The Antitumor Peptide CIGB-552 Increases COMMD1 and Inhibits Growth of Human Lung Cancer Cells

We have demonstrated that the peptide L-2 designed from an alanine scanning of the Limulus-derived LALF32-51 region is a potential candidate for the anticancer therapy and its cell-penetrating capacity is an associated useful property. By the modification in the primary structure of L-2, a second-generation peptide (CIGB-552) was developed. However, the molecular mechanism underlying its cytotoxic activity remains partially unknown. In this study, it was shown that CIGB-552 increases the levels of COMMD1, a protein involved in copper homeostasis, sodium transport, and the NF-κB signaling pathway. We found that CIGB-552 induces ubiquitination of RelA and inhibits the antiapoptotic activity regulated by NF-κB, whereas the knockdown of COMMD1 blocks this effect. We also found that CIGB-552 decreases the antioxidant capacity and induces the peroxidation of proteins and lipids in the tumor cells. Altogether, this study provides new insights into the mechanism of action of the peptide CIGB-552, which could be relevant in the design of future anticancer therapies.


Introduction
In previous work, a peptide-based approach was used to identify peptides devoid of LPS-binding capacity from LALF residues 32-51. Two peptides (L-2 and L-20) lost their ability to bind LPS and exhibited a differential cytotoxic activity, although a similar cell penetrating capacity was demonstrated for both peptides [1]. We introduced a chemical modi�cation in the primary structure of the peptide L-2 to improve the biological activity and speci�city. e chemical modi�cation included the substitution of a natural amino acid residue by an unnatural amino acid (D-con�guration) and blocked Nterminal by acylation. is modi�cation led to the development of a second-generation peptide (CIGB-552) with increased cytotoxic activity on murine and human tumor cells [2]. Although the antitumor effects of the peptides involve an increase in the apoptosis and a negative regulation of cell-cycle progression, little is known regarding this mechanism of action.
In this study, two complementary approaches were used: a yeast two-hybrid search for molecules that speci�cally interact with the peptides and a pull-down technique to

Peptides Synthesis.
Peptides were synthesized on a solid phase and puri�ed by reverse-phase-high-performance liquid chromatography to >95% purity on an acetonitrile/H 2 O-tri�uoracetic acid gradient and con�rmed by ion-spray mass spectrometry (Micromass, Manchester, UK). Lyophilized peptides were reconstituted in phosphatebuffered saline (PBS) for experiments in vitro. e sequences of peptides used were L-2: HARIKPTFRRLKWKYKGKFW; L-20: HYRIKPTFRRLKWKYKGKFA and CIGB-552 secondgeneration peptide Ac-HARIKpTFRRlKWKYKGKFW where proline and leucine were substituted by D-amino acid; and N-terminal blocked by acylation.

Yeast Two-Hybrid Screening.
Oligonucleotides with the sequences corresponding to peptides L-2 and L-20 were synthesized and cloned in-frame into pGBKT7 yeast twohybrid vector (Table 1). e recombinant clones pGBKT7-L2 and pGBKT7-L20 were veri�ed by DNA sequence and subsequently transformed into yeast strain AH109.
A matchmaker pretransformed liver cDNA library in Y187 (Clontech) was used to identify the protein interactions of L-2. Brie�y, 5 × 10 8 AH109 cells containing the plasmid pGBKT7-L2 were grown and matted with 5 × 10 8 Y187 cells containing the cDNA library from human liver and transferred to minimal medium plates SD-Trp-Leu-His-Ade and grown at 30 ∘ C 7 days. e positive colonies were grown in Trp-Leu medium and the plasmids recovered and transformed into DH10B cells. Each individual clone was transformed in Y187 strain and the interaction veri�ed by matting with the strain AH109 transformed with the plasmids pGBKT7, pGBKT7-L2, and pGBKT7-L20. e positive clones were sequenced and their sequences were analyzed using BLAST [14].

Western Blot and Immunoprecipitation
Analysis. 40 g of total protein extract was applied on 7.5% to 12.5% SDS polyacrylamide gels and Western blot conducted by standard procedures [16]. e primary antibodies used . cDNA was diluted 20-fold and 5 L was used in each quantitative PCR reaction (qPCR). Reactions were conducted in Rotor Gene 6000 equipment using ABsolute SYBR GreenQPCR kit (Abgene, Epsom). Primers sequences for reference genes: B2M, GAPDH, HMBS, ACTB, DDX5 and gene of interest COMMD1 were selected from http://primerdepot.nci.nih.gov/ database [18]. e gene expression of selected genes was conducted following MIQE recommendations [19]. Speci�city of all products was veri�ed by melting curve analysis. Reference genes were selected using GeNorm soware [20]. Quanti�cation was carried out using the ΔΔ CT method. A statistical analysis was conducted using REST soware [21].

Measurement of Oxidative Stress
Variables. All biochemical parameters of oxidative stress were determined by spectrophotometric methods using a Pharmacia 1000 Spectrophotometer (Pharmacia LKB, Uppsala, Sweden). Total protein levels were determined using the method described by Bradford with bovine serum albumin as standard [22]. SOD activity was determined using RANSOD kit (Randox Labs, Crumlin), where xanthine and xanthine oxidase were used to generate superoxide anion radicals (O 2 •− ), which react with 2-(4-iodophenyl)-3-(4-nitrophenol)-5-phenyltetrazolium chloride (INT) to form a red formazan dye. SOD activity was measured by the inhibition degree of this reaction [23]. e advanced oxidation protein products (AOPP) were measured as described previously [24]. Brie�y, 1 mL of samples in PBS was treated with 50 L of 1.16 M potassium iodide followed by the addition of 100 L of glacial acetic acid. e absorbance was immediately read at 340 nm. AOPP concentration was expressed as M of chloramines-T. e concentration of malondialdehyde (MDA) was determined using the LPO-586 kit obtained from Calbiochem (La Jolla, CA, USA). In the assay, the production of a stable chromophore was read at 586 nm aer 40 min of incubation at 45 ∘ C [25]. Freshly prepared solutions of MDA bisdimethyl acetal (Sigma St Louis, MO, USA) were employed to generate standard curves. Ferric reducing ability of plasma (FRAP) was assayed through the reduction of Fe 3+ to Fe 2+ by cell lysates or ascorbic acid as reference. e Fe 2+ -2,4,6-tripiridils-triazine complex was detected at 593 nm [26]. All results shown are the mean of duplicates of at least three independent experiments with SE.

Generation of Stable H460 COMMD1 Knockdown
Cell. Plasmids encoding a short hairpin control (shControl) and plasmid encoding a short hairpin targeting COMMD1 mRNA sequence (shCOMMD1) were obtained from Genecopeia (USA). Generation of H460 stable cell lines was performed with Lentivirus infection with the addition of 4 g/mL polybrene (Sigma). e selection was done in RPMI 1640 supplemented with 1 g/mL puromycin (Sigma) according to the manufacturer's instructions. e knockdown level of COMMD1 was determined by Western blot analysis using mouse monoclonal anti-COMMD1. Determination of the IC 50 values was performed as described previously [1].

Cell Cycle Assay.
Cells were seeded on 6 well plates during 24 h and then treated with 25 M of CIGB-552 for 24 h. e harvested cells were �xed gently by putting 100% ice-cold ethanol in freezer for 2 h. Subsequently, cells were resuspended in 300 L of PBS containing 40 g/mL of propidium iodide and 10 g/mL DNase-free RNase and incubated for 20 min at 37 ∘ C. Aer gating out cellular aggregates, the cell cycle distribution analysis was done on FACSCalibur �ow cytometer using Cell�uest soware (Becton Dickinson). For each sample, at least 20,000 cells were counted and plotted on a single parameter histogram.

Detection of Apoptosis.
Cells in early and late stages of apoptosis were detected with an Annexin V-FITC apoptosis detection kit from Sigma (041M4083). Cells were treated with CIGB-552 (25 M) and incubated for 24 h and 48 h prior to analysis. Brie�y, 2.5 × 10 5 cells were washed with PBS and adjusted in 1 × binding buffer to a concentration of 1 × 10 6 /mL. To 100 L of cell suspension, 5 L of Annexin V-FITC and 10 L propidium iodide (PI) were added and incubated for 10 min at room temperature prior to analysis. Samples were analyzed (20,000 events) using a Becton Dickinson FACSCalibur instrument. Cells that were positive for Annexin V-FITC alone (early apoptosis) and Annexin V-FITC and PI (late apoptosis) were counted.

Statistical
Analysis. e quantitative data in this paper are represented as means ± SD. Statistical evaluation was made using the Mann-Whitney test. Differences were considered to be signi�cant at 0 05.

�denti�cation of �roteins �at �nteract wit� �eptides
Derived from LALF 32−51 Region. To identify proteins that interact with the antitumor peptides L-2 and L-20, a yeast two-hybrid screening was performed. As the peptide L-2 has shown the major antitumor activity, the plasmid pGBKT7-L-2 was selected as bait in the yeast two-hybrid screening of a human liver cDNA library to identify peptide-binding partners. A total of 10 7 diploids were screened. Eighty-seven different positive clones were obtained. e interaction of each positive clone was veri�ed by matting with pGBKT7 as a negative control and with constructions pGBKT7 L-2 and pGBKT7 L-20. From the initial number of clones thirty-eight positive interactions were con�rmed. All positive clones were sequenced and their sequences analyzed using BLAST. Among them, a plasmid containing the sequence of COMMD1 (from amino acids 6 to 190) was identi�ed. Interestingly, the diploid of COMMD1 with the L-20 peptide on selection plate showed a lower growth indicating a lower strength of this interaction compared with L-2 ( Figure 1(a)).
A second GAL-4-based yeast two-hybrid screening identi�ed the region containing the amino acids 111-190 of COMMD1 as the potential interaction site for the peptides Figure 1(a), (bottom of the �gure). e proteolytic stability of natural peptides is one of the principal limitations for their use as drug candidates. In this study, the substitution of proline (Pro6) and leucine (Leu11) by an unnatural amino acid and blocking N-terminal ends by N-acylation from L-2 resulted in a second generation peptide named CIGB-552, which showed an increased antitumor activity in vitro and in vivo [2]. is �nding suggests that the incorporation of unnatural amino acids in the sequence could improve the metabolic stability of the peptide.
To elucidate the functional signi�cance of this chemical modi�cation, the interaction between COMMD1 and the peptides was examined. e interaction between COMMD1 and the synthetic peptides L-2, L-20, and CIGB-552 was evaluated by pull-down analysis in human lung cancer cells H460, following Western blot with speci�c COMMD1 antibody. As shown in Figure 1(b), COMMD1 was detected in the peptides-pull-down precipitation con�rming the existence of speci�c COMMD1/peptide complexes in cells that express endogenously COMMD1. �uanti�cation of the complexes COMMD1/peptides (RI) inversely correlates with the cytotoxic activity of the peptides expressed by its inhibitory concentration (IC 50 ), Table 2. e complex CIGB-552/COMMD1 increases in respect to the complex's L-2 and L-20/COMMD1 indicating that modi�cations done in the primary structure of the peptides increased the affinity to COMMD1 and this correlates with the increased cytotoxic activity. e results of two hybrid and pull-down experiments indicate that the interaction between the peptides and COMMD1 is speci�c and that the strength of this interaction may be relevant for the antitumor effect of the peptides. In addition, the interaction between CIGB-552 and COMMD1 was also con�rmed in whole-cell lysates of human cancer cells of different histological origins (data not shown). Pull-down assay demonstrating speci�c COMMD1/peptides complex in H460 cells. �ndogenous COMMD1 was precipitated from H460 cell lysates. e biotinylated peptides bond to streptavidin sepharose was used as bait and cell lysates as prey. e precipitated material was analysis by Western blot using antibodies directed against endogenous COMMD1. C (−) indicates streptavidin sepharose without peptide. Actin was used as input housekeeping. Relative levels of precipitated COMMD1 were determined by densitometry and normalized to actin. Ratios were depicted under each lane (RI) relative to COMMD1 levels immuprecipitated with peptide L-20. ImajeJ was used for quanti�cation.

CIGB-552 Accumulates COMMD1 in Human Cancer
Cells. Given that COMMD1 was identi�ed as a protein that associates to the above-mentioned antitumor peptides, we studied the role of COMMD1 in the mechanism of action of CIGB-552. First, the cellular expression of COMMD1 in whole-cell lysates of human cancer cells of different histological origin was determined using Western blot analysis, and the proteasome inhibitor MG132 which induces the accumulation of COMMD1 was used as a positive control [27]. ese experiments revealed an increase in the levels of COMMD1 aer 5 h of treatment with the peptide. Treatment with MG132 led to the increase of COMMD1 but not at a greater extent than CIGB-552 (Figure 2(a)). e cellular accumulation of COMMD1 was also found in situ immuno�uorescence in human cancer cells. Both cell lines assayed, MCF7 and HT29, showed the accumulation of COMMD1 aer 5 h of treatment with the peptide similar to the obtained results by Western blot (Figure 2(b)). Talking together these data con�rm that the peptide CIGB-552 induces the accumulation of COMMD1. COMMD1 undergoes constitutive nucleocytoplasmic transport and its nuclear localization is needed for negative regulation of NF-B signaling [28]. Since our interest in this study is focused on the human lung cancer, the levels of the protein COMMD1 in the cytoplasm and nucleus of the H460 cells treated with CIGB-552 were evaluated by Western blot. We found a low expression of COMMD1 in untreated cells. However, in response to CIGB-552, COMMD1 was increased in the cytoplasm and nucleus of cells at 40 min following treatment, and, most interestingly, this increment remained until up to 5 h aer treatment (Figure 2(c)). To examine whether the increase of COMMD1 levels is due to an increase in the RNA expression, a qPCR experiment was performed. e results showed that increased levels of COMMD1 were not accompanied by signi�cant changes in mRNA expression of the protein (Figure 2(d)), suggesting a posttranscriptional effect of CIGB-552 on COMMD1.

CIGB-552 Promotes the Ubiquitination of RelA Subunit of NF-B.
It has been known that overexpression of COMMD1 accelerates the ubiquitination and degradation of the RelA subunit of NF-B and decreases the activation of antiapoptotic genes [29]. Taking into account the above data, the second question to elucidate in this study was the effect of CIGB-552 on the NF-B signaling in human lung cancer cells. e effect of CIGB-552 on the endogenous levels of ubiquitinated RelA was investigated. Immunoprecipitation of endogenous RelA using a mouse anti-RelA followed by antiubiquitin Western blot con�rmed the increased amounts of ubiquitinated RelA in response to the peptide as early as 2 h aer treatment, and the increment remained for 12 h aer treatment (Figure 3(a)). Immunoprecipitation using a rabbit anti-IgG as negative control did not result in the recovery of ubiquitinated RelA, indicating the speci�city of the recovered material. As shown in Figure 3(a), CIGB-552 as well as MG132 (used as a positive control) increased the ubiquitinated forms of RelA. To investigate whether the CIGB-552 induces ubiquitination and subsequent degradation of RelA through a proteasome-dependent process, the effect of MG132 and CIGB552 on the basal levels of the protein was tested. Treatment with CIGB-552 led to a decrease in basal levels of RelA, while protein levels were accumulated in the presence of MG132 treatment and CIGB-552. is result suggests that CIGB-552 induces ubiquitination of RelA and promotes its proteasomal degradation (Figure 3(b)).

CIGB-552 Regulates Apoptosis-Related Proteins in H460
Cells. Further, we assess whether CIGB-552 could modulate the expression of proteins involved in the intracellular apoptosis signaling. As shown in Figure 3   facilitates cellular disassembly and serves as a marker of cells undergoing apoptosis. Our results provide molecular evidence of apoptosis induction by CIGB-552 treatment. e kinetics of responses to CIGB-552 was examined. We found that RelA ubiquitination occurred within 2 h of treatment (Figure 3(a)) and was associated with a nuclear accumulation of COMMD1 (Figure 2(c)). ese early events had no effect on the levels of apoptotic-related proteins (Figure 3(c)). Based on our observations so far, the accumulation of COMMD1 and ubiquitination of RelA are events that precede a repression of the antiapoptotic activity of NF-B.
�.5. C����� �e���e��y ��te�� �e��adat��� �� �e�� �ed�ated by CIGB-552. To assess if COMMD1 has a functional role in the antitumor activity of CIGB-552, knockdown of COMMD1 by siRNA was generated in H460 cell line and the cytotoxic effect of the peptide was determined. As shown in Figure 4(a), the cytotoxic activity of CIGB-552 in knockdown cells decreased in comparison to control cells. To con�rm that the observed reduction in the antitumor effect of CIGB-552 in knockdown cells was not an artifact, the levels of protein were determined by Western blot. As shown in Figure 4(a), COMMD1 expression decreased in respect to control cells (top of the �gure). However, the cytotoxic activity of the peptide was not completely blocked in knockdown cells. is result might be explained by the efficiency of knockdown but we do not discard that other factors could be involved in the cytotoxic activity of the peptide.
Journal of Amino Acids e reduced cytotoxic activity of the peptide in COMMD1 knockdown cells could be associated with decreased RelA ubiquitination and degradation. To study this possibility, the endogenous ubiquitination of the RelA levels in control and COMMD1 knockdown cells treated with the peptide was examined. As shown in Figure 4(b), COMMD1 de�ciency cells resulted in lesser levels of ubiquitinated RelA when treated with the CIGB-552 in respect to the control cells. Next, we examined the levels of endogenous RelA and Bcl-2 in cytosolic extracts of control and COMMD1 de�ciency cells using Western blot. As shown in Figure 4(c), COMMD1 knockdown cells showed greater levels of RelA and Bcl-2 proteins when treated with the CIGB-552 in respect to the control cells.
We have previously reported that the cytotoxic effect of the L-2 peptide involved cell cycle arrest followed by cell death [1]. erefore, the possibility that COMMD1 might be involved in the cell death induced by CIGB-552 was explored. To elucidate this, the alteration of the cell division cycle in control and COMMD1 knockdown cells was analyzed by �ow cytometry. As shown in Figure 5, control cells undergoing apoptosis aer 24 h of treatment showed a signi�cant increase in the sub-G 0 peak. In contrast, the apoptosis induced by the CIGB-552 was blocked in COMMD1 knockdown cells. Apoptosis can be assessed by using several characteristic features of programmed cell death. One variable of apoptosis is phosphatidylserine externalization, which can be measured by Annexin V staining.   Altogether, these results support the hypothesis of the effect of COMMD1 on the apoptosis induced by CIGB-552.

Altered Redox Status of H460 Treated with CIGB-552.
e maturation and activation of the antioxidant superoxide dismutase (SOD1) are highly regulated processes that require several posttranslational modi�cations. Recently, it has been described that COMMD1 is a novel interaction partner of SOD1. COMMD1 impairs SOD1 activity by reducing the expression levels of enzymatically active SOD1 homodimers late in the posttranslational maturation process of SOD1 [30].
Our previous results demonstrated that CIGB-552 induces the accumulation of COMMD1 in the cytosol. Next, we considered it important to determine the relevance of this accumulation in the enzymatic activity of SOD1, and the antioxidant capacity of the lung cancer cells. e activity of the superoxide dismutase SOD1 was measured and the total antioxidant capacity was evaluated by FRAP. Consistent with our expectation, the activity of SOD1 in knockdown cells is markedly increased compared to H460 cells aer 8 h of treatment with CIGB-552 (Figure 7(a)). In line with Vonk's report [30], we also found that endogenous COMMD1 modulates SOD1 activity. As shown in Figure 7(b), the total antioxidant capacity was signi�cantly diminished aer 8 h of treatment. Next, we examined the levels of protein and lipid peroxidation, as a sign of oxidative stress damage by the formation of MDA (malondialdehyde) and AOPP (advance oxidation protein products) [31]. Concordantly, the levels of MDA and AOPP markedly increased aer 8 h of treatment, likely in response to an imbalance in the antioxidant capacity Figure 7(c). Altogether, these data demonstrate that CIGB-552 might promote cell cycle arrest and apoptosis by inducing damage to proteins and lipids in lung cancer cells.

Discussion
Our previous studies showed that the substitution of alanine in speci�c amino acids of the region LALF 32−51 led to the design of peptides that lost the ability to bind LPS and exhibit a differential cytotoxic activity (L-2 and L-20). In this study, we developed the peptide CIGB-552 from the chemical optimization in the primary structure of the peptide L-2 with the purpose of reducing the elimination and biodegradation of the peptide and increasing selectivity or affinity to its potential target [32]. First we con�rmed that the peptides L-2 and L-20 bound the region 111-190 of COMMD1 protein. Particularly a higher capacity of binding was shown by the peptide L-2 and this correlated with its higher cytotoxic activity [1]. Interestingly, our subsequent studies using the pull-down technique demonstrated that modi�cations in the primary structure of the peptides increased the affinity to associate COMMD1 and this correlated with increased antitumor activity. For the �rst time, our results identi�ed COMMD1 as a potential target of the anticancer peptide CIGB-552 and indicated that targeting of a protein-peptide interaction may be a strategy to increase the speci�city and biological activity of novel anticancer peptides derived from LALF 32− 1 region.
Recently, it has been reported that a decreased expression of COMMD1 is frequently observed in a variety of cancers and that this correlates with tumor invasion as well as with the overall patient survival. is suggests that decreased COMMD1 expression might represent a novel mechanism that confers cancer cells with invasion potential and proliferating capacity [13]. CIGB-552 was undoubtedly found to accumulate COMMD1 in cancer cells and this accumulation was not accompanied by changes in the mRNA expression. ese data suggested the outstanding possibility that CIGB-552 may stabilize COMMD1 through posttranscriptional events. e basal levels of COMMD1 expression are tightly regulated by XIAP (X-linked inhibitor of apoptosis). e interaction between COMMD1 and XIAP have been mapped to the COMM domain and identi�ed leucine repetitions within the COMM domain are required for ubiquitination and proteasomal degradation of COMMD1 [33]. Interestingly, the peptides derived from LALF 32−51 region were found to bind COMM domain (amino acids 119-190) and, more important, our results provide the �rst indication that CIGB-552 could regulate the levels of COMMD1 to induce the cell death in cancer cells. e mechanism by which CIGB-552 induces the stabilization of COMMD1 is currently the issue of ongoing researches in our lab.
A central function of NF-B, in particular of RelA subunit, is the regulation of cellular growth and apoptosis [11]. Besides, the persistent activation of NF-B is a recognized contributory factor in a number of carcinomas and may provide the cancer cells with a survival advantage [34].
Recently, it has been demonstrated that ubiquitination and the degradation of RelA subunit by COMMD1-containing ubiquitin ligase is a critical mechanism of regulation of the antiapoptotic activity of NF-B and this event has considerable relevance to cancer prevention and therapy, as well as the postinduction regulation of RelA/NF-B [7,35].
It was found in this study that the treatment of lung cancer cells with CIGB-552 increased the levels of the protein COMMD1 in the cytoplasm and nucleus. is observation, along with the fact that COMMD1 can repress NF-B activation, suggests a signi�cant role for CIGB-552 in the regulation of RelA/NF-B in lung cancer cells. Our results demonstrated that CIGB-552 induces the ubiquitination of the RelA subunit of NF-B and our data also indicated that this effect promotes its proteasomal degradation. e stabilization and nuclear accumulation of COMMD1 mediated by CIGB-552 could be an attractive mechanism for the regulation of the constitutive activity of the transcription factor NF-B in cancer cells.
Consistently, our studies showed a role for COMMD1 in the cytotoxic effect of CIGB-552. H460 knockdown of COMMD1 was more resistant to the cytotoxic effect of the peptide, and this COMMD1 de�ciency correlates with reduced levels of ubiquitinated RelA and apoptosis.
Oxidative stress, involved in the etiology of cancer, results from an imbalance in the production of reactive oxygen species (ROS) and the cells' own antioxidant defenses. High levels of oxidative stress have been observed in various types of cancer cells. Accordingly, there is an aberrant regulation of redox homeostasis and stress adaptation in cancer cells and this event is associated with drug resistance [36,37]. Here, we found that the treatment of the lung cancer cells with CIGB-552 impaired the SOD1 activity and it could be associated with the accumulation of COMMD1. Moreover, the CIGB-552 induced a failure of the antioxidant defenses and this effect was accompanied by damage to proteins and lipids. ese observations suggest that CIGB-552 plays a signi�cant role in the regulation of the redox status of cancer cells.

Conclusions
Altogether, our results demonstrate that CIGB-552 could regulate the anti-apoptotic activity of NF-B and the oxidative stress in lung cancer cells, two biological processes involved in the survival and growth of tumor cells.

Con�ict of �nterests
No potential con�ict of interests was disclosed.