Multimodal evaluation of mannose engineered poly lactic glycolic acid nanoparticles with granulocyte colony stimulating factor focused delivery to bone marrow for neutropenia

Abstract

Neutropenia, characterized by a critical reduction in neutrophils, demands targeted therapeutic strategies to enhance the delivery efficiency of granulocyte colony-stimulating factor (G-CSF) specifically to bone marrow macrophages. This study focused on engineering mannose-modified poly(D, L-lactide-co-glycolide) (PLGA) nanoparticles (NPs) to achieve ligand-directed delivery of G-CSF. Mannose anchoring was achieved via Ethylenediamine (EDA)-mediated chemical ligation using N-hydroxysulfosuccinimide (NHS) and dicyclocarbodiimide as coupling agents, resulting in Mn-EDA-PLGA NPs. G-CSF-loaded and placebo NPs were fabricated through a multiple emulsion solvent evaporation method and subjected to comprehensive physicochemical characterization. The developed placebo Mn-EDA-PLGA NPs measured 199 ± 12 nm, while G-CSF-loaded Mn-EDA-PLGA NPs measured 153 ± 12.2 nm, both exhibiting a negative surface charge of − 40.07 ± 1.1 mV and − 34.9 ± 1.9 mV, respectively. Polydispersity index values were low (0.34 and 0.41), indicating uniform particle distribution. Entrapment efficiencies were significant, with the optimized G-CSF-loaded formulation achieving 72.6% drug encapsulation efficiency and drug loadings of 5 µg and 3 µg for placebo and active NPs, respectively. Scanning electron microscopy confirmed spherical morphology with smooth surfaces. Biological evaluation using scintigraphy and flow cytometry in J774.2 macrophage cells validated the targeting efficiency of mannose-modified NPs. Furthermore, molecular docking and molecular dynamics simulations substantiated the stability and interaction profile of G-CSF within the nanocarrier system. The convergence of in vitro, in vivo, and silico findings underscores the potential of Mn-EDA-PLGA NPs as a robust delivery vehicle for G-CSF, offering enhanced bone marrow macrophage targeting. This targeted approach holds promise for improving therapeutic outcomes in neutropenia by maximising drug localization and minimising systemic exposure.

Introduction

Neutropenia is a debilitating condition characterised by a significant decline in the count of neutrophils, a type of white blood cell crucial for immune defence. Specifically, neutropenia is diagnosed when the neutrophil count falls below 500 neutrophils/mcL or when it decreases to less than 1000 neutrophils/mcL with a subsequent decline to 500/mcL or lower within the following 48 h, resulting in febrile neutropenia. Febrile neutropenia is accompanied by a temperature increase equal to or greater than 38.3 °C orally or ≥ 38.0 °C within one hour¹. The underlying causes of neutropenia can be multifactorial. One common cause is the administration of myelosuppressive drugs, which inhibit the production of neutrophils in the bone marrow. Additionally, infections, whether bacterial, viral, or fungal, can contribute to the development of neutropenia. Furthermore, environmental factors, such as exposure to radiation or toxins, can also lead to neutropenic conditions.

Cytotoxic cancer chemotherapy is a significant factor that leads to a decline in neutrophil count, thereby hindering the effectiveness of the treatment regimen^2,3. Patients with additional risk factors such as HIV, viral or bacterial infections often experience neutropenia, which worsens when combined with high doses of chemotherapy⁴. Neutropenia presents significant challenges as it compromises the immune system’s ability to combat infections effectively. Patients with neutropenia are more susceptible to severe and life-threatening diseases. Therefore, managing neutropenia and its associated complications is of utmost importance in clinical practice. Treatment and disease-related neutropenia are common and have various adverse clinical effects, including hospitalization due to infection-related mortality and morbidity, febrile neutropenia, and reduced ability of patients to receive scheduled chemotherapeutic doses⁵. Delays in therapy or reductions in chemotherapeutic dosage can undermine the medication’s efficacy. These problems significantly impact patients’ quality of life and clinical outcomes.

Granulocyte colony-stimulating factor (G-CSF), or Filgrastim, is a glycoprotein that stimulates neutrophil progenitor cell production, proliferation, and differentiation⁶. G-CSF is a hematopoietic growth factor capable of reducing the prevalence and severity of neutropenia, febrile neutropenia, and associated complications such as cancer or HIV infection⁷. Another option, Pegfilgrastim, a longer-acting form of filgrastim, is used to mitigate neutropenic conditions during chemotherapeutic cycles⁸. Selective use of these colony-stimulating factors has proven effective in controlling neutropenic conditions, although it increases the cost of chemotherapy treatment. Several research studies have attempted to identify the risk factors for neutropenia and its related consequences. These studies have explored using myeloid factors, such as colony-stimulating factors, to counteract neutropenia, identify patients at risk for complications, and guide them towards more cost-effective applications⁹. G-CSF and its analogues have effectively alleviated chemotherapy-induced neutropenia, promoted patient compliance, and improved quality of life. However, challenges remain, including degradation in the gastrointestinal mucosa, for which injectables are available but ineffective for oral administration. Additionally, G-CSF is expensive in therapeutic quantities ($37.76 for 300 µg per chemotherapeutic cycle), which increases the overall cost of chemotherapy treatment. Moreover, overexposure of G-CSF to peripheral tissue can lead to adverse effects such as bone and joint pain¹⁰. The development of improved drug delivery systems and technologies, such as nanotechnology, has gained widespread interest in enhancing the efficacy and safety of formulations¹¹. Within nanotechnology, targeted delivery systems offer a promising approach to augment the drug’s effectiveness at the affected site while reducing side effects on peripheral, non-targeted tissues¹². Several publications have discussed various nanocarriers, such as solid-lipid, polymeric, dendrimers, liposomes, and nanotubes, loaded with different drugs that claim to improve the absorption and permeation of drugs in the gastrointestinal tract¹³. Targeting the bone marrow with G-CSF may provide a suitable method for administering the drug at reduced doses and with fewer adverse effects¹⁴. Conventional delivery of G-CSF results in peripheral drug deposition, and selective targeted drug delivery to the bone marrow eliminates adverse effects and improves efficacy at lower doses. To our knowledge, no publications on targeted delivery of G-CSF to the bone marrow are currently available at low concentrations. The use of biopolymers in the design of advanced drug delivery systems has gained substantial attention due to their inherent biodegradability, biocompatibility, and potential to serve as bioactive carriers^15,16. This study used the modification of synthetic poly(D, L-lactide-co-glycolide) (PLGA) with mannose to facilitate macrophage-specific targeting and the polymer’s degradation behaviour and overall biological performance. PLGA, while widely accepted for controlled drug delivery, exhibits limitations such as acidic degradation products and inconsistent degradation rates^17,18. Functionalising PLGA with biologically derived ligands such as mannose improves its physiological compatibility and aligns its degradation profile more closely with in vivo therapeutic needs^19,20. Biopolymers such as alginate, chitosan, carboxymethyl cellulose, and lignin derivatives have shown promising potential in sustaining the release of both hydrophilic and hydrophobic drugs. These polymers can be tailored structurally to efficiently encapsulate therapeutic agents and release them in a controlled manner over extended durations^21,22. Importantly, they also offer favourable interactions with mucosal tissues, enhancing permeability and retention at target sites. Incorporating such natural moieties into synthetic carriers extends the versatility of nanoparticle-based drug delivery, allowing dual functionalities active targeting and prolonged release^23,24. Biopolymer based nanoparticles, such as polysaccharides and modified cellulose systems, are increasingly employed in medical applications owing to their biocompatibility and ability to sustain release of bioactives, reduce burst effect and promote tissue integration. These carriers enable controlled degradation and drug diffusion ensuring therapeutic levels are maintained while minimizing systemic exposure.

By introducing mannose residues onto PLGA via EDA-mediated conjugation, we aim to confer both targeting specificity and enhanced biodegradability to the delivery platform. This hybrid approach leverages the advantages of synthetic polymer robustness with the functional benefits of bioactive modifiers. Such a strategy improves pharmacokinetics and therapeutic index, opens avenues for tissue-responsive release, and reduces systemic toxicity. Therefore, integrating biopolymer principles into our nanoparticle formulation was a deliberate design strategy to enable efficient targeting and sustained therapeutic action of G-CSF in the bone marrow microenvironment.

Materials and methods

Methodology for in-silico studies

Data collection and preparation

Data collection was initiated by the crystal structure of CRD-4 Macrophage Mannose Receptor (PDBID: 1EGI) from the Protein Data Bank (PDB)^3,25,26. Along with preparing the protein with the Protein Preparation Workflow tool from Maestro (Schrödinger suite)^27,28, for the preparation chain A, solvents and other metals/ions were removed, and the longest chain B was then taken further for the studies, as both chains are dimers. After the problem with Cα (802), we removed the water beyond 3.0 Å. The G-CSF Filgrastim (PDBID: 1PGR)²⁹ is used to treat neutropenia also downloaded from PDB for preparation we removed repeated chains like- C, D, E, F, G, H; kept only Chain A and B for further analysis, filled or missing gaps and we removed water beyond 3.0 Å. As far as ligands are concerned, we took Ethylenediamine (CID15908), Mannose (CID18950), Poly (lactic-co-glycolic acid), or PLGA (CID23111554) from PubChem Databases and prepared them from LigPrep tool in Maestro (Schrödinger suite) for the molecular docking in the 3D SDF^27,30.

Molecular docking

A molecular docking study investigated if the ligand (PG) bound to the Macrophage Mannose Receptor (PDBID: 1EGI). To perform protein-peptide docking by the Protein-Protein Docking tool in Maestro (Schrödinger suite) and for analysis and visualization by the Protein Interaction Analysis tool for the Macrophage Mannose Receptor (PDBID: 1EGI) and the peptide Granulocyte colony-stimulating factor (G-CSF) Filgrastim (PDBID: 1PGR). Also, the Macrophage Mannose Receptor (PDBID: 1EGI) investigated the behaviour of other carriers’ Mannose, PLGA, and Ethylenediamine nanoparticles. We performed molecular docking by Virtual Screening Workflow (VSW) in Maestro, and the Receptor Grid Generation generated the glide grid³¹ for the Filgrastim (PDBID: 1PGR)^3,12. Visualization for Interaction and bonding analysis of the protein-ligand complexes is visualized in the Ligand Interaction Diagram tool³².

Molecular dynamic simulation studies

The behaviour of the protein-peptide complex under simulated physiological conditions was analyzed using molecular dynamics (MD) simulation in Desmond (Schrödinger Maestro v 2020-4). The Macrophage Mannose Receptor (PDBID: 1EGI and the peptide Granulocyte colony-stimulating factor (G-CSF) Filgrastim (PDBID: 1PGR) complex were solvated in a 10 × 10 × 10 Å SPC water box. To neutralise the system, specific counter ions were added 8Na+, the systems were energy minimised and position-restrained with the OPLS4 force field³³. A 100 ns MD simulation was then run at 1 atm and 300 K, using the NPT ensemble and recording frames every 100 ps, producing 1000 frames per complex. Key parameters like fluctuations, structural deviations, and stability were analyzed to assess the system’s compactness and stability throughout the simulation^34–37.

Materials and methodology for in-vitro studies

The materials utilised in this study were sourced from various suppliers. Fetal bovine serum, filgrastim (G-CSF), poly(D, L-lactide-co-glycolide) [PLGA acid-terminated RG 502 H, Mw 7000-17 000], and RPMI 1640 were obtained from Sigma-Aldrich, India. Mannose (MAN) and dimethyl sulfoxide (DMSO) were acquired from SRL Chemicals in Bangalore, India, and Loba Chemie in Mumbai, India, respectively. 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide was purchased from Sigma Aldrich in the United States. Dicyclocarbodiimide and N-hydroxysulfosuccinimide (NHS) were obtained from Rankem in Mumbai, India. Himedia Laboratories provided the fluorescein isothiocyanate 98% (FITC) required for this study. Polyvinyl alcohol (PVA, Mw about 125,000) was purchased from SDFCL in India, while Tween 80 was obtained from Merck, India. Sodium stibogluconate (SSG) was sourced from Albert David Ltd. in Kolkata, India. Dichloromethane (DCM) and dimethyl sulfoxide (DMSO) were procured from Merck, India. Well plates for uptake studies were purchased from Scientific Laboratory Supplies in Wilford, UK. All other chemicals used were of analytical grade. Purified water from a three-stage Millipore Milli-Q Plus 185 purification system in Bedford, MA, USA, was utilizcomplieded in all experiments.

Ethics statement for animal experiments

Animal experiments were performed on female Wistar rats weighing 150 and 180 g at the Animal Facility of ICMR-National Institute of Pathology in New Delhi (NIP/IAEC-1506). All the animals were obtained from the animal house facility of All India Institute of Medical Sciences, New Delhi. All animal experiments were strictly complied to the relevant laws established by the Committee for Control and Supervision of Experiments on Animals (CPCSEA), Government of India. The Institutional Animal Ethics Committee (IAEC) of the National Institute of Pathology (NIP/IAEC-1506) approved the experimental protocols, and all procedures complied with committee guidelines. The animals were housed in polypropylene cages in three groups to allow for acclimatization. They were provided with a standard diet and water ad libitum. Additionally, the rats were maintained under a 12 h light-dark cycle to simulate natural environmental conditions.

Fourier transform-infrared spectroscopy (FT-IR)

A mixture of these components was prepared as an excipient blend to evaluate the compatibility between the drug, and excipients. In order to determine the drug excipient compatibility, all the excipients and drug were mixed in equimolar quantities. The mixture was then carefully sealed in a glass vial and subjected to a temperature of 50 °C for 15 days. Fourier Transform-Infrared Spectroscopy (BRUKER- Tensor 27) was employed to examine the physical and chemical changes of the mixture on the fifteenth day.

Optimization, synthesis and preparation of G-CSF embedded PLGA nanoparticles

Mn-EDA-PLGA NPs were synthesized and prepared using a previously reported method with slight amendments³⁸ by controlling the dependent variables to optimize a suitable formulation. QBD (Quality-by-design) software was used to optimize the developed NPs.

Optimization of different LAS parameters by QBD

QBD software optimized the developed NPs by analysing and understanding the associations and interactions between various dependent and independent variables. A response surface methodology (RSM) design was employed to minimise the integrated anticipated variance across the design space. Eighteen distinct combinations were constructed to evaluate three independent variables at three levels (Table 1). The investigated variables included PLGA concentration (10–70 mg), PVA surfactant concentration (0.5–5% w/v), and Sonication time (2–15 min). Regression analysis of the data was performed using the DESIGN EXPERT software (version 9.0.5) provided by Stat-Ease Inc., Minneapolis, MN.

Table 1.

Significance of the model and insignificance of lack of fit by QBD.

Batch no	PLGA conc. (mg)	Surfactant (%)	Sonication time in min	Particle size (nm)	Entrapment efficiency
1	60	2.75	8.5	217	94.82
2	60	2.75	8.5	213.3	94.66
3	10	2.75	2	426	90.51
4	40	0.5	2	283.2	94.25
5	60	2.75	8.5	169	94.25
6	70	2.75	15	374.7	93.17
7	60	2.75	8.5	199	95.08
8	60	2.75	8.5	198.8	95.36
9	10	0.5	8.5	406	89.76
10	70	0.5	8.5	452.1	91.63
11	70	2.75	2	283.2	93.1
12	40	5	2	216.6	91.63
13	70	5	8.5	270.3	89.48
14	40	5	15	99.1	93.31
15	10	5	8.5	223.2	89.23
16	10	2.75	15	117.7	90.51
17	40	0.5	15	219.1	92.14

Bold values indicate statistically significant terms (p < 0.05).

Synthesis of placebo and G-CSF embedded Mn-EDA-PLGA NPs

Synthesis of placebo Mn-EDA-PLGA NPs

It involves two phases, i.e organic and aqueous phases. In the organic phase, 50 mg PLGA (poly (l-lactic-co-glycolic acid), 5 ml of ethyl acetate in a ratio (75:25) were used. The aqueous phase consisted of PVA (Polyvinyl alcohol) solution, and the concentration was optimized from 0.5 to 2%. Filtration was conducted through a 0.2 µm syringe filter. Finally, 1% PVA solution was optimized for an aqueous phase. The organic phase was sonicated, and the sonication time was optimized to start at 2–15 mins. Consequently, the optimized time of sonication was found to be 8.5 min. The organic phase, consisting of PLGA and ethyl acetate solution, was stirred on the magnetic stirrer for 1 h. Subsequently, the aqueous solution (1% PVA) was kept under a cold jacket and the organic phase (PLGA + ethyl acetate) was added dropwise into it and was kept for stirring on a magnetic stirrer for the next 3 h. This solution was further subjected to ultrasonication for 1 min at 232 volts with an on-off cycle of 5 sec³⁹. After sonication, the solution was stirred overnight on a magnetic stirrer to evaporate the organic solvent. The solution obtained was a clear solution with a bluish tint.

Synthesis of Mn-EDA-PLGA NPs

A weighed quantity of 60 mg of PLGA was dissolved in 15 ml of dichloromethane (DCM). DCC and NHS were incorporated into the solution to activate the free carboxylic group at the end of the PLGA terminal. The resulting precipitate, dicyclohexyl urea, was removed by filtration, and the excess DCC and NHS were separated using a dialysis membrane against Milli-Q water for 6 h. Next, 160 μL (0.003 μmol) of Ethylenediamine (EDA) was added to the solution, and the pH was adjusted to 5.0 using 1N HCl. The amine group attached to the PLGA terminal was then conjugated with mannose, following the mannosylation method previously reported by Mitchell et al. in¹³, with slight modifications. In this step, 2 mL of tetrahydrofuran and D-mannose were used to dissolve the amine-terminated PLGA (0.1 mmol) (8 mM in 0.1 M sodium acetate buffer at pH 4.0). The reaction was allowed to proceed by stirring the solution for two days at room temperature. After completion of the reaction, the final solution with a weight of 12 kDa/mol was placed in a dialysis bag and dialyzed against Milli-Q water for 24 h. The complete synthesis of the nanoparticles is presented in Fig. 1. To form FITC-anchored PLGA, 250 μl of a 1 mg/ml FITC solution in dimethylformamide was added to a 5 mg (0.37 μmol) solution of amine-terminated PLGA with continuous stirring at 600 rpm for 1 h at a temperature of 40 ± 10 °C⁴⁰. Further, Fig. 1 shows the chemical ligation of PLGA with Mannose.

Fig. 1.

Chemical ligation of PLGA with mannose.

Preparation of placebo and G-CSF embedded Mn-EDA-PLGA NPs

The placebo and G-CSF embedded Mn-EDA-PLGA NPs were prepared using the emulsion solvent evaporation and the o/w emulsification techniques. The emulsion was probe-sonicated for 8.5 min at 40W (optimized, Table 1), followed by solvent evaporation. Initially, a calculated amount of 60 mg PLGA was dissolved with and without G-CSF in a 2 mL mixture of DMSO and DCM (30:70% v/v), and the pH was adjusted to 3.0 using acetic acid. This mixture was vortexed and then emulsified with 25 ml of 2.75% w/v PVA solution using a probe sonicator (5 min at 40 W output). Organic solvent was evaporated for 6 h at room temperature over a magnetic stirrer. NPs were recovered by centrifugation at 40 °C at 44,250Xg rpm for 15 min. Similarly, the FITC-labelled NPs were prepared by replacing PLGA with 500 μg of FITC-PLGA conjugate. The remaining procedure followed was the same as for Mn-EDA-PLGA NPs. Following two MiliQ water washes, the untrapped G-CSF was taken out of the NPs using Amberlite resin XAD 16, and then the NPs were lyophilized. The dried NPs were kept in a refrigerator at 4℃⁴¹.

Characterisation and evaluation of placebo and G-CSF embedded Mn-EDA-PLGA NPs

Particle size, charge and PDI index

The particle size and PDI (Polydispersity Index) of the Mn-EDA-PLGA nanoparticles were determined using the Malvern Zetasizer, which utilises the principle of light scattering through laser diffraction at an 173° backscatter angle. The nanoparticles were suspended in 1.5 ml of an aqueous medium at a concentration of 0.3 mg/ml and placed in a corvette to perform this analysis. The suspension was maintained at a temperature of 25 ± 10 °C, and the viscosity and refractive index of the medium were adjusted to be similar to those of water. The surface charge of the nanoparticles was determined using the Malvern Zetasizer based on the principle of Laser Doppler Anemometry. The nanoparticles were suspended in a 1 mM HEPES buffer at a pH of 7.4, and the pH was adjusted using 0.1 M HCl to maintain a constant ionic strength, as described by Biswaro et al. in⁴².

Drug entrapment efficiency, yield and actual drug loading

To precipitate the polymer, 2 ml of methanol was added to the mixture of placebo and Mn-EDA-PLGA nanoparticles after dissolving them in 1 ml of ACN (Acetonitrile). The resulting sample was then centrifuged at 21,000 x g for 5 min. The estimation of the G-CSF drug was performed using an HPLC method with slight modifications. A C18 column (250 × 34.6 mm; 5 mm particle size) with a guard column (45 × 34.6 mm) was utilised. The mobile phase consisted of a mixture of ACN, 1% acetic acid, and water in a ratio of 41:43:16 (v/v), and the flow rate was set at 1.5 mL/min at room temperature. The UV detection wavelength was set at 405 nm, and the retention time of the drug was 4.3 min. The per cent drug entrapment efficiency (%EE) was calculated using the following equation:

1

After purification, the NPs solution was subjected to ultra-centrifugation at 4 ± 10 °C and 30,000 x g for 1 h. The supernatant was then separated, and the pellet was freeze-dried. The yield of the nanoparticles was calculated using the following equation:

2

The actual drug loading (% w/w) was determined using the equation:

3

Scanning electron microscopy

The shape and size of Mn-EDA-PLGA NPs were assessed by scanning electron microscopy (SEM). In SEM analysis, the aqueous dispersion was placed over a carbon-coated copper grid with a 400 mesh size, and afterwards, negative staining was done by 3% w/v phosphotungstic acid solution. pH was adjusted to 4.7 with the help of KOH and later placed at an accelerating voltage of 95kV⁴⁰. Mounted on an aluminium stub, a random sample of produced G-CSF Mn-NPs underwent gold-palladium alloy sputtering to reduce surface charge. Pictures were taken using a scanning electron microscope (Leo Electron Microscopy Ltd., Cambridge, UK) at a 15 kV accelerating voltage & 15 mm working distance. SEM was conducted to confirm the shape and surface morphology of the prepared optimized formulations.

Thermogravimetric analysis

Thermogravimetric analysis was employed to evaluate the thermal stability of placebo and G-CSF embedded Mn-EDA-PLGA NPs. Accurately weight quantity (5 mg) of each samples was heated in a nitrogen atmosphere from 25 °C to 600 °C at a constant heating rate of 10 °C/min. To determine the thermal degradation profile of the formulations, the % weight loss as a function of temperature was observed. To illustrate encapsulation and improved thermal stability of G-CSF-loaded nanoparticles, derivative thermogravimetric (DTG) curves were further investigated to determine distinct decomposition steps corresponding to loss of moisture, protein degradation, and polymer backbone hydrolysis.

In-vitro drug release study

The in-vitro drug release study was conducted using dialysis bag method and HPLC analysis. Initially, 10 mg of G-CSF-loaded nanoparticles (NPs) were suspended in 2 ml of phosphate-buffered saline (PBS) at pH 7.4, containing 5% v/v dimethyl sulfoxide (DMSO), and placed inside a dialysis bag (14,000 MWCO). The dialysis was performed against 50 ml of PBS: DMSO mixture (95:5% v/v) at a speed of 50 rpm. During dialysis, 500 μl samples were collected at regular intervals, and the withdrawn volume was replenished with the PBS: DMSO mixture to maintain sink conditions. These samples were then analyzed using HPLC to determine the amount of G-CSF released from the NPs. Release behaviour was also evaluated in a simulated macrophage environment using 0.2 M sodium acetate buffer at pH 5.5, representative of the endosomal compartment.

Flow cytometric analysis

To evaluate the uptake of nanoparticles (NPs) by bone marrow macrophages, the J774.2 macrophage cell line was employed. Since mannose receptors are abundant on bone marrow macrophages, the uptake was studied using flow cytometric analysis³³. The J774.2 macrophage cell line was cultured as an adherent culture in humidified Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum, 100 IU/ml penicillin, 100 IU/ml streptomycin, Glutamax-I, and 4.5 g/L glucose. The cells were maintained at 37 °C with 5% CO2. To initiate the experiments, the cells were harvested and adjusted to achieve the desired number of viable cells, which were counted using a hemocytometer. At regular time intervals (0.5, 1, 2, 4, 6, 12, and 24 h), the J774.2 cells were incubated with FITC-labelled NPs and subsequently examined. At the end of each incubation period, the cells were collected, and any excess NPs were washed away using ice-cold phosphate-buffered saline (PBS) containing 0.01% sodium azide and 5% fetal calf serum. The flow cytometer, equipped with an argon ion laser for excitation at a wavelength of 488 nm, was calibrated and adjusted using Calibrite beads and the FACSCOMP software (v.5.1) (https://www.bdbiosciences.com/en-dk), following the manufacturer’s recommendations. The cells were resuspended in a fluorescent-assisted cell sorting (FACS) buffer. The flow cytometer collected ten thousand events for each sample, and the cell-associated FITC fluorescence was measured using the Cell Quest Software.

In-vivo efficacy study (methotrexate induced neutropenia)

In this method, acclimatised Wistar rats were divided into four groups with N = 6/group and were maintained in standard laboratory conditions of temperature (25 ± 2⁰C °C), light (12 h light/12 h dark) and humidity. The scrutiny of animals was done by Group 1: Placebo NPs; Group 2: MTX-induced; Group 3: Mn-NPs; Group 4: G-CSF embedded Mn-GCSF-NPs (equivalent to 30.83 µg/kg of G-CSF). They were fed a standard rat pelleted diet and water ad libitum. After a week of acclimatisation, a trial using twice-weekly injections of 1 mg/kg/bw of methotrexate (i.p) dissolved in MilliQ water for 30 days was carried out. Blood was withdrawn before the start of the experimental protocol, and the level of neutrophils was recorded. At the end of the experimental period, the animals were killed by CO₂ inhalation. CO₂ was administered in a controlled environment. The animals were placed in a small chamber where the concentration of CO₂ was gradually increased, leading to rapid loss of consciousness followed by cessation of vital functions, ensuring minimal distress. Blood was withdrawn from the retro-orbital plexus and collected in vials. Blood was processed for RBC, WBC, platelet count, and haemoglobin level. In addition, neutrophil and eosinophil content was also determined⁴³, and the placebo group received blank nanoparticles.

In-vivo biodistribution of radiolabeled nanoformulation

The study assessed the distribution of radiolabeled nanoparticles into bone marrow. Radiolabeling was performed using ^99mTc as the radiolabeling material. Labelling efficiency and stability in blood serum were evaluated using chromatography and a scintillation well counter⁴⁴. Radiolabeled nanoformulation was administered into a suitable animal model, and the percentage distribution of the prepared nanoformulation was assessed using a rediometry at predetermined time intervals. The quantitative uptake of nanoparticles by animal organs such as liver, spleen and bone marrow was conducted to determine distribution kinetics. The major organ tissue samples were taken, weighed, and counted for radioactivity in the scintillation well counter to calculate the biodistribution after the animals were euthanised 24 h after the injection.

Statistical analysis

A one-way ANOVA test was conducted using the GraphPad Instat software for statistical analysis. The post-hoc test used was Tukey’s multiple comparison test with p < 0.005 as statistically significant.

Results

Interaction analysis

The Interaction between the protein and the peptide with the Macrophage Mannose Receptor (PDBID: 1EGI) and the Granulocyte colony-stimulating factor (Filgrastim) Peptide (PDBID: 1PGR) (Fig. 2) shows a PIPER pose score of − 393.373 kcal/mol, and Interactions are shown below (Table 2). Interaction of the Granulocyte colony-stimulating factor (Filgrastim) Peptide (PDBID: 1PGR) and the Mannose (PubChem ID: 18950) (Fig. 3A) produced docking score of − 5.66 kcal/mol and the MM/GBSA score of − 11.29 kcal/mol that interact with seven hydrogen bonds among LYS17, ALA47, TYR78, GLN121, and GLU124 residues along five OH atoms of the mannose. The mannose section of the Granulocyte colony-stimulating factor (Filgrastim) Peptide (PDBID: 1PGR) and the PLGA [Poly(D, L-lactic acid-co-glycolic acid)] (PubChem ID: 23111554) (Fig. 3B) produced a docking score of − 4.90 kcal/mol and the MM/GBSA score of − 3.52 kcal/mol, interacting with two hydrogen bonds along GLN121 and GLU124 residues with the OH atom; also, the LYS17 residue contacts with a salt bridge with the O atom of the PLGA. The Interaction between the Granulocyte colony-stimulating factor (Filgrastim) Peptide (PDBID: 1PGR) and the Ethylenediamine [1-ethyl-3-(3-dimethylaminopropyl)] (PubChem ID: 15908) (Fig. 3C) produced docking score of − 2.28 kcal/mol and the MM/GBSA score of − 2.92 kcal/mol that interact with a hydrogen bond along the HIS80 with the N⁺H atom, and three pi-cation bonds contact PHE84, and TRP119 residues with the N⁺H atom of the Ethylenediamine.

Fig. 2.

Interaction of macrophage mannose receptor and peptide Filgrastim.

Table 2.

Showing the interaction of macrophage mannose receptor and peptide Filgrastim.

Macrophage mannose receptor	Filgrastim	Bonds
GLU20	ARG193	Hydrogen bond, and salt bridge
ASP113	ARG72	Hydrogen bond, and salt bridge
GLU20	TYR78	Hydrogen bond
GLN120, GLU124	GLN79	Two hydrogen bonds
LYS17	ASP102, ASP105	Hydrogen bond, and two salt bridges

Fig. 3.

(A) Showing the interaction of the mannose ligand and peptide filgrastim, (B) the interaction between PLGA and peptide filgrastim, and (C) the interaction between ethylenediamine and peptide filgrastim.

Molecular dynamics simulation

We conducted a 100 ns molecular dynamics (MD) simulation on a protein-peptide complex generated from molecular docking. Using the Desmond package, we analyzed the trajectory files to assess root mean square deviation (RMSD) and root mean square fluctuation (RMSF) for interactive stability insights.

Root mean square deviation (RMSD)

RMSD shows the average change in the displacement of an atom at a specific molecular level, confirmed with its reference confirmation data. The trajectory analysis for the Macrophage Mannose Receptor complex and Filgrastim showed slight deviations. Initial fluctuations were observed due to the system equilibration phase and changes in the simulation environment. The initial deviations were noticed due to the initial head and a change in the medium. The Macrophage Mannose Receptor (PDBID: 1EGI) and the Granulocyte colony-stimulating factor (Filgrastim) Peptide (PDBID: 1PGR) (Fig. 4) showed an initial deviation of 2.77 Å in 0.80 ns, notable protein produced deviation 4.91 Å at 100 ns while peptide stated initial deviation of 4.78 Å at 2.90 Å and at 100 ns showed 6.68 Å.

Fig. 4.

RMSD of macrophage mannose receptor and peptide Filgrastim.

Root mean square fluctuation (RMSF)

RMSF measures how much each residue fluctuates from its original position over time, giving insights into the movement of residues at specific nanoseconds. The Macrophage Mannose Receptor (PDBID: 1EGI) and the Granulocyte colony-stimulating factor (Filgrastim) Peptide (PDBID: 1PGR) were simulated to analyse the RMSF and found the most fluctuating residues that were ASP122, ALA1, GLY2, PRO121, ALA7, HIS650, LYS649, GLY689, GLY648, PRO742, THR743, TYR691, GLU651, THR686, ALA687, SER688, NMA175, HIS692, ASP741, LYS647, and SER690 and rest of the residues were under control. Also, the interacting residues completely controlled fluctuations (Fig. 5).

Fig. 5.

RMSF of macrophage mannose receptor and peptide Filgrastim.

Drug polymer compatibility

The FT-IR study was performed to evaluate the drug-excipients compatibility and to analyse the molecular level interaction between the drug and polymers, i.e. PLGA, and mannose. The FT-IR spectrum of filgrastim (Fig. 6A) show broad absorption band at ~ 3290–3300 cm⁻¹ which is in confirmation of the protein compound in the substance (corresponds to the O–H and N–H stretching vibrations). Additionally, the peak at 1650 cm⁻¹ and 1530 cm⁻¹ corresponds to the amide I (C = O stretching) and amide II (N–H bending) vibrations, respectively, confirming the presence of peptide bonds. The Fig. 6B and C show the FT-IR spectra of PLGA and D-mannose respectively. It was evident that there was no formation of new peak, and further all characteristic peaks of the filgrastim were also available in the FT-IR spectra of the physical mixture (Fig. 6D). This suggests physical compatibility among Filgrastim, PLGA, and D-mannose, confirming that blending these components does not alter their chemical structures.

Fig. 6.

FT-IR spectra of (A) Filgrastim, (B) PLGA, (C) D-mannose, and (D) physical mixture of Filgrastim, PLGA, and D-mannose showing the characteristic peaks corresponding to their functional groups and confirming the compatibility among the components.

In-vitro analysis

The novel conjugates of PLGA, viz-a-viz placebo Mn-EDA-PLGA and G-CSF embedded Mn-EDA-PLGA, were optimized, synthesized (Fig. 7A, B) and characterized. The nanoparticles were prepared using the emulsion-solvent evaporation method described in the methodology section. Then, using diverse techniques, the NPs were characterised.

Fig. 7.

Method for preparation of placebo and G-CSF embedded PLGA nanoparticles(A), Representation of the organic and aqueous phase (Bi) and Homogenisation of the organic phase containing the drug (Bii).

Optimisation of Mn-EDA-PLGA nanoparticles

Effect of independent variable on particle size:

An imperative parameter in preparation of PLGA NPs is particle size, as the release of the drug and permeation are governed by particle size. A targeted drug delivery system’s optimum desirable particle size is 200–250 nm. The quadratic model used in the study (Table 3) was found to be statistically significant (p < 0.05) with a model F-value of 22.95. A p value below 0.0500 indicates that the model terms are significant. The model terms A, B, C, AC, and A2 were significant in this case. On the other hand, p values above 0.1000 indicate that the model terms are not significant. The predicted R2 value of 0.5962, which represents the proportion of the total variation in the response variable explained by the model, is not as close to the adjusted R2 value (R² = 0.9251) as expected. A difference greater than 0.2 between these values suggests a lack of rational agreement.

Table 3.

Shows the R-squared value of different models.

Source	Sum of squares	Mean square	F-value	p value	Remarks
Model	67.20	7.47	24.27	0.0002	Significant
A – PLGA Conc	6.79	6.79	22.07	0.0022	Significant
B – surfactant	2.13	2.13	6.93	0.0338	Significant
C – sonication time	0.0162	0.0162	0.0527	0.8251	Not significant
AB (interaction)	0.6561	0.6561	2.13	0.1876	Not significant
AC (interaction)	0.0012	0.0012	0.0040	0.9515	Not significant
BC (interaction)	3.59	3.59	11.67	0.0112	Significant
A2 (quadratic)	35.64	35.64	115.85	< 0.0001	Highly significant
B2 (quadratic)	15.19	15.19	49.38	0.0002	Highly significant
C2 (quadratic)	0.0438	0.0438	0.1424	0.7171	Not significant
Residual	2.15	0.3077	–	–	–
Lack of fit	1.44	0.4817	2.72	0.1792	Not significant
Pure error	0.7087	0.1772	–	–	–
Total (corrected)	69.35	–	–	–	–

Bold values indicate statistically significant terms (p < 0.05).

The quadratic equation obtained from the model indicates that the model terms can have either a positive or an adverse effect on the quality attributes being studied. However, without further information about the specific coefficients of the equation and their corresponding signs, it is not possible to determine the exact nature and magnitude of these effects. It is worth noting that the information provided lacks detailed coefficients and interpretations of the model, which are necessary to fully understand the impact of the independent variables on the quality attributes. In this model, terms like PLGA concentration (A), surfactant (B), and sonication time (C) were taken. The following equation predicts AB = PLGA and surfactant with a substantial p value of 0.0002 and p < 0.05, respectively (Table 4). Based on the provided quadratic equation for particle size:

Table 4.

The data reflect that the model follows a quadratic equation.

Model	Sequential p value	Lack of fit p value	Adjusted R2	Predicted R2	Remarks
Linear	0.1013	0.0021	0.2247	–0.2171
2FI (two-factor interaction)	0.1742	0.0027	0.3731	–0.5909
Quadratic	0.0004	0.1110	0.9251	0.5962	Suggested
Cubic	0.1110	–	0.9666	–	Aliased

Bold values indicate statistically significant terms (p < 0.05).

The negative coefficients of variables B and C (− 68.90 and − 49.80, respectively) indicate that these variables have a negative impact on particle size, which means that an increase in the values of variables B and C would decrease particle size. Additionally, the positive interaction term BC (13.35) suggests an interaction effect between variables B and C on particle size. This means that the combined effect of increasing variable B and decreasing variable C (or vice versa) has a specific impact on particle size, which differs from the individual effects of these variables. It is important to note that without specific values for variables A, B, and C, it is impossible to determine the exact numerical impact of these variables on particle size. The coefficients in the equation indicate the relative importance and direction of the effects, but do not provide precise quantitative information. The coefficient variable for Variable B is at its maximum value, indicating that it has a stronger influence on particle size than Variables C and A, which have a lesser impact. Furthermore, variable A (+ 25.93) had a positive value, indicating a favourable impact on particle size. The range of particle sizes is 99.1–452.1 nm. Therefore, it was concluded that with 3D response graphs (Fig. 8A), the particle size was significantly impacted by PLGA surfactant concentration and sonication time. The plot between the predicted and actual value is shown in Fig. 8B.

Fig. 8.

Surface response graphs showing effect of surfactant concentration, PLGA and sonication time on particle size (A), Plot between predicted and actual values for particle size (B).

The effect of PLGA concentration, surfactant and sonication time on entrapment efficiency and particle size was studied. A noticeable positive effect on PLGA was observed while other independent variables like surfactant and sonication time, negatively impacted particle size after fitting the data of response in the model. Thus, we conclude that the concentration of PLGA, surfactant and sonication time significantly affected the particle size.

Effect of independent variable on entrapment efficiency

Based on the provided quadratic equation for entrapment efficiency:

The negative coefficients of variables B (− 0.5162) and C (− 0.0450) indicate that these variables negatively impact entrapment efficiency. This means that an increase in the values of variables B and C would decrease entrapment efficiency (Fig. 9A). On the other hand, the positive coefficient of variable A (0.9213) suggests that it positively impacts entrapment efficiency. An increase in the value of variable A would increase the entrapment efficiency (Fig. 9B). The quadratic terms, A, B, C, A2 and B2, also have coefficients with negative values. This indicates that the squared values of variables A, B, and C negatively impact entrapment efficiency. It is important to note that without specific values for variables A, B, and C, it is impossible to determine the exact numerical impact of these variables on entrapment efficiency. The coefficients in the equation indicate the relative importance and direction of the effects, but do not provide precise quantitative information. Additionally, the decrease in particle size as the surfactant concentration increases is related to the interfacial tension between the oil and aqueous phases. Higher surfactant concentrations can lower the interfacial tension, leading to smaller particle sizes. Increasing PLGA concentration also resulted in larger particle sizes, which suggests a relationship between PLGA concentration and particle size. The variables B and C negatively impact entrapment efficiency, while variable A has a positive impact, and both PLGA concentration and surfactant concentration influence the particle size.

Fig. 9.

Surface response graphs showing effect of surfactant concentration PLGA, and the sonication time on the entrapment efficiency of NPs (A). Plot between predicted and actual values for the entrapment efficiency of NP (B).

Characterization of Mn-EDA-PLGA nanoparticles

Particle size, charge and PDI index

The mean diameter of placebo NPs and Mn-NPs was 199 nm ± 12 and 374.7 nm, respectively (Fig. 10). The size of G-CSF MnEDA PLGA NPs did not increase considerably due to the encapsulation of the monomeric form of G-CSF in NPs. The PDI data suggested that the NPs formed were monodisperse. Zeta potential was found to be − 43.04 ± 1.4 and − 34.9 ± 1.90, respectively.

Fig. 10.

The mean diameter of placebo NPs (i) and Mn-NPs (ii).

Entrapment efficiency of G-CSF Mn-NPs

The percentage entrapment efficiency (%EE) of prepared formulation was found to be 89.23 to 95.36%. The effects of the independent variables on entrapment efficiency are reflected in the quadratic equation.

Scanning electron microscopy (SEM) of G-CSF Mn-NPs

SEM images of optimized formulation revealed that prepared nanoparticles were spherical, and a smooth morphology was seen in Fig. 11. In the case of 11B, the Mn-NPs are observed to have a uniform distribution of mannose. SEM is used to examine the size and shape of engineered NPs. SEM images showed that NPs are spherical and nanometre-sized. The Mannose link mannose NPs are less spherical, and smoother compared to the placebo.

Fig. 11.

SEM images of optimized formulation of G-CSF Mn-NPs (A), and Mn-NPs (B).

Thermogravimetric analysis (TGA)

Thermogravimetric data of placebo and G-CSF embedded Mn-EDA-PLGA NPs are given in Fig. 12A and Fig. 12B respectively. The data clearly indicate the in the placebo NPs, showed a single major decomposition pattern between 250 and 360 °C, which is due to the polyester chain degradation of the polymeric backbone. However in case of Mn-EDA-PLGA NPs, the DTG curve further confirmed a two-step degradation mechanism. There was additional degradation patter at the position of 150–250 °C corresponding to protein denaturation, followed by polymer decomposition at 260–350 °C. The drug-loaded nanoparticles also retained higher residual mass (~ 25%), indicating the presence of proteinaceous char and enhanced thermal stability. These findings confirm successful incorporation of G-CSF into the polymer matrix.

Fig. 12.

Thermogravimetry analysis of placebo (A) and G-CSF embedded Mn-EDA-PLGA NPs (B).

In-vitro drug release study

The release rate of plane (non-manosylated) (Fig. 13A) and Mn-EDA-PLGA NPs (Fig. 13B) was determined and release rates were measured over 16 h. The release of G-CSF from both the nanoparticles followed a sustained pattern. The release of G-CSF from the formulation over the 16 h was found to be 91.15 ± 0.14% for the plane NPs and 94.49 ± 0.42% for the Mn-EDA-PLGA NPs. This indicates that a significant amount of G-CSF was released from the NPs during this time, suggesting sustained release characteristics. In the 0.2 M sodium acetate buffer at pH 5.5, which mimics the endosomal compartment of macrophages with a lower pH, the release rate of G-CSF from both types of NPs was slightly higher compared to physiological pH. The release of G-CSF over the 16 h was measured to be 92.81 ± 1.59% for the plane NPs and 96.16 ± 0.28% for the Mn-EDA-PLGA NPs. This indicates that the NPs released G-CSF faster in the endosomal compartment with a lower pH. These observations suggest that the pH of the surrounding environment can influence the release of NPs and the drug (G-CSF) from the NPs. The lower pH in the endosomal compartment promotes faster drug release, which can be advantageous for targeted drug delivery to macrophages. Considering the experimental conditions and measurement variability, these values estimate the average release behaviour.

Fig. 13.

In-vitro release study of G-CSF from plain PLGA nanoparticles and Mn-EDA-PLGA NPs at pH 7.4 (A) and pH 5.5 (B).

Uptake and transport by bone marrow macrophages using the flow cytometry technique

To evaluate the uptake of nanoparticles (NPs) by bone marrow macrophages, the J774.2 macrophage cell line was employed. Since mannose receptors are abundant on bone marrow macrophages, the uptake was studied using flow cytometric analysis. Placebo NPs (drug-free) were used as negative controls, while FITC-labelled G-CSF-loaded Mn-EDA-PLGA NPs were used to assess mannose-mediated uptake. Flow cytometry results (Fig. 14) showed a sharp increase in fluorescence (49.5 ± 1.25%) in the G-CSF-loaded Mn-EDA-PLGA NPs group, compared to the negligible signal in the placebo NPs group, confirming receptor-mediated internalization. Placebo NPs (drug-free) were used exclusively for Physicochemical characterization (particle size, zeta potential). Cellular uptake controls (Fig. 14A). Biodistribution controls (Fig. 14B). These nanoparticles were negative controls throughout the in vitro and in vivo studies.

Fig. 14.

Flow Cytometry dot plots; dot plot (A) Placebo group showing no fluorescence (0.8%); dot plot (B) Mannose PLGA G-CSF group showing augmentation in fluorescence (49.5 ± 1.25%).

In-vivo experiment

Efficacy study (Methotrexate-induced neutropenia)

Methotrexate (MTX) causes bone marrow suppression as a potential side effect and lowers neutrophil levels in healthy animals after 30 days. G-CSF incorporated mannose PLGA nanoparticles effectively upregulate the RBC, WBC, eosinophil and neutrophil count (Table 5).

Table 5.

Haematological parameters of the animals from the different groups (n = 6).

S. No	Parameters	Group 1 Placebo NPs	Group 2 MTX-induced	Group 3 MTX-induced (Mn-NPs)	Group 4 MTX-Induced (Mn-GCSF- NPs)	ANOVA (p value)
1	Hemoglobin (mg/dl)	7.65 ± 0.47a	7.35 ± 0.43a	8.03 ± 0.51ab	11.33 ± 0.72c	0.03
2	RBC (× 106/mm3)	5.70 ± 0.35a	4.17 ± 0.24b	4.50 ± 0.33b	7.15 ± 0.48c	0.01
3	WBC (× 103/mm3)	8.06 ± 425ab	8083.3 ± 439ab	7083.3 ± 386a	10,300 ± 524b	0.14
4	Platelets(X103/mm3)	840.12 ± 51a	824 ± 48a	812.23 ± 43a	1231.3 ± 72b	0.001
5	Neutrophils (%)	15 ± 1.06a	15.7 ± 1.13a	16.83 ± 1.27a	25.8 ± 1.51b	0.63
6	Eosinophils (%)	69.5 ± 3.57a	65.3 ± 3.23a	68 ± 3.41a	84.5 ± 4.03b	0.53

Values are expressed as mean ± SEM (n = 6).

Different superscripts (a, b, c) within a row indicate statistically significant differences among groups (p < 0.05, one-way ANOVA followed by Tukey’s multiple comparison test).

Biodistribution study

Animals were injected with 99mTc-labelled G-CSF-loaded Mn-EDA-PLGA NPs—placebo NPs (drug-free) to assess distribution patterns. Tissue samples (brain, kidney, liver, spleen, lungs, heart, and stomach) were collected at 15, 30, 60, 120, and 240 min. Radioactivity was measured using a gamma counter (Fig. 15). To assess the distribution and localization of the NPs, the animals were sacrificed using CO₂ euthanasia. After sacrificing the rats, their organs of interest were removed. These organs were then cleansed to remove any external contamination or residual NPs on their surfaces. The counts of radioactivity associated with the NPs in each organ were obtained using a gamma counter. The specific gamma counter used in this study was the Caprac®-t wipe test / well counter manufactured by Capintech.

Fig. 15.

Biodistribution study of Mn GCSF-NPs in albino Wistar rats.

The gamma counter allows for detecting and quantifying gamma radiation emitted by the 99mTc-tagged NPs. This enables researchers to measure the radioactivity levels in each organ and obtain information about the distribution and accumulation of the NPs over time. By analysing the counts obtained from the gamma counter at different time intervals, researchers can evaluate the uptake and clearance kinetics of the NPs in various organs of interest. The G-CSF-loaded Mn-EDA-PLGA NPs showed preferential accumulation in bone marrow-rich tissues, indicating macrophage targeting via mannose-receptor interaction. Placebo NPs did not exhibit targeted accumulation, supporting their role as biodistribution controls. This information is crucial for understanding the biodistribution and pharmacokinetics of the NPs in the animal model, which can further inform the potential applications and effectiveness of the NPs for targeted drug delivery or imaging purposes.

Comparison with the existing literature

Unlike Pegfilgrastim, which relies solely on PEGylation to prolong systemic circulation, our Mn-EDA-PLGA NPs leverage mannose receptor-mediated endocytosis to target bone marrow-resident macrophages actively. This approach enables precise granulocyte colony-stimulating factor (G-CSF) localization while minimising off-target accumulation and associated adverse effects such as bone pain. Developed NPs demonstrate a controlled G-CSF release profile, achieving 94.28% release over 16 h (Fig. 13), which closely parallels the extended pharmacodynamics of Pegfilgrastim but is further enhanced by endosomal pH-triggered release (pH 5.5), aligning drug availability with physiological needs. High drug entrapment efficiency (72.6%) and targeted delivery enable therapeutic efficacy at a markedly lower dose of 3 µg per administration, in contrast to the standard 300 µg required for free G-CSF, thereby reducing systemic burden. In vivo studies using a methotrexate (MTX)-induced neutropenia model reinforce these findings, with neutrophil restoration reaching 25.8% using Mn-EDA-PLGA NPs versus 15.7% in the MTX control group (Table 5). While prior studies, such as Sou et al.¹⁴, Expert Opin. Drug Deliv.), have suggested the potential of PLGA carriers for bone marrow applications, our mannose-EDA conjugation represents a novel strategy for G-CSF targeting. Moreover, unlike PEGylated G-CSF systems that exhibit poor macrophage uptake⁴⁵, Front. Bioeng. Biotechnol.), our platform integrates receptor-mediated targeting, endosomal-triggered release, and dose minimisation three synergistic features not simultaneously achieved in existing formulations. Table 6 provides a comparative overview underscoring the innovation and translational promise of Mn-EDA-PLGA NPs for neutropenia management.

Table 6.

Comparative analysis of sustained-release G-CSF formulations.

Parameter	Mn-EDA-PLGA NPs (current study)	Filgrastim(Exp. Hematol., 1999)	Pegfilgrastim(Neulasta®, 2004)
Formulation approach	Mannose-anchored PLGA nanoparticles for active targeting of bone marrow macrophages	Unmodified recombinant G-CSF with empirical release kinetics	PEGylated G-CSF for passive extension of plasma half-life
Targeting mechanism	Receptor-mediated uptake via mannose receptors on macrophages reduces peripheral distribution	Non-targeted systemic exposure	Non-targeted; avoids renal clearance through PEGylation
Particle size/PDI	153 ± 12.2 nm; PDI: 0.41 (uniform size distribution)	Not applicable (solution form)	Not applicable (solution form)
Entrapment efficiency	72.6% drug encapsulation efficiency (DEE); effective at 3 µg G-CSF/dose	Not applicable (free protein)	Not applicable (PEG-conjugated protein)
Release profile	Sustained release (94.28% over 16 h); pH-sensitive release at endosomal pH (5.5)	Short plasma half-life (~ 3.5 h)	Extended half-life (~ 15–80 h); lacks controlled or targeted release
Biodistribution	9.3% injected dose (ID) in bone marrow (γ-scintigraphy); < 2% for non-targeted NPs	Diffuse systemic distribution	Primarily liver and spleen; minimal marrow accumulation
Therapeutic advantage	Reduced dose and cost (∼$37.76/cycle; ~ 3 × reduction); mitigates bone pain (NCT03407430)	Requires frequent dosing; higher cumulative exposure	Infrequent dosing; lacks marrow specificity; associated with bone pain
Clinical relevance	Potential for alternate routes (e.g., oral/needle-free delivery)	Injectable only	Injectable only

Discussion

Chemotherapy is widely used in managing diverse cancers, but myelosuppressive chemotherapeutic drugs often cause neutropenia. G-CSF is co-administered to counteract this, though limitations such as short half-life, systemic toxicity, and high cost remain major problems that contribute to the failure of chemotherapeutic drug effectiveness⁷. Mannosylated polymeric nanoparticles improve G-CSF availability in neutrophils and reduce dosing due to efficient drug delivery at the target site⁴⁶. This approach minimized the adverse effect on peripheral tissue exposed to a high concentration of G-CSF. Further, the cost and therapeutic quantity of G-CSF (300 µg, $ 37.76) can be reduced up to one-third, which augments cost-effective application of colony-stimulating factors. Similar studies, such as Kumari et al., prepared a PEGylated version of G-CSF to improve systemic circulation stability⁴⁵. Seung et al. developed G-CSF-loaded PLGA nanoparticles to prolong availability in systemic circulation⁴⁷. Alebouyeh et al. demonstrated formulation aspects to improve the stability of G-CSF against thermal stress and biological conditions⁴⁸. Some patents also enable effective targeting of G-CSF. A Patent (WO 2,004,100,983 A2) described the method of macrophage targeting through folate receptor binding to treat lupus erythematosus (Low Stewart Philip). Formulation consists of a ligand capable of binding to activated macrophages or other stimulated immune cells to deliver cytotoxin immunogen. Another patent, WO 1,999,041,285 A1, describes methods and compositions designed to select macrophages in a localized area⁴⁹. The compositions disclosed in this patent consist of an Fc receptor binding agent and a toxic or detectable agent. These compositions are utilized for depleting or inhibiting the activity of macrophages. The patent highlights that the described compositions can be employed for therapeutic and diagnostic purposes. By specifically targeting macrophages, which are immune cells involved in various physiological processes and diseases, these compositions offer potential applications in treating and diagnosing macrophage-related conditions. The Fc receptor binding agent in the compositions enables specific binding to macrophages, facilitating targeted delivery of the toxic or detectable agent. This targeted approach enhances the efficiency and accuracy of macrophage depletion or inhibition. Another patent (WO 2,002,087,424 A2) included details on monitoring, treating, or diagnosing a disease state brought on by activated macrophages⁵⁰. The approach includes giving an effective dose of a composition containing a conjugation or complex of a ligand capable of binding to activated macrophages and an immunogen to a patient suffering from a macrophage-mediated illness condition. The patent (WO 2,011,015,333 A2) covers using antibodies to treat or detect leukaemia, specifically using such antibodies to target an antigen expressed in bone marrow neovasculature⁵¹. In the current study, we use a mannose ligand on PLGA NPs and the Mn-EDA-PLGA are taken by the MPS (Mononuclear phagocyte system). This system’s structure enables the localization of the G-CSF medication entrapped inside the macrophages. Targeting particular receptors is a more effective way to raise the concentration of NP at the desired action location. A suitable location for the uptake of nanoparticles is the macrophage membrane, which contains mannose receptors. In the current investigation, PLGA and Mannose were directly conjugated with EDA (Mn-EDA-PLGA NPs). Then, using PVA as a surfactant, a double emulsion solvent evaporation process was used to create the nanoparticles. Mn-EDA-PLGA and placebo PLGA nanoparticles were discovered to have mean diameters of 153 ± 12.2 nm and 178 ± 10.4 nm, respectively.

The G-CSF present in its monomeric form during the formation of nanoparticles did not lead to a significant increase in particle diameter upon encapsulation. However, a slight increase in this parameter was observed for M-PLGA and Mn-EDA-PLGA nanoparticles, although their polydispersity index remained very low. The zeta potentials of PLGA and Mn-EDA-PLGA nanoparticles were measured to be 43.04 ± 1.4 mV and 34.9 ± 1.9 mV, respectively. Scanning electron microscopy (SEM) analysis confirmed that the spherical Mn-PLGA and Mn-EDA-PLGA nanoparticles were in the nanometric range. However, the engineered nanoparticles exhibited surfaces that were slightly less smooth and occasional minimal aggregation compared to the PLGA nanosystems. Cellular uptake and in vivo biodisposition patterns were evaluated to assess the macrophage targeting capabilities of the designed nanoparticles in macrophage-rich tissues. In the intra-amastigote macrophage model of methotrexate-induced neutropenia, both plain and engineered G-CSF-loaded nanoparticles demonstrated superior efficacy to the free drug. Specifically, Mn-EDA-PLGA nanoparticles outperformed Mn-PLGA nanoparticles in terms of their therapeutic effect. Biodistribution data obtained by measuring the maximum concentration of administered doses in macrophage-rich tissues indicated that mannosylated nanoparticles exhibited the best targeting properties. Similar strategies employing polymeric devices have been successfully employed by other research teams for effective G-CSF delivery. Furthermore, the MTC formulation exhibited superior pharmacological action compared to G-CSF and demonstrated higher biocompatibility on J774 macrophagic cell lines, highlighting the potential of mannosylation as a promising method for treating intracellular infections. Engineered Mn-EDA-PLGA NPs were created using polymer-ligand conjugates, and their ability to target macrophages was assessed.

Chemotherapy remains a cornerstone in cancer treatment, but is frequently associated with neutropenia due to myelosuppression, necessitating the use of G-CSF to stimulate neutrophil recovery. However, the clinical use of G-CSF is hampered by its short systemic half-life, off-target toxicity, frequent dosing requirements, and significant cost. In this study, a targeted delivery strategy using mannose-conjugated PLGA nanoparticles (Mn-EDA-PLGA NPs) was employed to improve the pharmacological efficiency of G-CSF²⁴. Mannose receptors, abundantly expressed on macrophages, served as specific docking sites, enabling enhanced accumulation of G-CSF at the mononuclear phagocyte system (MPS), especially in bone marrow-resident macrophages. The engineered Mn-EDA-PLGA NPs demonstrated optimal size (~ 153 nm), monodispersity, and surface charge suitable for systemic administration and cellular uptake. The direct conjugation of mannose and mannose-PLGA facilitated receptor-mediated endocytosis, increasing drug accumulation in target tissues while minimising systemic exposure. Encapsulation of G-CSF in these nanocarriers preserved its bioactivity, as confirmed by its therapeutic superiority over free G-CSF in methotrexate-induced neutropenic models. In vivo biodistribution studies revealed pronounced localization in macrophage-rich tissues, validating the targeting efficiency of the system. SEM analysis confirmed well-defined spherical morphology with minimal aggregation. Notably, mannosylated nanoparticles allowed for reduced dosing without compromising efficacy, translating into lower therapeutic cost and decreased off-target side effects. Furthermore, the MTC formulation displayed enhanced biocompatibility on macrophagic J774 cell lines and superior pharmacodynamic profiles compared to plain PLGA or non-targeted formulations. The present formulation strategy demonstrates a promising avenue for targeted cytokine delivery, leveraging ligand-mediated targeting to overcome the pharmacokinetic limitations of protein therapeutics. Overall, the development of Mn-EDA-PLGA NPs provides a scalable, cost-effective, and targeted platform for G-CSF delivery with potential applications in neutropenia management and broader immunotherapy contexts. The crucial result of our work is higher G-CSF entrapment at low concentration and decreased cost. Our strategy focused on mannose-anchored, EDA-linked PLGA nanoparticles (Mn-EDA-PLGA NPs) for targeted delivery of G-CSF to bone marrow macrophages. The rationale is that mannose receptors are highly expressed on macrophage surfaces, offering a homing mechanism. Our data showed substantial G-CSF entrapment at lower concentrations, suggesting cost-effectiveness.

Conclusion

This study presents a promising nanotechnology-driven platform using Mn-EDA-PLGA NPs for macrophage-targeted delivery of G-CSF in neutropenia. The nanoparticles were successfully engineered, optimized, and validated in silico, in vitro, and in vivo models. The nanoparticles, specifically Mn-EDA-PLGA and Mn-PLGA, exhibited favourable characteristics such as nanometric size, low polydispersity index, and stable zeta potentials. The encapsulation of G-CSF did not significantly alter the particle size, indicating successful formulation. The designed nanoparticles demonstrated efficient targeting of macrophage-rich tissues, as evidenced by their enhanced cellular uptake and in vivo biodisposition patterns. In the intra-amastigote macrophage model context, G-CSF-loaded nanoparticles exhibited superior therapeutic efficacy compared to free G-CSF, with Mn-EDA-PLGA nanoparticles showing the most promising results. Furthermore, the mannosylated nanoparticles displayed excellent targeting properties, as indicated by their preferential accumulation in macrophage-rich tissues. This targeting ability holds great potential for treating intracellular infections, as demonstrated by the superior pharmacological action and higher biocompatibility of the MTC formulation on macrophagic cell lines. This study highlights the significance of nanoparticle-based delivery systems for macrophage targeting and emphasises the efficacy of mannosylation as a valuable approach to improving therapeutic outcomes. These findings support advancing targeted nanomedicine platforms, particularly for treating macrophage-associated diseases and infections, and the computational studies further substantiate the proposed hypothesis. To fully establish the translational efficacy of our formulation, future investigations are needed to confirm the long-term preservation of the encapsulated G-CSF’s structure, biological activity and therapeutic effectiveness post-administration. We plan to conduct advanced stability studies and extended invivo evaluations to further validate the preserved function of the delivered protein.

Abbreviations

PLGA

Poly(D, L-lactide-co-glycolide)

G-CSF

Granulocyte colony-stimulating factor

NPs

Nanoparticles

SEM

Scanning electron microscopy

EDA

Ethylenediamine

NHS/DSS

N-hydroxysulfosuccinimide/dicyclocarbodiimide

DMSO

Dimethyl sulfoxide

PDI

Polydispersity index

MTX

Methotrexate

FT-IR

Fourier transformation-infrared spectroscopy

Author contributions

Ritu Karwasra; Conceptualisation, Data Collection, Analysis, Writing- first draft. Nagmi Bano; Computational Data Collection, Analysis, Writing- first draft. Shaban Ahmad; Computational Data Analysis, Writing- first draft. Surender Singh; Supervision, Writing- Reviewing and Editing. Kushagra Khanna; Analysis, Writing- first draft. Nitin Sharma; Data Collection, Analysis, Writing- first draft. Khalid Raza; In-silico study Supervision, Computational Resources, Writing- Reviewing and Editing. Saurabh Verma; Conceptualisation, Supervision, Resources, Project Administration, Writing- Reviewing and Editing.

Funding

This research work was funded by the Indian Council of Medical Research, ICMR, in the form of the ICMR-Centenary Postdoc Fellowship with file number 3/1/3PDF(19)/2019-HRD.

Data availability

All data supporting the findings of this study are included within the manuscript as figures, tables or text. Guidelines: The study is reported in accordance with ARRIVE guidelines. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.