Pro-Apoptotic and Immunotherapeutic Effects of Carbon Nanotubes Functionalized with Recombinant Human Surfactant Protein D on Leukemic Cells

Nanoparticles are efficient drug delivery vehicles for targeting specific organs as well as systemic therapy for a range of diseases, including cancer. However, their interaction with the immune system offers an intriguing challenge. Due to the unique physico-chemical properties, carbon nanotubes (CNTs) are considered as nanocarriers of considerable interest in cancer diagnosis and therapy. CNTs, as a promising nanomaterial, are capable of both detecting as well as delivering drugs or small therapeutic molecules to tumour cells. In this study, we coupled a recombinant fragment of human surfactant protein D (rfhSP-D) with carboxymethyl-cellulose (CMC) CNTs (CMC-CNT, 10–20 nm diameter) for augmenting their apoptotic and immunotherapeutic properties using two leukemic cell lines. The cell viability of AML14.3D10 or K562 cancer cell lines was reduced when cultured with CMC-mwCNT-coupled-rfhSP-D (CNT + rfhSP-D) at 24 h. Increased levels of caspase 3, 7 and cleaved caspase 9 in CNT + rfhSP-D treated AML14.3D10 and K562 cells suggested an involvement of an intrinsic pathway of apoptosis. CNT + rfhSP-D treated leukemic cells also showed higher mRNA expression of p53 and cell cycle inhibitors (p21 and p27). This suggested a likely reduction in cdc2-cyclin B1, causing G2/M cell cycle arrest and p53-dependent apoptosis in AML14.3D10 cells, while p53-independent mechanisms appeared to be in operation in K562 cells. We suggest that CNT + rfhSP-D has therapeutic potential in targeting leukemic cells, irrespective of their p53 status, and thus, it is worth setting up pre-clinical trials in animal models.


Introduction
The innate immune system plays a key role in the clearance of pathogens and synthetic compounds including nanoparticles [1,2]. Nanoparticles have numerous biomedical applications [3][4][5][6], which can serve as drug delivery carriers or vaccine adjuvants [7]. Among nanoparticles, carbon nanotubes (CNTs) have unique physico-chemical properties, and hence, they are amenable as therapeutic nanocarriers [8][9][10]. CNTs can be single-walled (SWCNTs) and multiple-walled (MWCNTs), depending on length, diameter, and structure, and the layers of single CNT the wall is composed of [11].
Human surfactant protein D (SP-D) is a humoral, pathogen pattern recognition molecule, which is found to be associated with pulmonary surfactant, as well as mucosal surfaces outside the lungs [12,13]. SP-D belongs to the collectin family, a collagen containing C-type (calcium-dependent) lectin [14]. The primary structure of SP-D comprises a cross-linking amino-terminal region, a triple-helical collagen region, a coiled-coil neck region, and a C-type lectin domain or carbohydrate recognition domain (CRD) as a trimeric unit [15,16]. SP-D can bind to various carbohydrate and/or charge patterns on the surface of pathogens and become involved in clearing them by recruiting phagocytic cells such as neutrophils and macrophages [15,16]. SP-D can also interact with a range of cancer cell lines (leukemic, lung, pancreatic, prostate, ovarian and breast). For example, a truncated form of recombinant human SP-D (rfhSP-D), composed of trimeric neck and C-type lectin domain, has been shown to interfere with tumour progression via apoptosis induction, invasion, and epithelial-to-mesenchymal transition [17][18][19][20][21][22]. These studies have thus suggested that SP-D has an immune surveillance role against tumors.
SP-D can associate with nanoparticles and modulate their uptake by macrophages [23,24]. SP-D can bind efficiently with oxidized (Ox) DWCNTs via their C-type lectin domain [2,25]. SP-D mediated enhancement of nanoparticle uptake by alveolar macrophages and dendritic cells in mice has been examined using polystyrene, carbon black and silica nanocarriers [23].
CNTs, when opsonized with rfhSP-D, can provoke a differential pro-inflammatory immune response [26]. Surface modifications of hydrophobic CNTs are used for their good dispersion via covalent or non-covalent surface coatings [27]. For instance, the dispersion of MWNTs via oxidation (Ox-CNT), or with carboxymethyl-cellulose (CMC-CNT), has been reported [27]. Soluble complement components, such as factor H and C1q, opsonize functionalized CNTs, suggesting that key innate immune molecules can bind CNTs and alter inflammatory response [27].
This study was aimed at examining the ability of CNT + rfhSP-D to induce apoptosis using an eosinophilic cell line, AML14.3D10 [28], and a chronic myelogenous leukemia cell line, K562, to assess if CNT + rfhSP-D nanomaterials are worth testing in animal models.

rfhSP-D Upregulates p53 Expression in AML14.3D10 Cell Line
p53, a transcription factor, regulates oncogenic responses including DNA damage, cell cycle arrest, and apoptosis. CNT + rfhSP-D or rfhSP-D alone treated AML14.3D10 cells showed increased transcript levels of p53 when compared to untreated cells. CNT + rfhSP-D treated cells showed log 10 8.2-fold increased mRNA levels, compared to rfhSP-D treated cells (approximately log 10 5.2-fold) ( Figure 7). p53 transcripts were not measured in K562 cells as these cells do not express wild type p53. These data suggest that CNT + rfhSP-D treatment can induce apoptosis in these cell lines regardless of their p53 status.

Apoptosis Induction in AML14.3D10 and K562 Cells by rfhSP-D-CNT via Intrinsic Pathway
Since apoptosis can be initiated via intrinsic or extrinsic pathways, expression of caspases was examined in AML14.3D10 or K562 cell lines treated with CNT + rfhSP-D (20 µg/mL) or rfhSP-D alone (20 µg/mL), using a fluorogenic substrate to detect the activation of caspase 3 and 7 ( Figure 8). Higher levels of caspase 3 and 7 were observed in CNT + rfhSP-D treated AML14.3D10 ( Figure 8A) and K562 ( Figure 8B) cells, when compared to rfhSP-D or CNT alone-treated cells. There was a time-dependent increase in caspase 3 and 7 activation, which peaked at 24 h. Cleaved caspase 9 level was observed in CNT + rfhSP-D (or rfhSP-D-treated) AML14.3D10 or K562 cells at 12h, reflecting an intrinsic pathway ( Figure 9).

Apoptosis Induction by CNT+rfhSP-D in AML14.3D10 and K562 Cell Lines
The quantitative analysis of apoptosis induction by CNT+rfhSP-D was perform using flow cytometry. A significant proportion of AML14.3D10 or K562 (

Up-Regulation of Cell-Cycle Inhibitors by CNT+rfhSP-D Treatment
To further understand the mechanism of apoptosis induced by CNT+rfhSP-D AML14.3D10 or K562 cells, we analysed the expression of cell cycle inhibitors by qR PCR. p21 was upregulated in CNT+rfhSP-D treated AML14.3D10 (log10 5.7-fold) and K (log10 2.7-fold) ( Figure 6) [compared to CNT alone: AML14.3D10 (log10 1.2-fold) and K (log10 1-fold)]. p27 transcripts were also upregulated in CNT+rfhSP-D challeng AML14.3D10 (log10 2.5-fold) and K562 (log10 2-fold) cells. The level of upregulation w considerably higher compared to CNT or rfhSP-D alone that were negative and posit controls, respectively, suggesting that coating rfhSP-D on CNTs enhanced rfhSP-D tency for targeting tumors.    (Figure 7). p53 transcripts were not measured in K5 cells as these cells do not express wild type p53. These data suggest that CNT+rfhSP treatment can induce apoptosis in these cell lines regardless of their p53 status.

Apoptosis Induction in AML14.3D10 and K562 Cells by rfhSP-D-CNT via Intrinsic Pathway
Since apoptosis can be initiated via intrinsic or extrinsic pathways, expression of caspases was examined in AML14.3D10 or K562 cell lines treated with CNT+rfhSP-D (20 μg/mL) or rfhSP-D alone (20 μg/mL), using a fluorogenic substrate to detect the activation of caspase 3 and 7 ( Figure 8). Higher levels of caspase 3 and 7 were observed in CNT+rfhSP-D treated AML14.3D10 ( Figure 8A) and K562 ( Figure 8B) cells, when compared to rfhSP-D or CNT alone-treated cells. There was a time-dependent increase in caspase 3 and 7 activation, which peaked at 24 h. Cleaved caspase 9 level was observed in CNT+rfhSP-D (or rfhSP-D-treated) AML14.3D10 or K562 cells at 12h, reflecting an intrinsic pathway (Figure 9).

Discussion
The involvement of innate immune mechanisms in cancer progression and resistance has opened up opportunities for using innate immune molecules as a part of anti-tumour therapeutic strategies. Immune system, innate as well as adaptive, is a double-edged sword that can either foster tumour progression via immunosuppression, angiogenesis, and metastasis, or resist oncogenesis [29,30]. SP-D, especially the trimeric CRDs in its recombinant form (rfhSP-D), has recently been shown to be protective against a range of cancer, based on in vitro studies. Coupling rfhSP-D with nanoparticles triggers a differential immune response [26]. rfhSP-D-bound CNTs upregulate the pro-inflammatory response (IL-1β, TNF-α, IL-6 and IL-12) in U937 and THP-1 cells [26]. Here, we examined the ability of CNT+rfhSP-D to act as a potent inducer of apoptosis in leukemic AML14.3D10 or K562 cell lines. CNT+rfhSP-D treatment reduced the cell viability of

Discussion
The involvement of innate immune mechanisms in cancer progression and resistance has opened up opportunities for using innate immune molecules as a part of anti-tumour therapeutic strategies. Immune system, innate as well as adaptive, is a double-edged sword that can either foster tumour progression via immunosuppression, angiogenesis, and metastasis, or resist oncogenesis [29,30]. SP-D, especially the trimeric CRDs in its recombinant form (rfhSP-D), has recently been shown to be protective against a range of cancer, based on in vitro studies. Coupling rfhSP-D with nanoparticles triggers a differential immune response [26]. rfhSP-D-bound CNTs upregulate the pro-inflammatory response (IL-1β, TNF-α, IL-6 and IL-12) in U937 and THP-1 cells [26]. Here, we examined the ability of CNT + rfhSP-D to act as a potent inducer of apoptosis in leukemic AML14.3D10 or K562 cell lines. CNT + rfhSP-D treatment reduced the cell viability of AML14.3D10 and K562 cells and induced apoptosis at 24 h in a dose-and time-dependent manner, peaking at 24 h and 20 µg/mL. A significant reduction in viability was observed in CNT + rfhSP-D treated AML14.3D10 (~37%) and K562 (~55%) cells compared to untreated control (cells + CNT), based on trypan blue and MTT assays.
FACS analysis revealed a significant increase in the percentage of Annexin V-/PIpositive leukemic cells following CNT + rfhSP-D treatment, characterized by the disruption of the asymmetric arrangement of the membrane, and appearance of PS on the outer side of the cell membrane in the cells undergoing apoptosis [31]. Annexin V, a 36 kDa protein, can bind PS, and also enter the entire plasma membrane in necrotic cells. CNT + rfhSP-D triggered the maximum apoptosis at 24 h [AML14.3D10 (~71%) and K562 (~66%)], when compared to CNT alone [AML14.3D10 (~12%) and K562 (~7%)]. However, no significant difference in terms of cell viability reduction/apoptosis induction was noticed following rfhSP-D treatment at 48 h in AML14.3D10 and K562 cells, suggesting recovery of the cells after 24 h. Apoptosis induction in AML14.3D10 and K562 cell lines by CNT + rfhSP-D may occur through the intrinsic pathway, supported by increased levels of caspase 3, 7 and cleaved caspase 9. This validates earlier studies on AML14.3D10, prostate and breast cancer cells [20,21,32], and the involvement of a mitochondrial pathway [20,21,32].
We also tried to understand the underlying mechanism of apoptosis induction by CNT + rfhSP-D and the associated signaling pathways. CNT + rfhSP-D caused increased transcript level of p53 in AML14.3D10 cell line, probably due to oxidative stress [17,33]. The upregulation of p53 in CNT + rfhSP-D treated AML14.3D10 cells may downregulate pAkt pathway, increasing Bad and Bax, which in turn, causes the release of the cytochrome c, and caspase 9 cleavage. In addition, the increased expression of p53 and cell cycle inhibitors (p21/p27) can cause inactivation of the cyclin B-cdc2 complex, crucial for G2/M cell cycle transition [17]. The existence of a lack of p53 wild type gene in K562 cell line, and its increased susceptibility to CNT + rfhSP-D, the protective effects of rfhSP-D bound to CNTs seem p53 independent. An involvement of cellular receptors expressed by these cancer cell lines is of paramount importance. SP-D interaction with HMGA1, CD14, CD91-calreticulin complex, SIRPα, EGFR, and GRP78 has been reported [20][21][22]33,34]. The presence of rfhSP-D on CNT as an array of therapeutic molecule is likely to have a clustering effect on these putative receptors, enhancing the potency of rfhSP-D.
In conclusion, CNT + rfhSP-D nanomaterial seems to be an attractive and novel therapeutic approach for targeting intracellular signaling cascades. There is a clear therapeutic potential of rfhSP-D against tumour cells. The advantage here is that the enhanced glycosylation of oncogenic targets can evade natural or therapeutic antibodies. Having established the specific nature of interactions between CNT + rfhSP-D and receptors found on leukemic cancer cells, we can hope to investigate host response in the murine models of cancer using wild type and SP-D knock-out mice.

Expression and Purification of rfhSP-D
A recombinant fragment of human SP-D (rfhSP-D) was expressed and purified as described previously [17,32]. Affinity purified rfhSP-D was then subjected to endotoxin level measurement using QCL-1000 Limulus amebocyte lysate system (Lonza, Slough, UK); the endotoxin levels were found to be~5 pg/µg of rfhSP-D.

Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE)
The binding of rfhSP-D to CMC-CNTs was assessed via SDS-PAGE (12% v/v). CNT + rfhSP-D samples were boiled in a treatment buffer containing SDS and β-mercaptoethanol at 95 • C for 10 min before loading on to the gel. The SDS-PAGE gel was stained for 2 h using brilliant blue stain containing methanol (50% v/v) and acetic acid (10% v/v). This followed submersion of the stained gel with gentle shaking with de-staining solution (staining solution without brilliant blue).

Trypan-Blue-Dye Exclusion Assay
AML14.3D10 or K562 cells (0.1 × 10 6 ) were seeded in a 12-well plate in complete RPMI complete medium overnight under 5% CO 2 at 37 • C. Next, the cells were washed with PBS and treated with CNT + rfhSP-D (5, 10 or 20 µg/mL), or rfhSP-D alone (20 µg/mL), in serum-free RPMI for 24 h. Cells + CNT and Staurosporine (1 µM/mL) were used as an untreated/negative and positive control, respectively. Cells were then washed, detached using 5 mM EDTA, and centrifuged (1200× g). The cell pellet, re-suspended in RPMI, was treated with Trypan blue (10 µL) (60%), and viable cells were counted using hemocytometer in 5 different optical fields with a threshold value of 200 cells per field.

Flow Cytometry
For apoptosis assays, AML14.3D10 or K562 cells (0.4 × 10 6 ) were seeded in culture petri dishes (Nunc) in complete RPMI medium for 24 h and treated with CNT + rfhSP-D (20 µg/mL), or rfhSP-D (20 µg/mL), in serum-free RPMI medium for 24 h. Other controls were used as described above. Detached, centrifuged and PBS washed cells were incubated with Alexa Fluor 488 (1:200) (Sigma-Aldrich/Merck, Dorset, UK) (15 • , RT) in dark, and the extent of apoptosis was measured using Novocyte Flow Cytometer. Compensation parameters were acquired using unstained, untreated FITC stained, and untreated PIstained samples for all the cell lines.
For proliferative studies, AML14.3D10 or K562 cells (0.4 × 10 6 ) were washed with PBS, probed with anti-mouse Ki-67 (BioLegend, San Diego, CA, USA) diluted in permeabilization reagent of the FIX&PERM kit (Fisher Scientific), and incubated for 30 min at room temperature (RT). Goat anti-mouse-FITC conjugate (1:200) (Fisher Scientific) was used as a probe at RT in the dark for 30 min. Cells (12,000) were acquired for each experiment and compensated before plotting the acquired data.

Target Gene
Forward Primer Reverse Primer

Statistical Analysis
The graphs were generated using the GraphPad Prism 6.0 software. A one-way ANOVA test was carried out for statistical significance analysis. values less than 0.05 were considered as statistically significant.