxylin-eosin.

Transmission Electron Microscopy

For transmission electron microscopy (TEM), these cell populations were washed with 0.1 M Sorensen buffer, pH 7.5, and fixed in 2.5% glutaraldehyde for 1.5 hours, then dehydrated and placed in Spurra resin. Then the block was cut into 60-100 nm ultrathin sections and mounted on copper grids. For routine analysis, ultra-thin sections were stained with 2% uranyl acetate and lead citrate. Electron micrographs were obtained on a Philips CM-100 transmission electron microscope (FEI, Hillsbrough, OR) using a Kodak 1.6 Megaplus digital camera.

Confocal and fluorescence microscopy: quantification of spot spots GFP-LC3

An eGFP-LC3 fusion product cloned into a pCNA expression vector was provided by Gabriel Nunez (University of Michigan Cancer Center). eGFP-LC3 and fragments of eGFP-Rab7wt, eGFP-Rab7T22N, eGFP-Rab5wt, eGFP-Cherry-LC3 and Cherry-LC3 were cloned into the pLVO-puro lentiviral vector for stable gene transfer. Cells originating from melanoma (i.e., SK-Mel-103) were infected with pLVO-eGFP-LC3 and selected with puromycin. GFP-LC3-associated fluorescence emission was visualized using a Leica AF6000 fluorescence microscope and images were analyzed using LAS AF V1.9 (Leica, Solms, Germany). For real-time confocal microscopy, the inventors used a Leica TCS-SP2-AOBS-UV ultra-spectral microscope, combined with an incubation chamber with a controlled CO ₂ content and temperature. Images were analyzed using LCS (Leica, Solms, Germany). For Lysotracker ™ co-localization experiments, red or blue (Invitrogen, Carlsbrad, CA) at a concentration of 50 nM or 200 nM and Hoechst 33342 dye (Invitrogen, Carlsbrad, CA) were added 10 minutes before imaging at a concentration of 5 μg / ml. Co-localization images were analyzed using LAS AF V1.9 (Leica, Solms, Germany).

Cytokine expression

Human interferon alpha was measured in culture supernatants by enzyme-linked immunosorbent assay (ELISA). The human IFN-α ELISA kit and recombinant hIFN-α were purchased from PBL Interferon Source (Piscataway, NY) and used in accordance with the manufacturer's protocol. The expression level of IFN-α was measured for macrophages from bone marrow (BMDM) and melanoma B16 cells by real-time PCR. BMDM cells were obtained, seeded and processed as previously described (Celada et al., 1984). Real-time quantitative PCR analysis of IFIT-1 RNA transcripts was performed using TaqMan primer and probes obtained from Applied Biosystems on an Applied Biosystems 7700 sequence detector after normalization with β-actin.

Statistical analysis

Viability data were expressed as mean values ± SEM, and statistical analysis of differences was performed using the two-tailed student criterion. A value of P <0.05 was considered significant. For a statistical assessment of tumor growth and in vivo metastases, the generalized Mann-Whitney-Wilcoxon test was used when comparing the values of continuous variables between two groups. Values of P <0.05 were considered significant.

Example 1. Identification of autophagy inducers in melanoma cells using discriminatory analysis based on LC3 fluorescence. Verification of autophagic cell death induced by BO-110 by ultrastructural analysis

As discussed previously, a hallmark of autophagy (both for the induction of cell death and for survival) is the movement of gene 8 of the autophagy protein (ATG8) / LC3 from the cytosol to newly formed autophagosomes. Based on this observation, changes in the cell distribution of the GFP-LC3 fusion protein (i.e., from the diffuse pattern to focal staining) are used as a marker for the early stages of autophagy. The presence of autophagosomes can also be confirmed by electron microscopy or light microscopy.

1.1. LC3 Based Fluorescence Analysis

To study the role of autophagy in drug response of melanoma, a GFP-LC3 based discriminatory assay was used to screen for commercially available chemotherapeutic agents and immunomodulators. Melanoma cells were stably transfected with lentiviral vectors expressing derivatives of the autophagosomal marker LC3 with GFP, such as pLVO-eGFP-LC3.

The human SK-Mel-103 cell line was chosen as a model system for primary screening based on its highly metastatic and chemoresistant phenotype (Soengas et al., 2001). Subsequent validation studies were performed on a panel of human cell lines of various genetic backgrounds (see below).

It has been found that various anti-cancer drugs induce focal GFP-LC3 fluorescence emission without a significant effect on cell viability. However, it was found that, among cell death inducers, the complex of PEI and the polyinosine-polycytidylic acid dsRNA mimetic (BO-110) is particularly effective for the formation of GFP-LC3 foci. In approximately 50% of the cells, a noticeable spot staining of GFP-LC3 was observed during 4-6 hours of incubation at low doses (0.5-1 μg / ml) of BO-110 (see representative micrographs and quantitative data in Figure 1A and Figure 1B ) . Indeed, kinetic analysis indicated faster formation of GFP-LC3 foci when using BO-110 than using rapamycin ( Figure 1B ), the classic positive control of autophagy induction (Klionsky et al., 2008).

Interestingly, at later times, BO-110 treatment was able to cause cell death, even in melanoma cell lines, which are inherently resistant to standard DNA damaging agents such as doxorubicin or etoposide, as is the case with SK-Mel-103 .

Analysis of endogenous LC3 showed changes in electrophoretic mobility ( Figure 1C ), corresponding to the characteristic lipidation of this protein during the formation of autophagosomes, and the need for ATG5 to generate focal formation of GFL-LC3.

The inventors are not aware of any previous reports that would associate pIC with autophagy in cancer cells. As a result, the following tests have focused on this compound, as it may reveal new elements that better understand potential intracellular dsRNA sensors for inducing autophagy and death of tumor cells.

1.2. Ultrastructural analysis of the effect of BO-110 on cells

Focal staining of GFP-LC3 in cells treated with BO-110, described in the previous paragraph, is consistent with the formation of autophagosomes. However, in order to rule out possible nonspecific aggregates of ectopically expressed GFP-LC3 (Klionsky et al., 2008), the response was independently analyzed by electron microscopy ( Figure 1D ).

Early responses (5 h) to BO-110 included a marked accumulation of membrane-bound electron-dense structures that isolate cell breakdown products ( Figure 1D ), a distinctive feature of autophagosomes. The size and number of these structures could be seen as intracellular granules even under an optical microscope ( Figure 1E ).

At later times, BO-110 treatment caused cell death ( Figure 1F , middle panel), which, as detected by electron microscopy, was associated with the formation of large phagocytic vacuoles with a diameter of more than 500 nm ( Figure 1F, right panel).

Figure 2, using selected fluorescence photographs taken at various points in time, presents the generalized results of the development of autophagy induction over time, as can be seen from the micrographs of the redistribution of EGFP over time: from diffuse staining to focal aggregates indicating the formation of autophagosomes, a process that culminates in the complete destruction of cells. Diffuse staining was observed in control cells throughout the entire testing period, which is a sign of basal levels of LC3 accumulation.

Interestingly, the integrity of the plasma and nuclear membranes was maintained, and cells treated with BO-110 showed chromatin condensation characteristic of apoptosis programs. Autophagy induction was BO-110 dependent, since PEI (control) had minimal effect on the number or size of autophagosomes ( Figure 1D-F ). Together, these data confirm the cytotoxic effect of BO-110, suggesting the formation of autophagosomes, followed by cell death with characteristics characteristic of apoptosis.

Example 2. Sensitivity analysis of various human and mouse melanoma cell lines using pIC conjugated to cationic molecules and selectivity for melanocytes

An additional set of cell lines was tested to determine whether the pro-autophagic activity of BO-110, detected in an initial study conducted with SK-Mel-103 cells, was a reflection of broader anti-melanoma activity.

Melanoma cells were selected based on the principle of correlation with frequent events associated with melanoma, such as mutations in BRAF or NRAS , deletion of INK4a / ARF or PTEN loci, or increased regulation of various antiapoptotic members of the Bcl-2 family, which are known to contribute to progression and chemoresistance melanomas.

In melanoma, P53 mutations are rare (Soengas and Lowe, 2003). However, since p53 can play a key role in activating apoptosis and autophagy programs, the inventors also tested on the SK-Mel-28 cell line, which expresses mutant p53, to determine whether this tumor suppressor is essential for the anti-melanoma activity of BO-110 . In addition, murine metastatic melanoma B16 cell lines were also included in the analysis as an example of a model widely used in melanoma immunotherapy (Wenzel et al., 2008) to evaluate differences in treatment response, presumably related to differences between species. In parallel with this, the inventors also analyzed melanocytes isolated from the human foreskin.

Table 1 shows the genetic background of cell lines of human metastatic melanoma:

The mutation status of p53 was determined by direct sequencing of exons 2-10 by RT-PCR.

Samples with P72R polymorphism are designated as R. P53 inducibility was determined by immunoblotting of extracts treated with doxorubicin (0.5 mg / ml, 12 h). Lines with high endogenous p53 levels are marked with an asterisk. The levels of PTEN, Apaf-1, Casp-8, Bcl-2, Bcl-xL and Mcl-1 were determined by immunoblotting and normalized to control melanocytes. The mutational status of BRAF and NRAS was determined by direct sequencing of PCR amplified genomic fragments of exons 15 and 3, respectively. Responses to doxorubicin (DOX; 0.5 μg / ml, 30 h) were assigned to the categories ++, +, - / +, -, corresponding to the percentage of cell death equal to 100-70, 70-50, 50-30 and < 30%, respectively. Responses to BO-110 (1 μg / ml, 30 h) were assigned to the categories +++, ++, +, corresponding to the percentage of cell death equal to 100-90, 90-60 and 60-40%, respectively. Lines with high endogenous p53 levels are marked with an asterisk.

With these cell lines, tests similar to those described in Example 1 for the SK-Mel-103 cell line were performed to test the sensitivity of each of these cell lines to pIC and PEI as separate substances, or used in combination. The results are shown in figure 3 .

Five tested melanoma cell lines (SK-Mel-19, -28, -103, -147 and B16) were destroyed with similar kinetics and efficacy after treatment with BO-110 ( Figure 3A ). It is important to note that in all tested melanoma cell lines, early autophagy activation with BO-110 was inevitably accompanied by cell death.

Electron microscopy clearly showed autophagosomes also in the case of the line with the mutant p53, SK-Mel-28 ( Figure 3B ). As shown in FIG. 3A and the dose response curves shown in FIG. 4B (which represent representative data for lines corresponding to four independent isolates), it is important to note that under conditions that led to the death of 70% of SK-Mel-103 melanoma cells through 24 hours after treatment, normal melanocytes remained viable and showed no signs of autophagy.

Moreover, as shown in FIG. 4A , no significant changes were observed in the morphology, granulation, or cytosolic distribution of GFP-LC3 in melanocytes over a wide range of BO-110 concentrations. Figure 4C also demonstrates that normal human skin fibroblasts were also more resistant to BO-110 than melanoma cells.

Interestingly, PEI is crucial in the selectivity for tumor cells. So, anti-melanoma pIC activity decreased by 70-80% if the treatment was carried out without PEI ( Figure 3A ). In the absence of PEI, the naked pIC was almost as ineffective in melanoma cells as in melanocytes, and showed minimal activity as an autophagy inducer ( Figure 3B ).

PEI is a classic solvent in gene therapy because of its ability to stimulate the uptake of DNA and RNA molecules through endocytosis (reviewed in Payne, 2007). The multilayer structures found in melanoma cells treated with BO-110 (see photos below in Figure 2B ) are indeed consistent with multiple fusion events involving the entry of endosome-lysosome hybrids (amphisomes) into autophagosomes (Maiuri et al., 2007).

Example 3. Quantitatively different activation of MDA-5 under the influence of pIC in the absence and presence of PEI

In addition to promoting the endocytosis of DNA or RNA molecules, PEI can promote endosome swelling and enable efficient delivery of genetic material to the cytosol (review by Payne, 2007). As a consequence, PEI may facilitate access of pIC to intracytosolic sensors. Gene-5 associated with differentiation of melanoma (MDA-5) is one of these sensors (Akira, 2006), and as a result, the inventors tested whether this protein is the driving force behind BO-110 mediated killing of melanoma cells.

MDA-5 activation was analyzed by monitoring proteolytic cleavage, which leads to the separation of its helicase and caspase activation recruitment domains (CARD) during cell death, as described (Kovasovics et al., 2002; Barral et al., 2007).

This analysis was carried out by immunoblotting of extracts from SK-Mel-28 and SK-Mel-147 cells treated with PEI, pIC or BO-110, after electrophoresis of these extracts. Bortezomib was used as a positive control for effective induction of cell death. The results are presented in figure 5 .

Interestingly, protein immunoblotting demonstrated the strong and robust ability of the PEI-pIC complex, BO-110, to induce MDA-5 processing ( Figure 3A ). The naked pIC was able to induce such processing, albeit at a much lower level and not in all test cells (none of the paths corresponding to SK-Mel-28 cells contained a 30 kDa band characteristic of proteolytic cleavage, which should appear at the height indicated by the arrow in panels A and B of figure 5 ) or not constantly (see figure 5A , in which the intensity of the 30 kDa band decreases with time when processing pIC in SK-Mel-147 cells).

To determine the contribution of MDA-5 to the cytotoxic activity of BO-110, short hairpin RNAs (kshRNAs) complementary to MDA-5 were transduced into melanoma cells using lentiviral vectors for stable knockdown of MDA-5 (see protein immunoblots in Figure 3B ). kshRNA to MDA-5 significantly reduced BO-110-induced death of melanoma cells without any detectable nonspecific effect on control cells ( Figure 3C , p <0.05).

It is important to note that the induction and processing of MDA-5 was not just a consequence of the activation of the death mechanism in melanoma cells. Treatment with bortezomib, a proteasome inhibitor capable of activating both the natural and death-mediated apoptotic pathways in melanoma, did not affect the levels or processing of MDA-5 ( Figure 5A ). These results illustrate the main differences in the mechanisms of the implementation of cell death programs in the case of BO-110 and other proapoptotic inducers.

Example 4. Pharmacological autophagy inhibitors inhibit the cytotoxic activity of BO-110

The following experiments focused on the mechanisms underlying the activation of BO-110 induced cell death. 3-methyladenine (3-MA) and chloroquine are often used to independently test autophagy mechanisms because of their ability to inhibit the formation of autophagosomes or autolysosomal activity, respectively (Maiuri et al., 2007; Klionsky et al. , 2008).

To check whether this obstruction existed in melanoma cells treated with BO-110, SK-Mel-103 melanoma cells were treated with 3-methyladenine or chloroquine 12 hours after treatment with BO-110 or buffer control (solvent). The results are presented in figure 6 .

As shown by fluorescence microscopy ( Figure 6A and Figure 6B ), 3-MA blocked the formation of foci of GFP-LC3 caused by BO-110. Autophagosomes accumulated in the presence of chloroquine, but, interestingly, this induction was not productive as an inducer of cell death ( Figure 6C , it was found that the percentage of dead cells observed in the presence of chloroquine was lower than that observed with pepstatin A, E64d, or a combination of both). Thus, these results confirm the scenario in which the cytotoxic activity of BO-110 is not a passive by-product of the formation of autophagosomes: the lytic activity of lysosomes is an important mediator of the destruction of BO-110 melanoma cells.

Example 5. BO-110 drives the fusion of autophagosomes / lysosomes for the subsequent launch of destruction programs

If autolysosomes are a key factor in BO-110, blocking lysosomal hydrolases should protect melanoma cells during BO-110 treatment. It is not possible to block all lysosome-dependent activity, since many enzymes with overlapping targets can be localized in this organelle (Fehrenbacher and Jaattella, 2005). However, useful information on lysosomal activity can be obtained thanks to a wide range of E64d protease inhibitors and pepstatin A, since these compounds effectively block various cathepsins (B, D, and L) in autolysosomes (Klionsky et al., 2008). As a result, a test was conducted to compare the effect of chloroquine, pepstatin A, or E64d on cell death 20 hours after treatment with buffer control or BO-110. The results are presented in figure 6C , mentioned earlier. It is noteworthy that pepstatin and E64d reduced the degree of cell death by 50% due to BO-110.

To confirm that the vesicles corresponded to the previously identified large multivesicular structures, which included large endosomes, which in turn recruited multiple autophagosomes (to create hybrid structures known as amphisomes), and these vesicles did not come from inactive autophagosomes in which lysosomes were either not recruited or dysfunctional, or autophagosomes were the result of accumulation of defective decomposing material, melanoma cells were transfected with food by the merger of GFP and Cherry-LC3. Cherry-GFP-LC3 signals led to red and green fluorescence in autophagosomes due to two fluorescent proteins (Cherry and GFP), however, they lost the GFP signal (green) in the acidic environment of autolysosomes.

Using this strategy, it was established that, in fact, BO-110, like rapamycin, induced the formation of autolysosomes in melanoma cells, as indicated by the presence of only red LC3 foci in Figure 6C. Chloroquine, in accordance with previous tests, blocked the death of melanoma cells caused by BO-110 ( Figure 6D ) without affecting the endosomal uptake of this dsRNA mimic (as determined by the co-localization of labeled Fluo Red BO-110 and GFP fused early endosomal protein Rab5 ( figure 6E). Similar inhibitory effects were observed when using the broad-spectrum protease inhibitors E64d and pepstatin a, and vacuolar ATPase blocker bafilomitsina (6D figure), indicating the benefit of the new mechanism of action BO-110, dependent on lysosomes.

To independently control lysosomal activity during BO-110 treatment, cells were tested for their ability to process DQ-BSA (a derivative of BSA whose green fluorescence is quenched if it is not cleaved by proteolytic enzymes). As shown in FIG. 6F , DQ-BSA was efficiently cleaved in the presence of BO-110. Note that DQ-BSA emission was detected in lysosomes, as indicated by co-localization with Lysotracker red, a dye whose pH penetration into the cell and which emits red fluorescence when functional lysosomes enter the acidic environment). The result is very different from the minimal DQ-BSA-induced fluorescence emission observed in SK-Mel-103 cells treated with BO-110 when lysosomal activity is blocked by chloroquine ( Figure 6F and Figure 6G ).

To further characterize the ability of BO-110 to trigger the onset and complete development of the autophagy process, the fusion of autophagosomes and lysosomes was visualized using confocal microscopy. For this, SK-Mel-103 cells stably expressing GFP-LC3 were treated with BO-110 or the appropriate buffer as a control and incubated in the presence of Lysotracker-Red. Analysis of the double emission of green and red fluorescence (for the fused product GFP-LC3 and Lysotracker, respectively), based on individual cells and cell populations showed a clear joint localization of autophagosomes and lysosomes (see representative microphotographs of fluorescence in figures 6H and 6I , as well as corresponding quantitative data in figure 6J ). It is important that this joint localization was an early event in the responses induced by BO-110 (already detectable 4-8 hours after treatment) and preceded organized cell death.

Having determined that autophagosomes fuse with active lysosomes in response to BO-110, the inventors evaluated whether these organelles interacted with endosomes and whether they were rectified into endosomes. First, the endosomal dynamics in melanoma cells expressing GFP fused to the late endosomal marker Rab7 was evaluated (Luzio et al., 2007). The basal creation and destruction of endosomes (i.e., a gradual reduction in size) was determined in untreated melanoma cells (left panel in Figure 7A ). However, treatment with BO-110 markedly enhanced endosomal activity, inducing stable and multi-wave formation of endosomes (middle and right panels of Figure 7A ). These endosomes were filled with lysosomes, which was determined by double visualization of GFP-Rab7 and Lysotracker red ( Figure 7B ). In addition, slow-motion microscopy revealed the rapid kinetics of multiple recruitment of lysosomes to GFP-Rab7-labeled endosomes, which is also demonstrated in the sequential series of fusion events in Figure 7C . It is important to note that, as shown in Figure 7B (right panels), fusion of endosome-lysosomes was significantly inhibited if cells overexpressed Rab7-T22N, a known dominantly negative mutant of this protein. In general, these results demonstrate the dynamic mobilization of endo / lysosomal compartments in tumor cells treated with BO-110.

Example 6. BO-110 binds autophagy with apoptotic caspases

Lysosomal proteases can affect cell death programs at various levels (Maiuri et al., 2007; Hoyer-Hansen and Jaatella, 2008). In the case of mitochondria, they may upregulate the production of reactive oxygen species (ROS) and / or involve classic apoptotic caspases (regulatory casp-9 and effector casp-3 and -7). Casp-8 dependent external pathways may also respond to lysosomal activation (Fehrenbacher and Jaattella, 2005). To determine the participation of ROS in the mechanism of action of BO-110, the treatment was performed in the presence of vitamin E, trolox or thyrone, radical scavengers with different antioxidant activity, and a pan-caspase inhibitor z-VAD-fmk. Data analysis of cell death results in each of these cases is shown in FIG. 8A , with data obtained with vitamin E as a representative case of results obtained with the mentioned chemical antioxidants at doses of these reagents that block ROS-controlled apoptosis in melanoma (Fernandez, 2006). As shown in the figure, in the presence of these antioxidants, no significant effect on cell death due to BO-110 was found. In contrast, the pan-caspase z-VAD-fmk inhibitor inhibited the fatal effects of BO-110 by 70%. Together, these results confirm a caspase-dependent mechanism activated after the autophagy program.

Caspase processing, in fact, was effectively stimulated by BO-110, which was determined by immunoblotting of cell extracts collected at different time periods either after no treatment (NT), except for buffer control without PEI, or after treatment with PEI, pIC, complex BO-110 or a known caspase cleavage inducer, bortezomib ( Figure 8B and Figure 8C ). Figure 8B compares the ability of PEI, pIC, and BO-110 to induce apoptotic processing of caspases 8 and 9 in different lines of metastatic melanoma: effective activation of caspases 9 and 8 was clearly visible 20 hours after treatment with BO-110 in all tested human melanoma cell lines ; similar to what was observed when using bortezomib. Moreover, the kinetics and degree of caspase processing using BO-110 were highly consistent ( Figure 5B ), regardless of the mutational status of BRAF (e.g. SK-Mel-19), NRAS (SK-Mel-103, -147), or p53 (SK-Mel-28).

Figure 8C presents the results of a similar test performed with the SK-Mel-103 line, in which a more complete analysis was performed, including, in addition to apoptotic caspases 9 and 8, effector caspases 3 and 7. A similar test was performed with the SK-Mel line -147, which gave similar results. Effective processing of casp-9, -3 and -7 in SK-Mel-103 and -147 cells was especially important. These lines have low Apaf-1 levels and are very ineffective in involving casp-9 / Apaf-1 apoptosomes (Fernandez et al., 2005; Soengas et al., 2006) in response to classic anticancer drugs such as doxorubicin, etoposide, and cisplatin (see graph in Figure 1C ). Thus, these results indicate the superior ability of BO-110 to activate apoptotic programs and bypass the internal mechanisms of resistance to standard chemotherapeutic drugs.

Example 7. Activation of cell death by BO-110 in the absence of compensatory effects on antiapoptotic members of the Bcl-2 family

For a more detailed analysis of the mechanism of action of BO-110 and to determine events that can be uniquely activated by this agent, the response to the drug was compared with the effects of bortezomib. This substance was chosen because it is also a strong activator of the apoptosis mechanism in melanoma cells (Wolter et al., 2007; Fernandez et al., 2006).

However, the inventors expected that bortezomib and BO-110 would differ in their mechanism of action. The targets of bortezomib are proteosomes, not lysosomes (Qin et al., 2005). In addition, as shown in FIG. 5A , bortezomib kills melanoma cells without inducing or processing MDA-5. Bortezomib is also interesting because it can promote massive accumulation of proapoptotic NOXA, but it also induces fast and strong upregulation of its anti-apoptotic factor antagonist MCL-1 (Fernandez et al., 2005), a member of the Bcl-2 family. It is important to note that MCL-1 acts as an internal compensatory mechanism of proteasome inhibition and blocks the antitumor effect of bortezomib in vitro and in vivo (Wolter et al., 2007; Qin et al., 2006).

To assess the similarities and differences between bortezomib and pIC (naked or complexed with PEI), melanoma cells were incubated with each of these compounds, and extracts were collected at various points in time after treatment to evaluate levels of NOXA, MCL-1, and other family members Bcl-2 (Bcl-xL or Bcl-2). The results are presented in figure 9 .

As shown in Panels A and B of Figure 9 , the naked pIC was not able to induce NOXA systematically or continuously in SK-Mel-28 or SK-Mel-147 melanoma cells (cells expressing p53 with a wild-type L145R or p53 mutation, respectively ) On the other hand, BO-110 induced NOXA at levels 35, 10 and 5 times higher than the basal levels in SK-Mel-28, SK-Mel-147 and SK-Mel-103 cells, respectively (see immunoblots in panels A and C of figure 9 , and representative quantitative data in figure 9B from the results obtained in SK-Mel-28 cells), which once again emphasizes the different activity of naked and in combination with PEI pIC.

Regarding NOXA inhibitory regulators, MCL-1 levels were minimally induced by BO-110 ( Figure 9A and the first graph from Figure 9B ). This differs from the situation with bortezomib, which strongly activates NOXA, but induces the simultaneous accumulation of MCL-1, as described earlier (Fernandez et al., 2005). Other anti-apoptotic members of the Bcl-2 family, such as Bcl-2 and Bcl-xL, were also not affected by BO-110 (see immunoblots for SK-Mel-103 in Figure 9C ).

In the absence of compensatory mechanisms, relatively lower levels of NOXA induced by BO-110 may be sufficient to promote cell death. To test this hypothesis, melanoma cells transduced cshRNA, which, as shown earlier, specifically inhibits mRNA and NOXA protein (Fernandez et al., 2005), and cells infected with a lentiviral vector expressing inactive control cshRNA were used as a control. As shown in FIG. 9D, a 50% decrease in NOXA protein expression due to kshRNA inhibited the upregulation of NOXA induced by BO-110, also by about 50%, and inhibited the toxicity of BO-110 ( FIG. 9E ).

Further, the inventors used anti-MDA-5 kshRNA to determine the need for a given protein to regulate NOXA via BO-110 and performed the same test as before, but quantifying NOXA levels. The results are shown in the graph of figure 9F . Interestingly, the kshRNA for MDA-5 inhibited NOXA protein levels by 70% ( Figure 9F ) without side effects on other members of the Bcl-2 family.

Together, these results revealed a new aspect of the action of MDA-5 in the apoptosis mechanism due to the induction of NOXA.

Example 8. Different efficacy of naked pIC and BO-110 in immunocompetent mice

Then the anti-melanoma activity of pIC and BO-110 was evaluated in vivo. In melanoma models, naked pIC should be administered either in high doses or in combination with other substances (for example, protein synthesis inhibitors) to effectively activate innate immunity programs. Data from previous experiments indicated that pIC would be much more effective in the presence of PEI. The treatment effect was first analyzed against an immunocompetent genetic background. Murine melanoma B16 cells, either non-transduced or transduced with GFP (to facilitate detection of fluorescence), were implanted in syngeneic normal mice. Two strategies were used: injection of tumor cells (i) subcutaneously (s / c) or (ii) intravenously to evaluate tumor progression in localized areas or as distant metastases, respectively. Mice received PEI, pIC or BO-110, or 100 μl of 5% glucose (NT group).

Figure 10A summarizes a strategy for experiments with sub xenografts leading to the formation of B16, as well as the dosage and treatment regimen. During treatment with peritumoral injections, 2 ng / kg of body weight of “naked” pIC or pIC was administered in combination with PEI.

It is noteworthy that BO-110, as it turned out, was superior to pIC in all studied cases. So, mice with subcutaneous growing B16 melanomas that received the solvent, either PEI or pIC alone, had to be euthanized 15-25 days after implantation due to excessive tumor growth ( Figure 10A ). Under the same conditions, subcutaneous melanomas in the BO-110 treatment group were either not detected, or were significantly smaller ( Figure 10A ).

Figure 10B summarizes a strategy for experiments with intravenous implantation of B16-eGFP melanoma cells and subsequent treatment with pIC, PEI, BO-110, or 5% glucose (NT group). This figure also shows fluorescence images of the lungs of euthanized animals ( Figure 10B ) and quantitative data on pulmonary metastases ( Figure 10C ). In this experiment, BO-110 was 5 times more effective than the “naked” pIC also in surrogate models of pulmonary metastases of melanoma, as determined by visualization of fluorescence.

Example 9. IFN does not reproduce causing cell death features of BO-110

pIC is the classic inducer of IFN-driven cellular immunity (Wenzel et al., 2008). However, the data presented in the previous examples indicate that pIC, being in complex with PEI, can also act autonomously in the cell, which may differ from IFN-mediated responses in “professional” immune cells. To assess this possibility, B16 melanoma cells and macrophages were tested for their ability to secrete and respond to IFN-α. RT-PCR results showed that both types of cells activated classic IFN-α targets, such as IFIT-1 (IFN-inducible protein with tetratricopeptide repeats), after treatment with BO-110 ( Figure 11A ). PEI was optional for pIC-mediated induction of IFN targets in macrophages ( Figure 11A ). This was expected because these cells can efficiently perceive viral dsRNA. However, melanoma cells were unable to induce IFIT-1 using only naked pIC ( Figure 11A ).

To directly evaluate IFN-α production by melanoma cells, an Elispot assay was performed using recombinant human IFN-α as a reference. The levels of IFN-α secreted by melanoma cells after treatment with BO-110 were lower than 10 pg / ml. To determine whether IFN-α can replace BO-110 (that is, whether IFN-α secretion is the main inducer of melanoma cell death), increasing amounts of this cytokine were added to melanoma cells. Interestingly, high doses of IFN-α (10 higher than the levels secreted after treatment with BO-110) were not able to affect the viability of melanoma cells ( Figure 11B ).

It is also interesting to note that tests with microarrays showed that in the response to pIC, in addition to being very transient, all the genes involved were expected to respond to interferon, as shown in Figure 11C ). In contrast, the effect of BO-110, in addition to its duration, extended to additional transcripts.

Thus, these results demonstrate internal differences in the recognition and perception of dsRNA mimics in macrophages and melanoma cells.

Example 10. BO-110 can inhibit metastatic growth amid highly weakened immunity.

Since melanoma cells are often immunoresistant, it has been tested whether the direct toxicity of BO-110 to melanoma cells is also effective against a background of highly weakened immunity. The most common effector mechanisms associated with immunotolerance of melanoma are signaling defects in NK, T, and B cells (Kirkwood et al., 2008). As a result, the ability of pIC (as a single substance or in combination with PEI) to block the growth of melanoma was tested on SCID Beige mice with impaired function of NK, T and B lymphocytes.

To monitor the effectiveness of treatment in the control with pulmonary metastases, melanoma cells were labeled with GFP and administered by intravenous injection following the treatment regimen described in FIG. 10 . Melanoma B16 ( Figure 12 of Panel A, B and C ) and SK-Mel-103 ( Figure 12 , Panel D and E ) were analyzed as representative examples of murine and human melanoma, respectively.

In both cell models, BO-110 showed the ability to inhibit the growth of melanoma in the lungs. Figure 12A shows a striking difference in the number of B16 metastases visible on the surface of the lungs after treatment with BO-110 (see quantitative data in Figure 12B ). Histological analysis also confirmed the reduced number and size of B16-induced pulmonary nodules in the BO-110 group ( Figure 12C ). Similar analyzes showed a clear antitumor effect of BO-110 (but not “naked” pIC) in the control of disseminated growth of SK-Mel-103 ( figure 12 of panel D and E ). Thus, the data of the inventors confirm a new mechanism of action of a dsRNA mimic that induces strong anti-melanoma activity in vivo in SCID beige mice that lack a competent immune system (Cray and Chapeau, 1990).

Example 11. Inhibition of BO-110 metastatic growth in animal models of human melanoma

To compare the differences between pIC and BO-110, a more adequate experimental design was used. Tyr :: NRAS ^Q61K × INK4a / ARF ^{- / -} mice develop melanomas with characteristics similar to human disease (Ackermann et al. , 2005). DMBA (7,12-dimethylbenz [a] anthracene purchased from Sigma) was topically applied to mice once. When the pigmented lesions reached 1 mm in diameter, the control PEI, naked pIC or pIC conjugated to PEI was administered in vivo as intravenous injection (ip) twice weekly.

Again, it was noted that the antitumor activity of BO-110 was significantly higher than that of pIC, as indicated by direct measurements of the size of the tumors ( figures 13A and 13B ), the metabolic activity of the tumors determined using PET-CT ( figure 13C ), and histological analysis ( Figure 13A ).

Interestingly, BO-110 doubled the life expectancy without any progressive lesions ( Figure 13A ) at therapeutic doses used without signs of secondary toxicity (see analysis in Figure 13D ).

These results confirm the feasibility of treatment methods based on the introduction of dsRNA analogues in order to combat the aggressive behavior of melanoma cells.

Example 12. The cytotoxic effect of BO-110 on various tumor cells

Since the genetic and epigenetic changes that occur in melanoma and affect the perception of dmRNA and autophagy may not persist in various types of cancer, it was not clear whether BO-110 could have a therapeutic effect in other malignant neoplasms. In particular, tumors of the pancreas, colon, bladder, brain, breast, prostate, lung and ovary are aggressive and resistant to various methods of treatment, partly due to pleiotropic inactivation of cell death programs.

To determine whether BO-110 could represent a new broad-spectrum antitumor strategy, a series of independently isolated cell lines belonging to the above cancers were selected from the well-known NCI-60 panel ( Figure 14 ). Thus, collectively, these cell lines span many tumors (i.e., pancreas, colon, bladder, breast, prostate, lung, and ovary) of various genetic backgrounds. As shown in FIG. 14 , the analyzed cell lines had the same sensitivity to BO-110 as the reference control from melanoma cells. From these data, it follows that BO-110 is capable of triggering a double induction of apoptosis and autophagy, leading to coordinated and selective destruction (without affecting the viability of normal compartments) not only of melanoma, but also of cells belonging to various other types of tumors, for example: pancreas gland, colon, bladder, breast, prostate, lung and ovary.

Example 13. BO-110 Induced Cell Death Depends on Activation of MDA-5, Noxa, and Autophagy in Tumor Cell Lines

Since sensitivity to BO-110 cannot be predicted in advance (that is, based on the type of tumor cells), it was necessary to determine the signaling cascades mediating the response to BO-110. High-performance genetic analyzes (based on cDNA matrices) in melanoma cells indicated that BO-110 is capable of causing strong upregulation of the MDA-5 dsRNA sensor, as well as proapoptotic factor NOXA. Interestingly, using immunoblot analyzes, the inventors demonstrated that in reality sensitivity and resistance to BO-110 (for example, in the HCT116 or MiaPaCa2 lines) correlates with the ability of cells to induce MDA-5 and NOXA ( Figure 15 ). In accordance with the pro-apoptotic role of BO-110 indicated above, a distinct processing of caspase-9 was observed in sensitive cell lines, which can be visualized as changes in electrophoretic mobility ( Figure 15 ).

Bibliography

LIST OF SEQUENCES

<110> Fundación Centro Nacional de Investigaciones Oncológicas Carlos III.

<120> METHOD FOR IDENTIFICATION OF COMPOUNDS FOR TREATMENT OF CANCER

<130> PCT-03837

<160> 3

<170> PatentIn version 3.3

<210> 1

<211> 21

<212> DNA

<213> Homo sapiens

<220>

<221> misc_feature

<222> (1) .. (21)

<223> The target sequence of kshRNA used to reduce

Regulation MDA-5

<400> 1

cgcaaggagt tccaaccatt t 21

<210> 2

<211> 19

<212> DNA

<213> Artificial

<220>

<223> Direct RT-PCR primer for detecting MDA-5 expression

<400> 2

ggcaccatgg gaagtgatt 19

<210> 3

<211> 20

<212> DNA

<213> Artificial

<220>

<223> RT-PCR reverse primer for detecting MDA-5 expression

<400> 3

atttggtaag gcctgagctg twenty

Claims (10)

1. A complex comprising a combination of polyinosine polycytidylic acid (pIC) and polyethyleneimine (PEI), wherein said complex induces autophagy in melanoma cells or in a cell line derived from melanoma cells, and where said complex comprises a combination of pIC and linear PEI with an N / P ratio 1 to 5. 2. The complex according to claim 1, which induces autophagy in a cell line selected from:
(i) Pancreatic cancer cell line groups: IMIMPC2, MiaPaCa2, Aspc1, A6L, SKPC1 and Panc-1;
(ii) groups of colon cancer cell lines: CACO, SW480 and SW1222;
(iii) bladder cancer cell line groups: RT112, MGHu4, 639V, 253J, MGHu3 and SW1170;
(iv) glioma cell line groups: U87MG, U251 and T98G;
(v) groups of breast cancer cell lines: MDA231, MCF7, and T47D;
(vi) groups of prostate cancer cell lines: LNCaP, PC3, and DU145;
(vii) lung cancer cell line groups: H1299 and NCIH460; and
(viii) groups of ovarian cancer cell lines: NCI H23, CHQK1 and SK-OV-3. 3. The complex according to claim 1 or 2, which does not affect normal melanocytes. 4. The complex according to claim 1 or 2, consisting of a combination of pIC having a chain length of at least 1000 nucleotides, and PEI. 5. The complex according to claim 1 or 2, which contributes to the double induction of apoptosis and autophagy. 6. A pharmaceutical composition for use in the treatment of melanoma, containing an effective amount of a complex according to any one of claims. 1-5. 7. The pharmaceutical composition according to claim 6 for use in the treatment of one of the following cancers: pancreas, colon, bladder, breast, prostate, lung, ovary and glioma. 8. The pharmaceutical composition according to claim 6 for use in the treatment of melanoma in immunocompromised patients. 9. The use of the complex according to any one of paragraphs. 1-5 as a treatment for melanoma. 10. The use of the complex according to any one of paragraphs. 1-5 as a treatment for at least one of the following cancers: pancreas, colon, bladder, breast, prostate, lung, ovary, and glioma.

Images