Doxorubicin Conjugated to Immunomodulatory Anticancer Lactoferrin Displays Improved Cytotoxicity Overcoming Prostate Cancer Chemo resistance and Inhibits Tumour Development in TRAMP Mice

Advanced, metastatic, castration resistant and chemo-resistant prostate cancer has triggered change in the drug development landscape against prostate cancer. Bovine lactoferrin (bLf) is currently attracting attention in clinics for its anti-cancer properties and proven safety profile. bLf internalises into cancer cells via receptor mediated endocytosis, boosts immunity and complements chemotherapy. We employed bLf as an excellent functional carrier protein for delivering doxorubicin (Dox) into DU145 cells, CD44+/EpCAM+ double positive enriched DU145 3D prostaspheres and drug resistant ADR1000-DU145 cells, thus circumventing Dox efflux, to overcome chemo-resistance. Successful bLf-Dox conjugation with iron free or iron saturated bLf forms did not affect the integrity and functionality of bLf and Dox. bLf-Dox internalised into DU145 cells within 6 h, enhanced nuclear Dox retention up to 24 h, and proved significantly effective (p < 0.001) in reducing LC50 value of Dox from 5.3 μM to 1.3 μM (4 fold). Orally fed iron saturated bLf-Dox inhibited tumour development, prolonged survival, reduced Dox induced general toxicity, cardiotoxicity, neurotoxicity in TRAMP mice and upregulated serum levels of anti-cancer molecules TNF-α, IFN-γ, CCL4 and CCL17. The study identifies promising potential of a novel and safer bLf-Dox conjugate containing a conventional cytotoxic drug along with bLf protein to target drug resistance.


Sorting of cells for cancer stem cells
Isolation and enrichment of stem cell populations in advanced prostate cancer cell line of DU145 was carried out using magnetic bead based separation technique. In this technique, cancer stem cells were specifically labelled for the expression of stem cell markers such as EpCAM and CD44 using antibodies linked to magnetic beads and were separated using the magnetic columns (MACS® Miltenyi Biotec, Germany) according to manufacturer's instruction. CD44+ and EpCAM+ double positive DU145 cells were thus obtained by enriching the cells with EpCAM positive cells using first round of separation which were then re-enriched for CD44 positive population. The sorted cells were analysed by flow cytometry for expression of respective stem cell markers.

Cell lysis and isolation of total protein for Western blot analysis of protein expression changes
DU145 cells were plated out in a 6 well plate at a density of 2x10 5 cells/mL and incubated at 37 °C, 5% CO 2 and were allowed to grow overnight. Once confluent, the cells were treated with different concentrations of Dox alone or the bLf-Dox conjugates. The cells were then incubated with the treatments for 24 h at 37 °C at 5% CO 2 . After incubation with the treatments, the supernatant was discarded and the cell layer was washed with PBS. 100 μL of RIPA buffer containing 1x protease inhibitor was added to each well and the plate was placed on ice for 20 min 68 . The cell layer was scraped using a sterile cell scrapper and collected in microcentrifuge tubes. The lysates were first sonicated at 75% amplitude for 30 s in ice to dislodge any membrane bound proteins and were centrifuged at 13,000 rpm for 20 min at 4 °C. The supernatants were collected after centrifugation and stored in a fresh microcentrifuge tube. Samples were then measured for their protein concentration using Bradford's reagent. 100 µg of total cell lysate was loaded for each treatment and Western blotting was carried out using standard protocol with appropriate primary and secondary antibodies.

Western Blotting
The transfer of proteins from the poly acrylamide gel to a PVDF membrane (GE Biosciences, Australia) was carried out on a Trans-Blot Turbo semi dry transfer system (Bio-Rad, Australia). After the blocking process, the membrane was incubated with the primary antibodies at a 1:1000 dilution in TBS for 1 h at 37 °C. Following incubation, the membrane was washed with TBS-T and incubated with respective HRP conjugated secondary antibody (Sigma-Aldrich, Australia) at 1:10,000 for 1 h at 37 °C. The washes were repeated and finally the membrane was developed using HRP substrate based ECL Chemi-luminescence reagents (GE Biosciences, Australia); the membrane was imaged under dark using Chemi-Doc XRS+.

TUNEL assay
DNA fragmentation post treatments were evaluated using Terminal Deoxynucleotidyl Transferase dUTP Nick End Labelling [TUNEL] kit. 1X10 4 cells were seeded on an 8 well slide and allowed to grow until they attained 90% of confluency and cells were treated accordingly. Untreated cells were taken as controls along with a negative control containing only the fluorescent nucleotides without the enzyme. After treatment, cells were washed with PBS and permeabilised with 0.1% Triton X-100 and resuspended in 50 μL of TUNEL reaction (Containing the transferase enzyme and the FITC labelled nucleotides) mixture. The cells were incubated for about 60 min at 37 °C in dark. The slides were then counter-stained for the nucleus and imaged under the confocal microscope.

Clonogenic assay
ADR1000-DU145 cells were plated out in a 6 well plate and treated with appropriate treatments for 24 h. After this, the treatment was removed and the cells were trypsinised. About 200 cells were counted from each treatment and plated out as single cell suspension in a 6 well plate. The cells were allowed to grow for 10-14 days to allow them to form colonies. After that, the cells were fixed using paraformaldehyde and stained with 1% crystal violet to visualise the colonies which were then counted manually using an illuminated colony counter. Representative photographs were taken for each treatment.

Migration assay
Transwell membrane Boyden chambers (Thin-certs) with a pore size of 8 μm were placed in a 24 well plate and 1mL of 10 5 cells (ADR1000-DU145) were plated out in the top chambers with 2 mL of media in the bottom chamber 66 . To the wells, media with varying treatment concentrations were added including one well as a control with media alone. Cells were incubated with treatments for 24 h at 37 °C with 5% CO 2 . Cells were then fixed with 4% PF for 20 min at room temperature following which they were washed with PBS and then stained with 0.1% crystal violet. The non-migrated top layer of cells was removed with the help of a cotton. The cells were then viewed, counted and imaged using an inverted microscope.

Immunohistochemical analysis
Following dissection, brain and small intestine from all groups were fixed in 4% freshly prepared paraformaldehyde (in PBS pH 7.4) for 18 h. Samples were then washed, dehydrated, and embedded in paraffin using Leica Paraffin Embedding Sample Prep Processor and Paraffin Embedding Station.
Sections of 7μm thickness were stained either with Hematoxylin and Eosin (H&E) for histological observation. Demonstration of CCL11 immunostaining in brain and intestine sections of normal and treated mice was carried out by deparaffinizing the sections and incubation with 5µg/mL of goat polyclonal anti-mouse CCL11 antibody from R&D systems (AF-420-NA), for one hour at 37 0 C. The primary antibody binding was detected using R&D Systems Anti-Goat HRP-DAB Cell & Tissue Staining Kit protocol as per manufacturer's instructions. The immunoreactivity in terms of color intensity detected with (3, 3-diaminobenzidine tetrahydrochloride DAB) was visualized and imaged with the Leica bright field microscope. A total of 10 microscopic fields for each section were visualized for immunopositive zones of the tissue.

Fe-bLf-Dox conjugates were less cytotoxic to non-cancerous cells compared to the Dox alone treatment
In comparison to the cytotoxic Dox alone treatment, the Apo-bLf-Dox and Fe-bLf-Dox showed lower toxicity (P<0.001) in the lower concentration range of 1.5 µM and 3 µM. At 3 µM concentration Apo-bLf-Dox induced a 43.8% cytotoxicity which was significantly higher (P<0.001) than the 14.2% cytotoxicity induced by Fe-bLf-Dox at the same concentration, whereas it was much lower than the 81.5% cytotoxicity induced by Dox alone at the 3 µM concentration (Supplementary Information). At the high concentration of 6 µM there was no significant difference between the cytotoxicity of Dox alone and Apo-bLf-Dox, but, the cytotoxicityof Fe-bLf-Dox (55.32%) was significantly lesser than either of Apo-bLf-Dox and Dox alone treatments (P<0.001).  formed per field of analysis are given in the form of a histogram. Treatments were carried out in triplicates and the assay was performed thrice independently. Data has been expressed as mean ± SD. One-way ANOVA was performed to evaluate statistical significance followed by post-hoc analysis by Tukey's test. C.) The transwell chamber migration assay was used to determine the ability of DU145 CD44+/EpCAM+ double positive cells to migrate through an 8 micron transwell thincerts following exposure to treatments. The migrated cells were then stained with 0.1% cyrstal violet and imaged. D.) Migrated cells were manually counted under a microscope and the numbers of colonies formed per field of analysis are given in the form of a histogram. Treatments were carried out in triplicates and the assay was performed thrice independently. Data has been expressed as mean ± SD. One-way ANOVA was performed to evaluate statistical significance followed by post-hoc analysis by Tukey's test. E.) DU145 CD44+/EpCAM+ double positive cells were allowed to form prostaspheres for 7 days. Following spheroid formation, they were treated once at 0 h and again at 48 h and the spheroids were imaged under the microscope. F.) The size (Diameter) of the tumour spheroid was measured at 24 h, 48 h and 96 h post first treatment was measured using Image J (NIH) software tool. The fold change in the reduction of spheroid diameter was calculated and represented as histogram. The experiment was carried out thrice with 5 spheroids per treatment. Two-way ANOVA was performed to evaluate statistical significance followed by post-hoc analysis by Tukey's test. G.) Trypan blue analysis was performed post 96 h treatment to analyse the percentage viable cells left in the spheroids.

Dox induces neurotoxicity, upregulates CCL11 in mice brain while Fe-bLf-Dox conjugate is neuroprotective and inhibits CCL11 expression
We performed immunohistochemical analysis to determine, if the significantly high serum presence of CCL11 could have induced adverse actions in the brain tissue of Fe-bLf-Dox treated mice. The elevated circulating CCL11 can contribute to cognitive decline in aging by directly inhibiting adult hippocampal neurogenesis 1 . The mechanism by which circulating CCL11 induce neurodegeneration is not clear.
Recently, a study has reported that CCL11 crosses the BBB (blood brain barrier), its transport occurs throughout the brain with varied transport rates among brain regions, and transport is affected by interactions with the cellular components of blood. The authors suggests that circulating CCL11 levels can be regulated by the BBB and could therefore effect its actions in the CNS 2 .
On the other hand, Dox is known for its neurotoxicity and neuron damaging effects due to induction of oxidative stress and ROS generation 3 . As shown in Figure S.3 A and B, Dox treated mice brain tissue sections in addition to upregulated CCL11 Expression, showed eosinophil infiltration, and loss of neuronal tissue and architecture. Fe-bLf-Dox treatment (Figure S.3 D) restored very well the tissue architecture and no CCL11 expression was observed, though untreated mice brain tissue showed very low CCL11 expression.
The lack of CCL11 expression in Fe-bLf-Dox fed mice brain despite its significantly higher concentration in serum than the Dox (i.p.) treated animals, is intriguing. For its anti-inflammatory activity, lactoferrin (a glycosaminoglycan-binding molecule) has been shown to compete with cytokines and chemokines for its binding to proteoglycans. For example, it binds with a very high affinity than IL-8, to endothelial glycosaminoglycans involved in chemotaxis 4 . Thus it could be possible that the transport of serum CCL11 to BBB was either competitively inhibited by Fe-bLf-Dox conjugate through its binding to proteoglycans on BBB endothelia or Fe-bLf's binding to Lf receptors present on BBB and blood cells. As mentioned above, circulating CCL11 transport to brain regions is affected by interactions with the cellular components of blood 2 . An interesting future research can be considered based on our current findings.
The observations of well-preserved brain tissue architecture with Fe-bLf-Dox fed animals, could be explained in terms of the neuroprotective and antioxidant nature of bLf molecule which is gaining scientific attention, recently. We have reported that Fe-bLf involved both the PI3K and ERK signaling for inducing neuronal differentiation 5 , while bLf's neuroprotective function in preventing prion proteininduced cell death, as an antioxidant due to the scavenging of ROS has also been reported 6 . Human lactoferrin (hLf) protects vulnerable dopamine neurons from degeneration in Parkinson's disease by preserving calcium homeostasis in mitochondria 7 .Further the efficacy of (hLf) was comparable to that of glial cell line-derived neurotropic factor (a prototypical neurotrophic factor for dopamine neurons. The retained neuroprotective ability of Fe-bLf, while fed orally as a protein-drug conjugate to circumvent Dox induced neurotoxicity, is promising and warrants further research to delineate the exact mechanism which is beyond the scope of current study. A & B) CCL11 expression within doxorubicin (I.P.) treated mice brain tissue. Blue Star shows neuronal loss area and neurotoxicity of Dox. Blue star in A indicates loss of brain tissue architecture and neuronal loss region of brain. Red arrows in B show the eosinophil infiltration.

C)
Control group mice brain reveals patchy and very low CCL11 expression.

D)
Fe-bLf-Dox conjugate treated mice brain tissue revealed no CCL11 expression, and normal brain architecture was observed