Xinqidi,Your biological research partner
Xinqidi Biotech Co.,Ltd,Wuhan,China 2008-2021
R&D 13th year

Bladder cancer therapy using a conformationally fluid tumoricidal peptide complex

Issuing time:2021-06-10 09:55


Partially unfolded alpha-lactalbumin forms the oleic acid complex HAMLET, with potent tumoricidal activity. Here we define a peptide-based molecular approach for targeting and killing tumor cells, and evidence of its clinical potential ( NCT03560479). A 39-residue alpha-helical peptide from alpha-lactalbumin is shown to gain lethality for tumor cells by forming oleic acid complexes (alpha1-oleate). Nuclear magnetic resonance measurements and computational simulations reveal a lipid core surrounded by conformationally fluid, alpha-helical peptide motifs. In a single center, placebo controlled, double blinded Phase I/II interventional clinical trial of non-muscle invasive bladder cancer, all primary end points of safety and efficacy of alpha1-oleate treatment are reached, as evaluated in an interim analysis. Intra-vesical instillations of alpha1-oleate triggers massive shedding of tumor cells and the tumor size is reduced but no drug-related side effects are detected (primary endpoints). Shed cells contain alpha1-oleate, treated tumors show evidence of apoptosis and the expression of cancer-related genes is inhibited (secondary endpoints). The results are especially encouraging for bladder cancer, where therapeutic failures and high recurrence rates create a great, unmet medical need.


Targeted cancer therapies have made significant advances but the lack of tumor specificity remains a significant concern1,2. Few current therapies kill cancer cells without harming healthy tissues, and severe side effects have become accepted as a necessary price to pay for survival or cure. The notion of successfully combining efficacy with increased tumor selectivity is justly regarded with skepticism. Yet, a serendipitous discovery has provided insights into mechanisms of tumor-specific cell death, induced by unfolded polypeptide chains, which acquire tumoricidal activity by forming fatty acid complexes3,4. Extensive observations in tumor models and clinical studies have further defined the protein–lipid complexes as an interesting class of molecules with significant therapeutic potential5.

The findings are challenging, as protein unfolding and loss of structural definition is associated with a gain of toxicity, due to the formation of amorphous aggregates and amyloid fibrils6,7. Native protein structure is often regarded as a prerequisite for biological function, by epitope-specific interactions and molecular fitness for a finite number of cellular targets. Yet, a lack of structural definition may, in some cases, result in a gain of function, in part by uncovering different conformations and exposing peptide motifs that are unavailable in the native state8,9. Such effects have been predicted for membrane perturbing α-helices in antimicrobial peptides, where the ability to destabilize lipid bilayers has been proposed to reside in the three-dimensional conformation rather than the amino acid sequence10.

Alpha-lactalbumin is crucial for the survival of lactating mammals. In its native state, the protein serves as a substrate specifier in the lactose-synthase complex11, defining the nutritional content and fluidity of milk. Partially unfolded alpha-lactalbumin, in contrast, forms an oleic acid complex, named HAMLET, with potent tumoricidal activity3,4,8,9. The HAMLET complex kills a range of tumor cells with rapid kinetics and shows therapeutic efficacy in animal models of colon cancer, glioblastoma, and bladder cancer12,13,14,15. Early, investigator-driven clinical studies demonstrated that HAMLET is active topically, against skin papilloma and induces tumor cell shedding into the urine in patients with bladder cancer5,16.

This study presents a synthetic, peptide-based drug candidate derived from alpha-lactalbumin, which reproduces the tumoricidal properties of HAMLET and allows for a full translation of these findings into the clinic. Through complementary nuclear magnetic resonance (NMR) analysis and computational modeling, the molecular basis for this “gain-of-function” is defined, including the three-dimensional structural motifs that determine fatty acid-binding efficiency and tumoricidal activity. The therapeutic efficacy of the complex is demonstrated in patients with non-muscle invasive bladder cancer (NMIBC), in a fully controlled clinical trial.


Peptide-specific tumoricidal activity

To understand the involvement of specific peptide motifs in tumor cell death, we synthesized the N-terminal alpha-helical domain (residues 1–39, alpha1) or the beta-sheet (40–80, beta) domains of human alpha-lactalbumin (Fig. 1a). The alpha1 peptide formed complexes with oleate (alpha1–oleate, 1:5) and circular dichroism (CD) spectra detected an increase in alpha-helical structure content in these complexes (Fig. 1b). The beta–oleate complex remained structurally unchanged (Fig. 1b). Alpha1–oleate triggered a rapid, dose-dependent death response in human lung- and kidney carcinoma cells and in murine bladder cancer cells (Fig. 1c, d). The beta–oleate complex lacked tumoricidal activity and tumor cells subjected to the naked alpha-helical peptides (35 μM) or oleate (175 μM) controls were not tumoricidal (Fig. 1c, d and Supplementary Fig. 1). The loss of cell viability was irreversible, as shown after 10 days, by using colony assays (Fig. 1e and Supplementary Fig. 1). Membrane blebbing occurred in tumor cells within minutes of exposure to alpha1–oleate but the naked peptide- and oleate controls were not active (Fig. 1f and Supplementary Fig. 1). Rapid K+ fluxes were recorded, further defining the membrane response (Fig. 1g). Pretreatment of the cells with Na+ and K+ flux inhibitors reduced cell death by 40–50%, linking the membrane response to tumor cell death (Fig. 1h). The alpha1–oleate complex was rapidly internalized by tumor cells and by TUNEL staining, alpha1–oleate was shown to induce double-strand DNA breaks in the tumor cells, indicative of apoptosis (Fig. 1i, j).

Fig. 1: Tumoricidal activity of two non-homologous alpha-helical peptide–oleate complexes.

a Ribbon representation of the crystallographically determined three-dimensional structure of human α-lactalbumin (PDB ID: 1B9O), indicating the alpha1 (blue), beta (green), and alpha2 (gray) domains. The calcium ion is not shown. b Far-UV circular dichroism spectra of synthetic alpha1 peptide, beta peptide, and their respective peptide–oleate complexes. c, d Death response in human lung (A549), kidney (A498), and murine bladder (MB49) carcinoma cells, quantified as a reduction in ATP levels (c, P = 3.26E−5 for A549, 0.013 for A498 and 0.005 for MB49, alpha1–oleate compared to beta–oleate) or PrestoBlue fluorescence (d, P = 0.007 for A549, 0.003 for A498 and 0.002 for MB49, alpha1–oleate compared to beta–oleate). Cells were treated with the alpha1–oleate complex (blue) or the beta–oleate complex (green), (3 h, 35 μM, cell death compared to PBS controls). For controls exposed to the naked peptides or oleate alone, see Supplementary Fig. 1d. eColony assay showing dose-dependent long-term effects of alpha1-oleate. A representative image is shown from two independent experiments. Scale bar = 5 mm. f Alpha1–oleate triggers rapid membrane blebbing in A549 lung carcinoma cells (35 μM, 10 min). Scale bar = 10 μm. A representative image is shown from three independent experiments. g K+ efflux in A549 lung carcinoma cells exposed to alpha1–oleate and inhibition with BaCl2. h Inhibition of cell death by the ion flux inhibitors Amiloride and BaCl2 (100 μM), measured by PrestoBlue fluorescence (P = 0.031 for 21 μM + BaCl2, 0.005 for 21 μM + Amiloride, 0.028 for 35 μM + BaCl2, and 0.014 for 35 μM + Amiloride, compared to no inhibitor). i DNA strand breaks detected by TUNEL staining in alpha1–oleate-treated A549 lung carcinoma cells (n = 50 cells per group). Scale bar = 20 μm. jAlexaFluor568-labeled alpha1–oleate (red) is internalized by A549 lung carcinoma cells. Nuclei are counterstained with DAPI (blue) (n = 52 cells per group). Scale bar = 10 μm. Data are presented as mean ± SEM from three independent experiments, *P < 0.05, **P < 0.01, ***P < 0.001, analyzed by two-tailed unpaired t-test (c, d, h, j) and 2-way ANOVA using Dunnett’s correction (i).

In a screen of proteins with membrane-integrating properties, SAR1 was found to form tumoricidal complexes with oleic acid, reproducing effects of alpha1–oleate (Supplementary Figs. 1 and 2). SAR1 is a membrane-integrating protein of the COPII complex that induces membrane tubulation by insertion of its N-terminal amphipathic α-helix17,18,19. The N-terminal alpha-helical peptide (residues 1–23, sar1alpha) formed an oleate complex, which efficiently killed tumor cells (Supplementary Figs. 1 and 2). The naked peptide- and oleate controls were not active (Supplementary Fig. 1). Sar1alpha–oleate triggered membrane blebbing in tumor cells, rapid K+ fluxes were recorded, and tumor cell death was partially inhibited by Na+ and K+ flux inhibitors, suggesting a similar mode of action of the two complexes, despite low sequence homology (Supplementary Fig. 2). The sar1beta–oleate complexes and naked peptide controls did not trigger tumor cell death, however (Supplementary Figs. 1 and 2).

In preparation for the clinical trial, the safety of alpha1–oleate was investigated in C57BL/6J mice carrying MB49-induced bladder tumors13. Therapeutic efficacy in 100% of treated mice compared to the sham group and a lack of toxicity was demonstrated, providing the necessary background to plan the clinical trials13.

Biomolecular NMR analysis of the peptide–oleate complexes

1H NMR spectra of the alpha1–oleate and sar1alpha–oleate complexes detected a shift from sharp signals for the naked peptides to broad signals and poor chemical shift dispersion for the oleate complexes (Fig. 2a, b), suggesting a conformational change from a random-coil fast-exchange time regime to an intermediate millisecond timescale. Broadening in the amide, side-chain methyl and aromatic regions suggests that interactions between fatty acids and peptides occur throughout the molecules. Two-dimensional nuclear Overhauser effect spectroscopy (2D NOESY) spectra identified non-covalent, relatively short through-space interactions between the respective peptides and fatty acids. Important nuclear Overhauser effects (NOEs) were detected between the olefinic protons (5.23 ppm) of oleic acid and the Hα and aromatic protons of alpha1 and between the sar1alpha aromatic region and the oleic acid olefinic protons (Fig. 2c, d). The downfield chemical shift of amide protons observed between 7.6 and 8.8 ppm suggests the presence of secondary structure in alpha1 and alpha1–oleate. Well-resolved signals obtained from the one-dimensional 1H NMR spectra provided a stoichiometry of 3.7 oleate molecules per alpha1 peptide. Chemical shift mapping revealed a cluster of residues with aliphatic side chains that change upon the binding of oleate, providing further evidence of interactions between peptides and fatty acids (Fig. 2e, f).

Fig. 2: Biomolecular NMR analysis of naked alpha1- and sar1alpha peptides and their oleate complexes.

a, b One-dimensional 1H NMR spectra. The naked alpha1- (a, black) and sar1alpha- (b, black) peptides assume an ensemble of structures that interconvert rapidly and are therefore seen as sharp peaks. The alpha1–oleate (a, red) and sar1alpha–oleate complexes (b, red) show broader peaks. Arrows indicate the indole 1H signals arising from the three Trp side chains present in the sar1alpha peptide. c, d Two-dimensional 1H NOESY spectra of alpha1–oleate and sar1alpha–oleate complexes, showing atomic-level proximities of the fatty acid to the respective peptide. The spectra highlight NOEs between the 9,10 olefinic protons (5.23 ppm) of oleic acid with the Hα protons and aromatic protons of the alpha1–oleate complex (c) and the sar1alpha–oleate complex (d). e, f Two-dimensional 1H–13C Heteronuclear Single Quantum Coherence (HSQC) spectra overlays of the alpha1 peptide (red) and alpha1-oleate complex (black). Chemical shift perturbation is detected in the aromatic side-chain region and the imidazole ring protons (e, green circled regions), and in the aliphatic side chain regions (f). g Size-exclusion HPLC (SE-HPLC) of the alpha1 peptide and the alpha1–oleate complex, mapped onto a standard calibration curve. h Diffusion-ordered NMR spectroscopy (DOSY) of the alpha1 peptide, alpha1–oleate complex, human serum albumin (HSA), and oleate suspension.

Hydrodynamic volume measurements carried out with size-exclusion high-performance liquid chromatography (SE-HPLC) and diffusion-ordered NMR spectroscopy (DOSY) showed that the alpha1–oleate complex RH was considerably larger (27.4 and 29.3 Å, respectively) than the naked peptides (16.1 Å from SE-HPLC and 14.4 Å from DOSY) (Fig. 2g, h and Supplementary Figs. 3 and 4). The distinctly larger R2 values (transverse relaxation rate) for the complex than those of alpha1 peptide and human serum albumin (HSA) suggested slower millisecond to microsecond exchange processes (Supplementary Table 1 and Supplementary Figs. 5 and 6). Importantly, the R2 values for the complex were also different from oleate in an aqueous solution, suggesting that the dynamics of the complex were clearly different from micelle/vesicle-like particles formed from oleic acid/oleate with no peptide component.

Computational analyses of the peptide- and peptide–oleate system

Computational simulations also pointed toward structural heterogeneity, showing that the naked peptides and peptide–oleate complexes belonged to wide conformational spaces with relatively deep basins (Fig. 3a, b). Representative structures mapped to different free energy surface minima, revealed prominent alpha-helical secondary structural elements and a hydrophobic oleate core for the peptide–oleate complexes (Fig. 3c, e). The peptide–oleates fold upon this core differently from the naked peptides, which exhibit multiple local minima (Fig. 3d, f, and Supplementary Tables 2 and 3). Naked alpha1 ensembles were characterized by various partially folded helix-turn conformations, whereas naked sar1alpha ensembles exhibited a mixture of the random coil, alpha-helical, and beta-sheet structures.

Fig. 3: Free energy surface analyses of the peptide- and peptide–oleate system.

a, b Superimposition of dihedral principal component analysis (PCA) plots of alpha1 (black points) and alpha1–oleate (red points) systems (a), and sar1alpha (cyan points) and sar1alpha–oleate (magenta points) systems (b). Principal component (PC)1 and PC2 represent the axes of the two greatest variances after mathematical transformation for dimension reduction. cf Free-energy surfaces as a function of the first two principal components for alpha1–oleate (c), naked alpha1 (d), sar1alpha–oleate (e), and naked sar1alpha (f). The representative structures of peptides or peptide–oleate complexes, along with their respective local minima annotations, are colored from the N termini (blue) to the C termini (orange/brown). The free-energy surface of the alpha1–oleate complex contains 2 minima basins, A1 and B1, with A1 representing the major conformational ensemble. The free-energy surface of the sar1alpha–oleate complex contains 3 minima basins, A3, B3, and C3 (with the A3 basin harboring the major structural ensemble), which are characterized by a prominent alpha-helical secondary structural element, as shown from simulation calculated alpha-helical propensities. By contrast, the free-energy surface of the naked sar1alpha shows large structural heterogeneity. Here, minima basins A4 and D4 are represented by helical structures, B4 by beta structure, and C4 and E4 by random coil structures.

A contact probability analysis revealed that the interactions between alpha1 or sar1alpha and oleate were mainly hydrophobic, with a >0.9 contact probability with olefinic protons (Supplementary Tables 4 and 5). The peptide–oleate complexes displayed relatively wide and deep free energy minima basins, suggesting that a multitude of confirmations would be equally possible to visit (Fig. 3c, e). When combined with the R2 relaxation rates, the possibility of multiple sampling of various conformations within a short period of time provides an argument that rather than targeting specific partners, these alpha-helical complexes may potentially be interacting with multiple putative binding partners available on cancer cell surfaces20.

Based on these extensive investigations and the strong agreement of the experimental aspects with the simulated predicted ensembles, it was clear that the apparently unrelated peptides alpha1 and sar1alpha can form complexes with shared structural characteristics, involving a flexible peptide moiety and a fatty acid cluster. This notion resonates with the sequence-function inconsistency among certain antimicrobial amphipathic alpha helices, where peptides with similar overall properties, such as hydrophobicity or charge, can have dramatically different levels of activity10.

A placebo-controlled, randomized clinical trial of alpha1–oleate in patients with NMIBC

NMIBC is common and despite current treatment protocols, recurrence rates are high21,22. To address if the therapeutic effects observed in the murine MB49 bladder cancer model can be translated into the clinic, the investigational product alpha1–oleate was produced under GMP conditions. The alpha1–oleate complex was further subjected to formal toxicity testing and the results have been published13. Toxicity for bladder tissue was not detected at concentrations ranging from 1.7 to 17 mM13.

The clinical safety and therapeutic potential of alpha1–oleate were tested in a single-center, placebo-controlled, double-blind Phase I/II trial (EudraCT No: 2016-004269-14, NCT03560479, Supplementary Note 1, Supplementary Table 6). Patients with suspected NMIBC were randomized 1/1 to receive alpha1–oleate or placebo during a period of 22 days, prior to endoscopic removal of the tumor by transurethral resection (TURB), (Fig. 4a, b). Alpha1–oleate (1.7 mM) or placebo (PBS) was administered intravesical on six occasions (30 mL, days 1, 3, 5, 8, 15, and 22). The placebo solution was identical in appearance to the active treatment. Demographic data, medical history, and vital signs did not differ between the treatment and placebo groups (for details see Supplementary Table 7).

Fig. 4: Clinical study protocol, demographic data, and adverse events.

a Study CONSORT diagram. b Study protocol. After diagnosis and informed consent, the subjects received intravesical instillations of either alpha1–oleate or placebo on six occasions during one month preceding a scheduled transurethral resection (TURB). A safety follow-up was performed 52 days after the first instillation. c Number of adverse events (AEs) in the active and placebo groups. No drug-related adverse events were recorded. There were totally 29 AEs reported by 12 subjects in the active group and by 11 subjects in the placebo group. None of the AEs were related to the investigational product. One AE was severe and two were moderate in the placebo group. The active group had one moderate AE. Two subjects in the placebo group reported severe AE (SAEs). The AEs were evaluated descriptively, and the AE profiles were similar between the placebo and the active groups.

Primary study endpoints

Adverse events (AEs) were recorded and coded according to MedDRA (version 21.1) with a safety follow-up after 52 days (Supplementary Table 8). Procedure-related AEs, such as dysuria and bacteriuria, occurred at a similar rate in the treatment and placebo groups. AEs specific for the treatment group were not detected, suggesting low toxicity of the study medication (Fig. 4c). Furthermore, there was no evidence of a toxic response in healthy tissue samples from patients treated with alpha1–oleate, defined by histopathology or TUNEL staining.

Tumor cell shedding and release of tumor cell clusters were recorded. Cells with uroepithelial morphology were quantified in urine at each visit, before instillation and about 2 h after the instillation of alpha1–oleate or placebo. Alpha1–oleate triggered a rapid increase in cell shedding compared to the pre-instillation sample in all treated patients, at all visits (Fig. 5a–c and Supplementary Fig. 7). In addition, tumor cell clusters were released into the urine in the treatment group. The clusters were relatively large and the presence of tumor stroma in some samples supported their tumor origin (Fig. 5d–f). The cells shed in urine were assigned a pathology score as per the Paris classification (classes 1–6, urine cytology was a secondary endpoint). In the treatment group, an increase in the Paris score was detected in post-inoculation samples compared to pre-inoculation samples (Fig. 5g). A low level of cell shedding in the placebo group was attributed to the instillation procedure and changes in pathology score or shedding of cell clusters were not observed (Fig. 5a–gand Supplementary Fig. 7).

Fig. 5: Primary endpoints: shedding of tumor cells and reduction in tumor size following intra-vesical instillation of alpha1–oleate.

ac Cell shedding increased significantly after alpha1–oleate instillation. a Scatterplot showing individual means of six visits per patient in the treatment group (n = 20) compared to patients receiving placebo (n = 20). Line represents the median. b Comparison of cell numbers in urine before (pre = white) and after (post = black) alpha1–oleate inoculation on visits 1–6 showing increased cell numbers post-inoculation in the treatment group (n = 20 patients per group, P = 0.0030 for visit 1, 0.0098 for visit 2, <0.0001 for visits 3 and 4, 0.0073 for visit 5 and 0.0336 for visit 6) but not in the placebo group. Data are presented as mean ± SEM. c Representative images, illustrating the increase in cell shedding after alpha1–oleate instillation. Magnification = ×400. Scale bar = 50 μm. df Difference in the shedding of tumor cell clusters between the treatment and placebo group. d Scatterplot showing individual means of six visits per patient in the treatment group compared to patients receiving placebo. Line represents the median. e Increased numbers of cell clusters in post-inoculation samples of patients receiving alpha1–oleate (n = 20 patients per group, P = 0.9743 for visit 1, 0.0212 for visit 2, <0.0001 for visits 3, 4, and 5, and 0.0005 for visit 6). Data are presented as mean ± SEM. f Representative images of cancer cell clusters after alpha1–oleate instillation. Magnification = ×400. Scale bar = 50 μm. g Paris grade of shed cells before or after alpha1–oleate instillation. An increase is observed in the treatment group (χ2 test). hReduction in tumor size in patients receiving alpha1–oleate treatment. Images were compared between the time of diagnosis and the time of TURB (P = 0.04, χ2 test for trend compared to placebo, n = 19 for treatment group and n = 20 for placebo group). i Examples of cystoscopy photographs obtained by A.B. at the time of diagnosis and after treatment at the time of TURB. Scale bars = 5 mm. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. The data were analyzed by two-tailed unpaired Mann–Whitney U-test (a, d) or by repeated-measures two-way ANOVA with Sidak’s correction (b, e).

The endoscopic appearance of the tumors was recorded at the time of diagnosis and prior to surgery using a flexible cystoscope with white light-band and narrow-band imaging. Sizes were assessed by an experienced endourologist using fully opened clamps of the flexible forceps and a measuring device close to the tumor. Paired images from 39 patients were evaluated in a blinded manner, by an independent NMIBC expert using a simplified Delphi method23 addressing changes in lesion size, superficial necrosis, and tissue vascularization. A reduction in lesion size was detected in the treatment group (n = 19, Fig. 5h, i). In 1 patient, paired pre-treatment and post-treatment images could not be collected for technical reasons. No difference in superficial necrosis or vascularization was observed and there was no change in lesion size in the placebo group (n = 20).

Secondary end points

Evidence of tumor cell apoptosis was obtained by staining of tumor biopsies obtained at the time of surgery. Biopsies were examined for evidence of treatment-induced apoptosis, quantified by the TUNEL assay (Fig. 6a–c and Supplementary Fig. 8). A significant increase in net mean fluorescence was detected in the treatment group, compared to placebo (see also Fig. 1). Staining was most intense adjacent to the lumen suggesting that a gradient might be formed, from the lumen towards the center of the tumor. TUNEL staining intensity was significantly correlated to cell shedding and alpha1–oleate uptake in individual patients (Fig. 6d) but not to the tumor grade. In healthy tissue biopsies from the treated patients, TUNEL staining was low. Tumors from the placebo group did not show increased TUNEL staining, suggesting that tumor cell apoptosis may be induced by the alpha1–oleate treatment (Fig. 6c).

Fig. 6: Apoptotic response to alpha1–oleate and cellular uptake by tumor cells.

Apoptosis was quantified in tumor biopsies, using the TUNEL assay. Arbitrary units were calculated after subtraction of background staining in TUNEL negative healthy tissue samples. aRepresentative image of TUNEL staining (green = TUNEL, blue = DAPI) in tumor tissue from individual patients receiving alpha1–oleate instillations. Scale bars = 200 μm. b Representative images of TUNEL staining in tumor tissue from individual patients receiving placebo. Scale bars = 200 μm. c Scatter plot demonstrating elevated TUNEL staining intensity in tumor biopsies from patients receiving alpha1–oleate instillations compared to placebo. TUNEL staining was not significantly altered in healthy tissue biopsies from patients receiving alpha1–oleate instillations or placebo (n = 40 tumors and 38 healthy biopsies, two data points were further removed due to medical conditions from patients and confirmed by Grubbs’s outlier test) (two-tailed unpaired Mann–Whitney U-test). Line represents the median. d Correlation of TUNEL staining intensity with cell shedding (P = 0.03, 95% CI 0.0220– 0.6010) and alpha1–oleate uptake (P = 0.01, 95% CI 0.0957–0.6461) (Spearman correlation, two-tailed, approximate P-value, n = 20 for alpha1–o and n = 19 for placebo). e Representative images of alpha1–oleate (red) uptake with counter-stained nuclei (blue). Alpha1–oleate uptake by tumor cells was quantified by staining of shed cells in urine with polyclonal anti-alpha1–oleate antibodies. Scale bars = 20 μm. f Scatterplots of cellular uptake in individual patients receiving alpha1–oleate. Each dot represents the mean fluorescence intensity of six post-instillation samples per patient treated with alpha1–oleate. Comparison of alpha1–oleate uptake by the cell in urine before (pre = white) and after (post = black) alpha1–oleate inoculation on visits 1-6 (repeated-measures two-way ANOVA with Sidak’s correction, P = 0.0049 for visit 1, 0.1913 for visit 2, 0.0067 for visit 3, 0.0025 for visit 4, 0.3807 for visit 5 and 0.0043 for visit 6, n = 20 per group and time point). Line represents the median and bars represent mean ± SEM.

The alpha1–oleate content of shed cells in urine was quantified by immunohistochemistry, using alpha1-specific antibodies. Alpha1-staining was detected in 70% of post-inoculation samples in the treatment group (Fig. 6e, f and Supplementary Fig. 9). Uptake correlated with cell shedding and cluster grade but not with the tumor grade or stage (Supplementary Fig. 9).

The response to alpha1–oleate was further evaluated by RNA-seq, using RNA from tumor biopsies and comparing the treatment to the placebo group. A strong treatment effect was detected (Fig. 7a–c and Supplementary Fig. 10). Cancer-related genes accounted for about 80% of the significantly regulated genes in the treated patients (cut off fold change > 2.0, P < 0.05), confirming the effect of alpha1–oleate on the tumor environment. Genes regulating tumor growth and invasion were inhibited and Ras signaling was suppressed, consistent with known effects of the complex on Ras family members24 (Fig. 7d and Supplementary Fig. 10). Bladder cancer genes were specifically regulated, including metalloproteinases, solute carriers, WNT complex constituents, and thrombospondin, which affects angiogenesis25(Fig. 7e). Furthermore, Fatty Acid Desaturase 6 (FADS6) and transcriptional activator CREB3L4 were affected, suggesting that the tumors respond to the constituents of the alpha1–oleate complex. FADS6 regulates oleate biosynthesis and CREB3L4 the unfolded protein response to conformationally fluid proteins, such as the alpha1 peptide. Interesting targets also included the gap junction alpha1 protein, which was inhibited, potentially promoting cell detachment (GJA1/CXA1 encoding Connexin 43, Supplementary Fig. 10). No difference in tumor grade was observed between the treatment or placebo groups, defined by WHO 1973 and 2004/2016 criteria (Supplementary Table 9). Data regarding two secondary end-points are not reported. As this is an interim analysis, the long-term treatment effects will be evaluated when the entire study has been completed. The urine proteomics data set has not been fully analyzed.

Fig. 7: Reprogramming of gene expression.

RNA sequencing was used to compare gene expression profiles in tumor tissue biopsies from the treatment or placebo groups. a Pie chart of genes regulated in response to treatment (cut-off FC > 1.5, P < 0.05 compared to the placebo group). In the treatment group, 82% of all regulated genes were cancer-related and 14% were bladder cancer-related. Gene categories were identified by biofunction analysis. b Heatmap of specific cancer- and bladder cancer-related genes regulated in tumor biopsies from the treatment group (red = upregulated, blue = downregulated, cut-off FC > 1.5, P <0.05 compared to placebo group). About 60% of all regulated genes were inhibited in the treatment group. c Detailed analysis of data in (a, b). Top regulated, cancer-associated functions are shown. Inhibition is indicated by negative z-scores (blue) and significance by P values (orange). The expression of genes involved in tumor invasion, neoplasia, tumor growth, and urinary tract tumors was strongly inhibited. d Inhibition of Ras signaling in the treatment group compared to placebo. e Bladder cancer gene network regulated specifically in patients receiving alpha1–oleate treatment compared to placebo.


Bladder cancer is the fourth most common malignancy in the United States and the fifth in Europe, with a prevalence of about 1/400026. Due to high recurrence rates and a lack of curative therapies, “bladder cancer has the highest lifetime treatment costs per patient of all cancers, followed by colorectal-, breast-, prostate-, and lung cancer”27. More than 80% recur after complete surgical removal of the first tumor and 15% progress to muscle-invasive disease28. Intravesical chemotherapy and Bacillus Calmette–Guérin (BCG) immunotherapy have limited efficacy and significant side effects29,30. Systemic administration of PD-1 and PD-L1 inhibitors is considered only in BCG unresponsive patients where the experience is limited. Therapeutic options are also limited by the inadequate supply of immunotherapy and chemotherapy drugs worldwide31. In this study, we identify conformationally fluid peptide–fatty acid complexes as additional tools in cancer therapy and show that intra-vesical inoculation of alpha1–oleate is safe and effective in patients with bladder cancer.

The tumor response to alpha1–oleate was analyzed in-depth, using cellular and molecular tools to detect changes induced by the complex. Treatment triggered the shedding of cells and tissue fragments into the urine and alpha1–oleate internalization by tumor cells confirmed the affinity of the complex for the tumor. Further analysis of tissue biopsies suggested a lasting effect of the alpha1–oleate instillations, as several tumor samples showed a gradient-like pattern of apoptosis, starting from the bladder lumen. Dysfunctional apoptosis has been identified as a key to tumor development, especially in environments where oncogenes such as MYC drive tumor cell proliferation32. Numerous attempts have been made to develop apoptosis-inducing therapeutics with tumor selectivity, but this has proven challenging, probably due to the heterogeneity of individual tumors as well as their intrinsic resistance to activating cell death pathways. The ability of alpha1–oleate to stimulate apoptosis in the majority of bladder tumors is, therefore, encouraging and consistent with the apparent lack of toxicity for bladder tissue.

RNA sequencing revealed profound molecular changes in treated tissues, attributable to alpha1–oleate. Classical cancer gene networks were strongly inhibited in the treated patients, compared to the placebo group, including Ras, previously identified as a target for HAMLET; the oleate complex formed by the alpha-lactalbumin holoprotein24. HAMLET binds activated Ras at the plasma membrane of tumor cells and inhibits the Ras signaling pathway, in part through effects on b-Raf phosphorylation. Significant effects on adaptive immunity were not detected and innate immunity was largely inhibited, including granulocyte activation pathways. Notably, genes involved in oleate metabolism and the unfolded protein response were affected, possibly reflecting a direct response to the constituents of the alpha1–oleate complex. In addition, treatment inhibited GJA1, a gap junction protein that has been proposed to promote cancer development and metastasis33,34. The effect occurred specifically in tumor tissue, potentially providing the alpha1–oleate complex with a mechanism to trigger cell shedding, as observed here. It is interesting to speculate that cell shedding may serve as a “tip-of the iceberg” marker of the profound changes in tumor biology that include activation of programmed cell death, transcriptional reprogramming, and inhibition of tumor progression.

Alternative therapeutic tools are actively being developed and tested in patients with NMIBC, particularly in patients with disease recurrence after BCG treatment35,36,37. Device-assisted hyperthermia was shown to increase the efficacy of intra-vesical chemotherapy but treatment was accompanied by side effects, reducing compliance38,39,40. An oncolytic-virus-based intra-vesical therapy was recently reported to achieve a complete response in 53.4% of patients with BCG-unresponsive carcinoma in situ, in a phase III trial41. The authors discuss the assessment of side effects and the development of biomarkers to help select patients suitable for this therapy. In patients with BCG unresponsive disease, treated with systemic Pembrolizumab, a 41% response rate was reported but side effects were prevalent, limiting compliance (Keynote-676 trial35). The present study identifies alpha1–oleate as an active drug candidate with low toxicity. Further dose-finding clinical studies and adjuvant therapy protocols will be essential to define the therapeutic window of this complex.

Cancer cells are aggressive, outcompete healthy cells, and ruin tissue integrity. It is generally assumed that treatments must be equally aggressive and highly toxic substances are often used, despite their lack of selectivity and the severe side effects that they cause. The protein–lipid complexes studied here are attractive to cancer cells, which actively internalized them, but end up being killed. Healthy cells are less responsive and extensive toxicity studies have failed to detect adverse effects in the bladder13. This low toxicity was confirmed here, as no drug-related side effects were observed in the treatment group. The results, therefore, identify the alpha1–oleate treatment of NMIBC as an interesting therapeutic option. In view of the low toxicity observed so far, liberal intra-vesical administration in early-stage NMIBC might be an interesting approach to postponing the introduction of more toxic and invasive therapeutic options.

Article classification: Biological abstract
Share to:
Add:Room A11-329, 1st Floor, No.1, SBI Venture Street, Optics Valley, East Lake
New Technology Development Zone, Wuhan, China.
Certificate NO.:U18Q28010569R0S