Volume 181, Issue 8 p. 1238-1255
RESEARCH ARTICLE
Open Access

Development of a therapeutic monoclonal antibody against circulating adipocyte fatty acid binding protein to treat ischaemic stroke

Boya Liao

Boya Liao

Department of Pharmacology and Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China

State Key Laboratory of Pharmacological Biotechnology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China

School of Chinese Medicine, Hong Kong Baptist University, Hong Kong, China

Contribution: ​Investigation (equal), Writing - original draft (equal)

Search for more papers by this author
Shilun Yang

Shilun Yang

Institute of Biomedicine and Biotechnology, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China

Contribution: ​Investigation (equal), Writing - original draft (equal)

Search for more papers by this author
Leiluo Geng

Leiluo Geng

State Key Laboratory of Pharmacological Biotechnology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Department of Medicine, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Contribution: ​Investigation (supporting)

Search for more papers by this author
Jiuyu Zong

Jiuyu Zong

Department of Pharmacology and Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China

State Key Laboratory of Pharmacological Biotechnology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Contribution: ​Investigation (supporting)

Search for more papers by this author
Zixuan Zhang

Zixuan Zhang

Department of Pharmacology and Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China

State Key Laboratory of Pharmacological Biotechnology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Contribution: ​Investigation (supporting)

Search for more papers by this author
Mengxue Jiang

Mengxue Jiang

Department of Pharmacology and Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China

State Key Laboratory of Pharmacological Biotechnology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Contribution: ​Investigation (supporting)

Search for more papers by this author
Xue Jiang

Xue Jiang

State Key Laboratory of Pharmacological Biotechnology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Department of Medicine, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Contribution: ​Investigation (supporting)

Search for more papers by this author
Simeng Li

Simeng Li

Institute of Biomedicine and Biotechnology, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China

Contribution: ​Investigation (supporting)

Search for more papers by this author
Aimin Xu

Aimin Xu

Department of Pharmacology and Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China

State Key Laboratory of Pharmacological Biotechnology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Department of Medicine, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Contribution: Writing - review & editing (supporting)

Search for more papers by this author
Junlei Chang

Corresponding Author

Junlei Chang

Institute of Biomedicine and Biotechnology, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China

Correspondence

Ruby Lai Chong Hoo, Department of Pharmacology and Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, Rm 208a, Laboratory Block, 21 Sassoon Road, Pokfulam, Hong Kong, China.

Email: [email protected]

Junlei Chang, Institute of Biomedicine and Biotechnology, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, B1112, 1068 Xueyuan Avenue, Nanshan Xili, Shenzhen, China.

Email: [email protected]

Contribution: Conceptualization (equal), Resources (equal), Supervision (equal), Validation (equal), Writing - original draft (lead), Writing - review & editing (lead)

Search for more papers by this author
Ruby Lai Chong Hoo

Corresponding Author

Ruby Lai Chong Hoo

Department of Pharmacology and Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong, China

State Key Laboratory of Pharmacological Biotechnology, Faculty of Medicine, The University of Hong Kong, Hong Kong, China

Correspondence

Ruby Lai Chong Hoo, Department of Pharmacology and Pharmacy, LKS Faculty of Medicine, The University of Hong Kong, Rm 208a, Laboratory Block, 21 Sassoon Road, Pokfulam, Hong Kong, China.

Email: [email protected]

Junlei Chang, Institute of Biomedicine and Biotechnology, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, B1112, 1068 Xueyuan Avenue, Nanshan Xili, Shenzhen, China.

Email: [email protected]

Contribution: Conceptualization (equal), Resources (equal), Supervision (equal), Validation (equal), Writing - original draft (lead), Writing - review & editing (lead)

Search for more papers by this author
First published: 10 November 2023

Boya Liao and Shilun Yang have contributed equally to this work.

Abstract

Background and Purpose

Adipocyte fatty acid-binding protein (A-FABP) exacerbates cerebral ischaemia injury by disrupting the blood–brain barrier (BBB) through inducing expression of MMP-9. Circulating A-FABP levels positively correlate with infarct size in stroke patients. We hypothesized that targeting circulating A-FABP by a neutralizing antibody would alleviate ischaemic stroke outcome.

Experimental Approach

Monoclonal antibodies (mAbs) against A-FABP were generated using mouse hybridoma techniques. Binding affinities of a generated mAb named 6H2 towards various FABPs were determined using Biacore. Molecular docking studies were performed to characterize the 6H2-A-FABP complex structure and epitope. The therapeutic potential and safety of 6H2 were evaluated in mice with transient middle cerebral artery occlusion (MCAO) and healthy mice, respectively.

Key Results

Replenishment of recombinant A-FABP exaggerated the stroke outcome in A-FABP-deficient mice. 6H2 exhibited nanomolar to picomolar affinities to human and mouse A-FABP, respectively, with minimal cross-reactivities with heart and epidermal FABPs. 6H2 effectively neutralized JNK/c-Jun activation elicited by A-FABP and reduced MMP-9 production in macrophages. Molecular docking suggested that 6H2 interacts with the “lid” of the fatty acid binding pocket of A-FABP, thus likely hindering the binding of its substrates. In mice with transient MCAO, 6H2 significantly attenuated BBB disruption, cerebral oedema, infarction, neurological deficits, and decreased mortality associated with reduced cytokine and MMP-9 production. Chronic 6H2 treatment showed no obvious adverse effects in healthy mice.

Conclusion and Implications

These results establish circulating A-FABP as a viable therapeutic target for ischaemic stroke, and provide a highly promising antibody drug candidate with high affinity and specificity.

Abbreviations

  • A-FABP, AP2, FABP4
  • adipocyte fatty acid binding protein
  • AP-1
  • activator protein 1
  • BBB
  • blood brain barrier
  • E-FABP
  • epidermal fatty acid binding protein
  • ELISA
  • enzyme-linked immunosorbent assay
  • FBS
  • fetal bovine serum
  • FFAs
  • free fatty acids
  • H-FABP, FABP3
  • heart fatty acid binding protein
  • HRP
  • horseradish peroxidase
  • IgG
  • non-immunoglobulin G
  • JNK
  • c-Jun N-terminal kinase
  • KLH
  • keyhole limpet haemocyanin
  • KO
  • knockout
  • mAbs
  • monoclonal antibodies
  • MCAO
  • middle cerebral artery occlusion
  • MOE
  • molecular operating environment
  • mTE
  • mechanical thrombectomy
  • rA-FABP
  • recombinant adipocyte fatty acid binding protein
  • SPR
  • surface plasmon resonance
  • TC
  • total cholesterol
  • TJ
  • tight junction
  • TMB
  • 3,3′, 5,5;-tetramethylbenzidine
  • tPA
  • tissue plasminogen activator
  • TTC
  • 2,3,5-triphenyltetrazolium chloride
  • VH
  • heavy chain variable regions
  • VL
  • light chain variable regions
  • WT
  • wildtype
  • ZO-1
  • zonula occluden-1
  • What is already known?

    • A-FABP exaggerates ischaemic stroke outcomes through inducing disruption of the blood-brain-barrier by increasing microglia/macrophage-derived MMP-9.
    • Circulating A-FABP is positively correlated with infarct size and early death of ischaemic stroke patients.

    What does this study add?

    • Replenishment of recombinant A-FABP in the circulation of A-FABP deficient mice exaggerated ischaemic stroke outcome.
    • Treatment with mAb 6H2 against circulating A-FABP attenuated transient MCAO-induced ischaemic stroke injury in mice.

    What is the clinical significance?

    • 6H2 is a promising antibody drug candidate with superb affinity and specificity towards A-FABP.
    • Neutralization of circulating A-FABP is a potential therapeutic strategy for alleviating ischemic stroke outcome.

    1 INTRODUCTION

    Ischaemic stroke is the major leading cause of death and disability worldwide (Feigin et al., 2021). Currently, the only pharmacological intervention approved by US Food and Drug Administration for ischaemic stroke treatment is tissue plasminogen activator (tPA) (Marler & Goldstein, 2003). However, the short treatment window and side effects of tPA such as high prevalence of haemorrhagic transformation and the fatal consequence stringently limit its clinical application (Dewar & Shamy, 2020). Thus, only a small number of patients benefit from this thrombolytic therapy. For those who are ineligible for tPA treatment, mechanical thrombectomy can be performed to remove blood clots but this is only feasible for patients with large artery occlusion. Moreover, the risks of mechanical thrombectomy, including intracerebral haemorrhage, vessel injury, vasospasm of the access vessel and stent-related complications outweigh the benefits (Behme et al., 2014). Therefore, novel and effective therapies against ischaemic stroke are critical and warranted.

    Adipocyte fatty acid-binding protein (A-FABP), also called AP2 or FABP4, is an adipokine abundantly expressed in adipocytes and macrophages (Hoo et al., 2017) and can be secreted into the circulation (Xu et al., 2006). Besides its physiological function as a fatty acid chaperone, A-FABP is also a key mediator of inflammation (Hui et al., 2010). High levels of circulating A-FABP are found in patients with metabolic complications including obesity, atherosclerosis and Type 2 diabetes which are the risk factors of stroke (Tso et al., 2007; Xu et al., 2006; Yeung et al., 2007). In addition, circulating A-FABP is closely associated with early stroke recurrence (Li et al., 2019), early death (Tso et al., 2011) and is an independent prognostic biomarker in patients with acute ischaemic stroke (Tu et al., 2017). We confirmed that circulating A-FABP is elevated in stroke patients and positively associated with infarct size (Liao et al., 2020). In animal studies, we have demonstrated that there was an elevated circulating and cerebral A-FABP in mice upon ischaemic stroke insult (Liao et al., 2020). Circulating A-FABP was increased 2 h after middle cerebral artery occlusion (MCAO), reached a peak after 48 h and gradually returned to baseline after 72 h. The expression of A-FABP was also significantly induced in the ischaemic brain 2 h after stroke and remained high for at least 72 h. Peripheral blood monocytes, microglia and infiltrated monocytes were identified as the major cellular source of circulating and cerebral A-FABP, respectively. A-FABP exacerbates ischaemic stroke injury through promoting expression of matrix metallopeptidase 9 (MMP-9), via activating the JNK-c-Jun signalling pathway thus mediating blood–brain barrier (BBB) disruption (Liao et al., 2020). Genetic deletion of A-FABP or pharmacological inhibition of A-FABP using its selective inhibitor BMS309403 significantly alleviated the outcomes of ischaemic stroke, associated with reduced BBB disruption and post-stroke inflammation (Liao et al., 2020). Thus, data from both human and animal studies suggest that A-FABP is a promising therapeutic target to ischaemic stroke injury.

    Although blocking of A-FABP activity using BMS309403 showed beneficial effects in animal models of Type 2 diabetes and atherosclerosis (Furuhashi et al., 2007), steatohepatitis (Hoo et al., 2013), liver fibrosis (Wu et al., 2021) as well as ischaemic stroke (Liao et al., 2020), this compound (BMS309403) exhibited low specificity and off-target activities, including suppression of cardiomyocyte contractility (Look et al., 2011) and stimulation of glucose uptake in C2C12 myotubes (Lin et al., 2012), which limited its further development for clinical use.

    There is an increasing trend of using antibodies as drugs for treatment of a range of diseases, because of their higher specificity, fewer off-target side effects, fewer drug–drug interaction and relatively long half-life (Behrens et al., 2022; Gao et al., 2022; Lu et al., 2020). A previous proof-of-concept study showed that a monoclonal antibody CA33 targeting secretory A-FABP, despite showing low affinity (micromolar range) and low specificity to A-FABP, exhibited an anti-diabetic effect in obese mice, indicating the feasibility of antibody-based A-FABP inhibitors (Burak et al., 2015). Thus, given the important role of A-FABP in ischaemic stroke as described above, developing a monoclonal antibody targeting circulating A-FABP, with high affinity and high specificity, could provide a potential therapeutic strategy for ischaemic stroke.

    In the current study, we describe the generation of neutralizing monoclonal antibodies against A-FABP. The specificity, selectivity, and binding affinity towards various FABPs of the identified monoclonal antibody candidate 6H2 were assessed and the structure of the complex A-FABP-6H2 and their interaction were predicted. The therapeutic effect of 6H2 on ischaemic stroke was determined using mice subjected to MCAO. The safety of chronic 6H2 administration was evaluated in healthy mice.

    2 METHODS

    2.1 Animal studies

    All animal care and experimental procedures complied with the guidelines and were approved by the Animal Ethics Committee of the University of Hong Kong (licence no. 4847-18). Animal studies are reported in compliance with the ARRIVE guidelines (Percie du Sert et al., 2020) and with the recommendations made by the British Journal of Pharmacology (Lilley et al., 2020). Efforts were made to minimize animal suffering and reduce the number of animals used per experimental group.

    A-FABP knockout (KO) mice on the C57BL/6N background were generated as previously described (Zhou et al., 2015). Eight-week-old male A-FABP KO mice, their wild-type (WT) littermates or C57BL/6N mice (from the Centre for Comparative Medicine Research (CCMR), The University of Hong Kong) were used in this study. The mice were group-housed (5 mice per cage) under a regular 12-h light/dark cycle, with temperature control (21-23oC) and 60%–70% humidity. Mice had free access to water and food (standard chow diet with 13.12% calories from fat). Our study adhered to the CAMARADES study quality score (Macleod et al., 2004) with randomization of mice to treatment groups and blinded assessment of stroke outcomes, to minimize bias.

    2.2 Transient middle cerebral artery occlusion (MCAO)

    Anaesthesia of the mice was induced by 5% isoflurane in 2 L·min−1 of O2, in a plastic anaesthesia chamber. Following induction, an adequate depth of anaesthesia for the surgery was achieved with 2% isoflurane in 30% O2 delivered by a face mask combined with a scavenger system to lower the surgeon's exposure to isoflurane. Body temperature was maintained at 37 ± 0.5°C with a thermostat-controlled heating pad. The procedure was performed using an intraluminal filament-based approach as previously described (Liao et al., 2020) and the cortical blood perfusion was monitored using laser-Doppler flowmetry. A sudden fall in blood flow to below 20%–30% of the baseline value was considered to sufficient occlusion. After 1-h of occlusion, the blood supply was restored. After 24 h or 7 days reperfusion, mice were killed humanely with an overdose of pentobarbital (Dorminal, USA).

    2.3 Determination of the detrimental effect of circulating A-FABP in ischemic stroke

    To determine the effect of circulating A-FABP in ischaemic stroke, eight-week-old male A-FABP KO mice were infused with recombinant A-FABP (rA-FABP; 2 μg·h−1, diluted with PBS), or vehicle (PBS) as control, using an osmotic pump (Alzet) for 7 days (Shu et al., 2017). Alzet osmotic pump model 2002 was used which exhibited a pumping rate of 0.5 μl·h−1. Based on our previous study, circulating A-FABP is increased and reaches a peak level (~450 ng·ml−1), 48 h after ischaemic stroke (Liao et al., 2020). As there is ~4 ml of blood in a mouse, the concentration of rA-FABP in the osmotic pump was set to 4 μg·μl−1 which was gradually released into the circulation of the A-FABP KO mice at the rate of 0.5 μl·h−1, thus the amount of rA-FABP released is 2 μg·h−1. This maintained the circulating rA-FABP concentration in A-FABP KO mice at ~ 450–500 ng·ml−1 which mimicked the elevated circulating A-FABP levels in WT mice after experimental stroke. After 7 days of infusion of rA-FABP or vehicle, mice were subjected to sham or MCAO surgery and reperfusion for 24 h or 7 days (Liao et al., 2020). A variety of systemic parameters were analysed.

    2.4 Determination of the therapeutic potential of 6H2 on ischaemic stroke and its safety

    For evaluating the effectiveness of the A-FABP neutralizing antibody 6H2 in alleviating stroke outcomes, C57BL/6N mice or A-FABP KO mice and their wildtype (WT) littermates subjected to MCAO, were intravenously injected with 6H2 (1.8 mg·kg−1 or 3.6 mg·kg−1) or mouse non-immunoglobulin G (ImmunoDiagnostics Cat# 221116, RRID:AB_3073816; IgG; 1.8 mg·kg−1) as control, 1 h after the stroke onsite and continuously every 3 days. Briefly, 6H2 concentration was calculated based on the highest level of A-FABP (~450 ng·ml−1) induced by MCAO surgery in the mice in previous study (Liao et al., 2020). The molar concentration was calculated according to the mass of A-FABP, and the concentration of 6H2 is calculated according to the molecular weight of the monoclonal antibody which is about 150. The molar ratio of concentration of 6H2 to capture A-FABP at highest concentration upon stroke is 1:1 or 1:2 as indicated in the Figures. To determine the safety of 6H2 treatment, healthy C57BL/6N mice were intravenously injected with 6H2 (1.8 mg·kg−1), or IgG as control, every 3 days for total 21 days. Blinded assessment was performed to determine a range of systemic parameters including histological/morphological analysis, neurological and behavioural analysis.

    2.5 Neurological score assessment and cerebral infarct volume measurement

    The severity of the neurological deficit was assessed every 24 h after the reperfusion, according to the following criteria: 0 = no deficit; 1 = failure to extend the forepaw; 2 = spontaneous contralateral turning; 3 = spontaneous contralateral circling; 4 = loss of walking ability; 5 = dead. The mouse brains were removed and cut into coronal slices. The infarct volume was evaluated by 2,3,5-triphenyltetrazolium chloride (TTC) staining as described previously (Liao et al., 2020). Infarct volume was quantified by Image J software (Image J 1.42 software, U.S. National Institutes of Health). The infarct size, corrected for oedema, was calculated using the formula [contralateral hemisphere (mm3)-undamaged ipsilateral hemisphere (mm3)]/area of contralateral hemisphere× 100% (Endres et al., 2000; Liao et al., 2020; Nishimura et al., 2008).

    2.6 Corner turning test and survival rate

    For corner turning test assessment, a corner was made by attaching two boards (30 cm * 20 cm *1 cm) at an angle of 30°. Mice were placed midway from the corner. When the animals reached the corner and they reared upward and then turned to either side. The number of left (ipsilateral side) turns during the trial was recorded. For the survival rate, the number of living animals in each group was recorded and the percentage was calculated.

    2.7 Brain water content

    Mouse brains were collected on the first and seventh day after MCAO surgery. The tissue samples were weighed immediately after death (wet weight). Brain samples were dried in an oven at 110°C for 48 h and the dried samples were weighed. The brain water content was calculated as [(wet tissue weight-dry tissue weight)/wet tissue weight] × 100%.

    2.8 Biochemical and immunological analysis

    Blood glucose was measured using an Accu-Check blood glucose meter (Roche, Germany). Serum triglyceride (TG) and total cholesterol (TC) concentrations were quantified using commercial kits (Stanbio Laboratory, Cat # 2100430; Cat #1010, USA) according to the manufacturer's manual. Serum free fatty acids (FFAs) concentrations were measured with the FFAs, Half Micro Test (Roche, Cat # 11383175001, Germany).

    2.9 Glucose tolerance test (GTT)

    For GTT, mice were fasted overnight (19:00 PM–9:00 AM) followed by an intraperitoneal injection of D-glucose (Sigma #G5767, 2-g·kg−1 body weight). The blood was collected from tail vein at various time points as specified in the experiment for the measurement of glucose levels using an Accu-Check blood glucose meter.

    2.10 Production of the A-FABP monoclonal antibody

    The A-FABP monoclonal antibodies were generated by immunizing BALB/c mice with the keyhole limpet hemocyanin (KLH)-conjugated synthetic human A-FABP peptide. The peptide sequence is KEVGVGFATRKVAG-Cys (aa 22–35 of human A-FABP with an additional cysteine for conjugation with KLH), which is part of the two-alpha helices of A-FABP acting as the “lid” to control entry of the FFA into the A-FABP binding pocket (Gillilan et al., 2007). The C-terminal Cys was added to allow KLH coupling of the peptide. Seventy micrograms of antigen were diluted in 500 μl sterile water and mixed with 600-μl complete Freud's adjuvant. The mixture was sonicated to a milky mixture with an ultrasonic cell pulverizer to generate the Immunogen A. On the other hand, 140 μg antigen was diluted in 1-ml sterile water, and mixed with 1-ml confidential adjuvant of AbMax Biotechnology company (Beijing). The mixture was shaken at 80–100 rpm at room temperature for 2 h to generate Immunogen B. Immunogen A was administered (i.v.) to BALB/c mice every 3 days for 2 weeks. Immunogen B was injected i.p. and i.m. injections into BALB/c mice every 3 days for 2 weeks. Blood samples were collected from tail vein of Immunogen A- or Immunogen B-treated mice 14 days after injections for ELISA to determine the presence of antibodies against A-FABP.

    The ELISA plate was coated with 1-μg·ml−1 human or mouse A-FABP recombinant protein (Shu et al., 2017) at 4°C overnight. A hundred microliters of serum (100 μl) were added to each well of the ELISA plate and incubated overnight at 4°C. The plate was then washed three times with phosphate buffered saline (PBS) followed by blocking in 5% milk in PBS at room temperature for 1 h. When the serum antibody titre reached twice the background value at 1:10000 dilution, single spleen cells from the immunized mice were fused with myeloma cells SP2/0 (CLS Cat # 400481, RRID:CVCL_2199). Ten days after cell culture in hypoxanthine-aminopterin-thymidine selection medium, monoclonal colonies were picked into 96 well plates. The concentration of antibodies in the supernatant from monoclonal colony was further determined by ELISA 7 days later. The positive hybridoma cells were injected into the peritoneal cavity of mice for 10 days to generate ascites fluid containing the desired antibody. The ascites fluids were collected and centrifuged to remove the lipid. The monoclonal antibody in the ascites fluid was further purified by protein G beads (GenScript, Cat # L00209, China) according to the manufacturer's manual.

    2.11 Determination of the binding affinities of 6H2 using Biacore

    The binding kinetics of 6H2 with human and mouse A-FABP antigens were analysed by surface plasmon resonance (SPR) using a Biacore 8 k (GE HealthCare, USA) instrument. Briefly, A-FABP antigens were covalently immobilized onto a sensor chip using standard amine coupling chemistry. The serial concentrations of the antibody 6H2 (125 nM-2 μM) were injected over the channels at a flow speed of 30 μl·min−1 for 3 min and allowed to dissociate for another 5–10 minutes before regeneration with 25 μl injection of 10 mM glycine-HCl (pH 2.5) at a flow rate of 30 μl·min−1. All the data was processed using the Biacore 8 K Evaluation Software version 4.0. Blank immobilized flow cell (without A-FABP antigens) and blank injections of running buffer in each cycle were used as double reference for Resonance Units subtraction.

    2.12 Determination of the specificity of 6H2 towards various FABPS

    To evaluate the reactivity and specificity of the monoclonal antibody with various FABPs, mouse or human A-FABP, epidermal-FABP (E-FABP, FABP5; ImmunoDiagnostics, Cat # 41040, Cat # 42040), heart-FABP (H-FABP, FABP3; SinoBiological Cat # 51233-MNAE, Cat # 12476-HNAE) recombinant protein (1 μg·ml−1) was coated on 96-well plate overnight for ELISA binding assay. A hundred microliters of 6H2 (1 μg·ml−1) were added to the well and incubated for 1 h at room temperature. The plate was then washed three times with PBS followed by incubating with 0.2 μg·ml−1 of HRP conjugated-Goat anti-Mouse IgG H + L (Proteintech Cat # SA00001-1, RRID:AB_2722565) as detection antibody for 1 h at room temperature. After washing with PBS, the plates were developed by adding TMB chromogen solution (Beyotime, Cat # P0209, China). The enzyme-induced colour reaction was stopped by adding 2 M H2SO4 to the wells. The absorbance of light intensity at 450 nm was measured by a microplate reader.

    2.13 Isolation of peritoneal macrophages

    Primary macrophages were isolated from the C57BL/6N mice peritoneum. Briefly, the abdominal area was aseptically prepared by spraying with 70% ethanol. Sterile PBS (5–10 ml) was intraperitoneally injected, and gentle abdominal massage was performed to ensure even distribution. The collected peritoneal lavage fluid was centrifuged at 400 × g for 10 minutes at 4°C. Peritoneal lavage supernatant was carefully aspirated and the peritoneal macrophage pellet was resuspended in RPMI-1640 medium (Thermo Fisher Scientific, Cat # 11875093, USA) supplemented with 10% FBS and 1% antibiotics. Cultures were incubated at 37°C in a humidified atmosphere containing 5% CO2. The isolated peritoneal macrophages were utilized for subsequent experiments.

    2.14 Determination of the neutralizing activity of monoclonal antibody

    To evaluate the neutralizing activity of monoclonal antibody 6H2 on suppressing A-FABP mediated inflammation, peritoneal macrophages were isolated from eight-week-old C57BL/6 N mice. The primary peritoneal macrophages were cultured in high glucose Dulbecco's Modified Eagle's Medium (DMEM) (Thermo Fisher Scientific, Cat # 11960044) supplemented with 10% fetal bovine serum (FBS) and 100 -U·ml−1 penicillin and 100 μg·ml−1 streptomycin at 37°C, 5% CO2 incubator. The medium was changed 6 h after seeding in the 6-wells plates or 96-wells plates. To determine the efficiency of 6H2 in suppressing A-FABP-mediated activation of JNK signalling pathway and production of MMP-9, 1-μg·ml−1 mouse A-FABP protein were pre-incubated with either 6H2 (A-FABP: 6H2, molar ratio of 1:1 or 1:10), or mouse non-immunoglobulin G (A-FABP: IgG, molar ratio of 1:1 or 1:10) in DMEM for 30 min at 37°C. Subsequently, the primary peritoneal macrophages were treated with the above mixtures for another 30 min at 37°C. The activation of JNK signalling and production of MMP-9 in the treated primary peritoneal macrophages were further analysed by Western blot analysis and ELISA.

    2.15 Western blot analysis

    Proteins were extracted from the primary cultures of mouse peritoneal macrophages. Cell lysates were separated by SDS-PAGE. Proteins were transferred to polyvinylidene difluoride membranes, and probed with primary antibodies against MMP-9 (Abcam Cat# ab38898, RRID:AB_776512), p-JNK (Cell Signaling Technology Cat# 4668, RRID:AB_823588), t-JNK (Cell Signaling Technology Cat# 9252S, RRID:AB_2250373) and β-actin (Proteintech Cat# 66009-1-Ig, RRID:AB_2687938). Membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies. The intensities of protein bands were quantified using NIH Image J software. All experimental details provided comply with the BJP Guidelines (Alexander et al., 2018).

    2.16 Determination of half-life of the monoclonal antibody 6H2

    To determine the half-life of the monoclonal antibody 6H2, 6H2 was first labelled with biotin using the Sulfo-NHS-Biotin (Thermo Fisher Scientific, Cat # A39256). Unbound biotin was eliminated by desalting on a ZebaSpin desalting spin column (Thermo Fisher Scientific, Cat # 89882). 1.8 mg·kg−1 of biotin-labelled 6H2 was intravenously injected to eight-week-old C57BL/6N mice and blood samples were collected 1 h and 1–15 days after injection. Western blot analyses were conducted to detect 6H2 heavy (6H2-H) and light (6H2-L) chains using HRP-conjugated streptavidin (Thermo Fisher Scientific, Cat # N100 USA). The contents of serum biotin-labelled 6H2-H or 6H2-L at different time points are presented as percentage, and the content of serum biotin-labelled 6H2-H or 6H2-L at 1 h after injection served as 100%, and blank serum as 0%. The half-life of 6H2 was the time corresponding to 50% the content of serum biotin-labelled 6H2-H or 6H2-L.

    2.17 Molecular docking

    The crystal structure of human A-FABP was retrieved from the Protein Data Bank (Protein Data Bank [PDB] code: 2NNQ) (Sulsky et al., 2007). The predicted structure of 6H2 was generated using SAbPred (http://opig.stats.ox.ac.uk/webapps/newsabdab/sabpred/) (Dunbar et al., 2016). The molecular docking studies were performed using Molecular Operating Environment (MOE) software 2014.9. For the preparation of A-FABP and 6H2, the water and the co-crystallized ligand were removed and hydrogens were added and subsequent minimization was adapted by the MMFF94x force field (Dunbar et al., 2016; Halgren, 1999). The A-FABP and 6H2 were then further processed to model missing loop regions, calculate protein ionization and protonate the protein structure. The prepared 6H2 was defined as the receptor and the binding site was restrained to complementarity-determining regions. The prepared A-FABP will attack the complementarity-determining regions of 6H2, after 100 trials, till the most stable docking complexes are reached.

    2.18 Data and statistical analysis

    Data and statistical analysis in this study complied with the BJP's recommendations and requirements on experimental design and analysis (Curtis et al., 2022). Statistical analysis was performed using GraphPad Prism (v.8.4.3) software (GraphPad Software, San). For all animal experiments, only male mice were utilized to minimize potential sex-related variations, and technical replicates were incorporated to enhance the reliability of individual measurements. Any mice experiencing excessive bleeding during MCAO were excluded from further tissue collection and data analysis. Group sizes represent the number of independent animals or cell cultures, and statistical analyses were based on these independent values. The sample size was at least five animals in each group to ensure an adequate number for statistical testing while minimizing the animal suffering and number used in each experiment. The n number used to determination of neurological score, survival rate and potential adverse effect of 6H2 were based on our previous related study (Liao et al., 2020). Data normality was accessed by the Shapiro–Wilk test. All data are expressed as mean ± SD. No data were excluded from statistical analysis and presentation. Only sample size more than or equal to 5 were subjected to statistics analysis. For neurological score and survival rate, Mann Whitney Wilcoxon rank-sum test and Log-rank test were used for data analysis, respectively. Student's t test was used to compare the difference between two groups with continuous variables. One-way ANOVA test followed by Dunnet's post hoc test was used for multiple comparisons. Significance was established at a threshold of P < 0.05. In multigroup studies with parametric variables, post hoc tests were conducted only if F in ANOVA achieved the statistical significance (P < 0.05) and there was no significant variance in homogeneity. All data conformed to the statistical test assumptions.

    2.19 Materials

    The recombinant mouse A-FABP (Cat # 42030), human A-FABP (Cat # 41013); mouse E-FABP (Cat # 42040), human E-FABP (Cat # 41040) were supplied by ImmunoDiagnostics (Hong Kong, China). Mouse H-FABP (Cat # 51233-MNAE) and human H-FABP (Cat # 12476-HNAE) were supplied by SinoBiological (Beiging, China).

    2.20 Nomenclature

    Key protein targets and ligands in this article are hyperlinked to corresponding entries in the IUPHAR/BPS Guide to PHARMACOLOGY (http://www.guidetopharmacology.org) and are permanently archived in the Concise Guide to PHARMACOLOGY 2021/22 (Alexander, Fabbro, Kelly, Mathie, Peters, Veale, Armstrong, Faccenda, Harding, Pawson, Southan, Davies, Beuve, et al., 2021; Alexander, Fabbro, Kelly, Mathie, Peters, Veale, Armstrong, Faccenda, Harding, Pawson, Southan, Davies, Boison, et al., 2021; Alexander, Kelly et al., 2021).

    3 RESULTS

    3.1 Circulating A-FABP exaggerates ischaemia-induced cerebral injury

    Circulating A-FABP was elevated in humans and mice in response to ischaemic stroke (Liao et al., 2020; Tso et al., 2011). To determine the effect of circulating A-FABP on stroke outcome, 8-week-old A-FABP KO mice were infused with recombinant A-FABP (rA-FABP) at 2 μg·h−1 or vehicle for seven consecutive days followed by MCAO surgery for 1 h and reperfusion for 23 h or 7 days (Figure 1a). TTC staining showed that replenishment of rA-FABP significantly increased the MCAO surgery-induced infarct volume in A-FABP KO mice (Figure 1b). Consistent with this finding, the brain oedema and neurological deficits of rA-FABP-infused A-FABP KO mice were significantly increased and accompanied by reduced survival rate (Figure 1c–e), compared with the vehicle-infused mice. These data confirmed the deleterious role of circulating A-FABP in mediating cerebral ischaemia injury.

    Details are in the caption following the image
    Circulating A-FABP exacerbates cerebral ischaemia injury. Eight-week-old A-FABP KO mice were infused with recombinant A-FABP (rA-FABP; 2 μg·h−1) or vehicle (veh; PBS) for 7 days followed by MCAO surgery and reperfusion. The infarct volume and brain water content were measured at 24 h after MCAO or sham operation. Neurological score and survival rate were monitored for 7 days. (a) Schematic diagram for the schedule of rA-FABP infusion using osmotic pump and MCAO surgery in mice. (b) Representative photographs of coronal brain sections of A-FABP KO mice infused with Veh or rA-FABP stained with TTC at 24 h after MCAO or sham operation (n = 5) and the relative infarct volume. (c,d) Percentage of water content (n = 5) and (d) neurological score (n = 8) of A-FABP KO mice infused with Veh or rA-FABP at 24 h after MCAO or sham operation. (e) The survival rate of rA-FABP- or vehicle-infused A-FABP KO mice after MCAO surgery (n = 9). Data are presented as mean±SD, with (b-d) individual values. *P < 0.05, significantly different as indicated.

    3.2 Development and characterization of A-FABP mAb

    As circulating A-FABP exaggerated ischaemic stroke outcome, we next generated neutralizing monoclonal antibodies (mAbs) against A-FABP. In order to generate antibodies that can potentially inhibit the biological activity of A-FABP, we selected a peptide fragment from the two α-helices of human A-FABP that controls the entry of its substrates into the internal binding pocket to immunize mice (Gillilan et al., 2007). Murine anti-human A-FABP mAbs were generated by conventional hybridoma techniques. Among the mAbs generated, pre-incubation of the positive hybridoma clone named 6H2 with rA-FABP (in 1:10 ratio) markedly attenuated A-FABP elicited JNK/c-Jun activation in mouse primary peritoneal macrophages (Figure S1). Thus, 6H2 was selected for further characterization in this study.

    3.3 Binding affinities and neutralizing activity of 6H2

    The real-time binding affinities of A-FABP mAb 6H2 to human and mouse rA-FABP were determined by Biacore 8 K system. 6H2 showed moderately fast association rates, but very slow dissociation rates to both hA-FABP and mA-FABP with the steady state affinity KD values as 7.41E-10 M (0.74 nM) and 2.51E-09 M (2.51 nM), respectively, which suggests high binding affinities and that the antibody–antigen complex is highly stable over time (Figure 2a–c). ELISA was performed to evaluate the specificity of the 6H2 towards FABPs of different species. The results demonstrated that 6H2 is highly specific to human and mouse A-FABP, with no cross reactivity with human and mouse epidermal-FABP (E-FABP) and heart-FABP (H-FABP) (Figure 2d).

    Details are in the caption following the image
    Characterization of anti-A-FABP monoclonal antibody 6H2. (a) BIAcore analysis of the binding affinity of 6H2 with mouse A-FABP. (b) BIAcore analysis of the binding affinity of 6H2 with human A-FABP. (c) the kinetic parameter association rate constant (ka) and dissociation rate constant (kd) were determined by BIAevaluation software. KD values were calculated from ka and kd values. (d) Antigen-binding specificity of 6H2 with recombinant mouse or human A-FABP, E-FABP, H-FABP evaluated using ELISA (n = 4). (e) Neutralizing activity of 6H2 or mouse non-immunoglobulin G (IgG) on rA-FABP-induced activation of JNK and production of MMP-9. The band intensity of each protein relative to their respective control protein or β-actin, and normalized to their relative control (Ctrl) (n = 5). (f) Half-life of 6H2 in serum. The contents of serum biotin-labelled 6H2-H or 6H2-L at different time points were presented as percentage, and the content of serum biotin-labelled 6H2-H or 6H2-L at 1 h after injection served as 100%, and blank serum as 0%. The half-life of 6H2 was the time corresponding to 50% the content of serum biotin-labelled 6H2-H or 6H2-L (n = 5). Data are presented as mean ± SD, with (d, e) individual values. # P < 0.05, significantly different from Ctrl; * P < 0.05, significantly different from A-FABP.

    A-FABP forms a finely tuned positive feedback loop with activator protein 1 (AP-1) and JNK to modulate the inflammatory response induced by lipopolysaccharide in macrophages (Hui et al., 2010). During ischaemic stroke, elevated A-FABP in macrophages also activates JNK/c-Jun signalling pathway leading to the increased production of MMP-9, thus exaggerating BBB disruption (Liao et al., 2020). To evaluate the neutralizing activity of 6H2 on the biological action of A-FABP, Western blots were performed to determine the p-JNK/JNK ratio and MMP9 production in mouse primary peritoneal macrophages after treatment with rA-FABP pre-incubated with 6H2 or mouse IgG as a negative control (1:1 and 1:10 ratio). Treatment with rA-FABP (1 μg·ml−1) alone significantly induced the p-JNK/JNK ratio in macrophages which was markedly suppressed when rA-FABP was pre-incubated with 6H2 in 1:1 ratio. A further reduction of p-JNK/JNK ratio was observed when rA-FABP was pre-incubated with 6H2 in 1:10 ratio (Figure 2e). Consistent with these data, A-FABP-induced expression of MMP-9 was significantly suppressed by 6H2 treatment in 1:1 and 1:10 ratio (Figure 2e). However, pre-incubation with mouse IgG did not show any suppressive effect on rA-FABP-mediated p-JNK/JNK ratio. These data indicated that 6H2 exhibited strong neutralizing activity against A-FABP.

    Next, the half-life of the 6H2 was evaluated. C57BL6/N mice were injected intravenously (i.v.) with biotin-labelled-6H2 or -mouse IgG at a concentration of 1.8 mg·kg−1 followed by collection of blood samples at different time points. Western blot analysis demonstrated that the 6H2 mAb has a half-life of 2–3 days and could be found in the serum for about 2 weeks (Figure 2f).

    3.4 Prediction of the A-FABP-6H2 complex and target epitope

    To assess and characterize the structure of A-FABP-6H2 complex and the target epitope, protein–protein docking analysis was performed using Molecular Operating Environment (MOE) software (Figure 3a). Prediction of the target epitope showed that 6H2 heavy chain variable regions (VH, purple) and light chain variable regions (VL, green) bound an epitope that is spread out over the secondary structure elements α1 and α2 (red) of A-FABP, and these two consecutive α-helices are the peptides used for 6H2 antibody generation. In addition, the slow dissociation of the 6H2 can be explained by the predicted target epitope. Asn33 in the heavy chain of 6H2-VH forms a hydrogen bond with A-FABP backbone at Phe27, and the other two hydrogen bonds are formed through the Asn28 and Asn30 in the light chain of 6H2 with the A-FABP backbone at Aps18 and the carboxyl oxygen at Asn15, respectively (Figure 3b). The structure also showed that 6H2 bound directly to the ‘lid’ (Ser 14 to Ala 37) of the β-barrel formed by the two consecutive α-helixes (red) in A-FABP (Burak et al., 2015), which has been postulated to control the access of substrates to the binding pocket (Gillilan et al., 2007). Cumulatively, the docking analysis suggested that the neutralizing activity of 6H2 may arise from interference with the lipid binding and transport function of A-FABP.

    Details are in the caption following the image
    Characterization of the structure of the A-FABP-6H2 complex and the target epitope. (a) Ribbon diagram depicting the secondary structure elements of the 6H2 variable region of light chain (purple) and heavy chain (green) in complex with A-FABP (α-helix: red, β-sheet: yellow). (b) High-resolution mapping of the 6H2 epitope (variable region of heavy chain: green; variable region of light chain: purple) on β barrel (Ser14 to Ala37 of A-FABP) (red). Interacting residues in both molecules are shown as stick models. Hydrogen bonds are dashed lines (red arrow).

    3.5 Treatment with 6H2 attenuates MCAO-induced cerebral injury

    To evaluate the therapeutic potential of the A-FABP mAb 6H2 on cerebral ischaemia, C57BL/6N mice were subjected to MCAO surgery for 1 h followed by intravenous injection of 6H2 or mouse IgG followed by reperfusion for 23 h or 7 days. The dosage of 6H2 was calculated based on the MCAO-induced serum level of A-FABP previously reported (Liao et al., 2020). The TTC staining result at 24 h after MCAO onset showed that treatment of 1.8-mg·kg−1 6H2 significantly reduced the ischaemia-induced infarct volume, compared with that in mice treated with 1.8-mg·kg−1 IgG (~12% vs. 40%) (Figure 4a) which aligned with ~30% improvement in the stroke outcome observed in most of the previous therapeutics research in the field of experimental stroke (Macleod et al., 2004). Further increase of 6H2 dosage to 3.6 mg·kg−1 did not show a further reduction in infarct volume (Figure 4a), while being associated with a lower survival rate in mice, compared with those treated with 1.8-mg·kg−1 6H2 (Figure S2). Thus, 1.8-mg·kg−1 6H2 was used in the following studies. The ELISA result indicated that treatment with 6H2 significantly reduced ischaemic stroke-induced elevation of serum A-FABP in mice, compared with those treated with mouse IgG treatment (Figure 4b). This data suggested that 6H2 effectively captured circulating A-FABP leading to a reduced free A-FABP to be detected by the ELISA. Also, 6H2-treated mice exhibited a significant decrease in ischaemia-induced brain oedema, when compared with mouse IgG-treated mice (Figure 4c). Long-term neurological deficits during 7 days after MCAO were also improved, accompanied by an increased survival in the 6H2-treated mice (Figure 4d,e). Furthermore, MCAO-induced cerebral expression of pro-inflammatory cytokines in 6H2-treated mice were markedly attenuated (Figure 4f), compared with that in IgG-treated mice. Importantly, treatment with 6H2 did not show a further alleviation in the MCAO-induced cerebral injury in A-FABP deficient mice when compared with that in wildtype mice suggesting the therapeutic effect of 6H2 ws due to its specific targeting of circulating A-FABP (Figure S3). These data suggested that treatment with 6H2 neutralized MCAO-induced circulating A-FABP along with reduced pro-inflammatory cytokine secretion and alleviated cerebral injury with improved functional outcome. These data underscore the broad potential applicability of A-FABP neutralization to combat injury due to cerebral ischaemia.

    Details are in the caption following the image
    Treatment with 6H2 attenuates cerebral ischaemia injury in mice. Eight-week-old C57BL/6N mice were subjected to MCAO surgery followed by intravenous injection of mouse non-immunoglobulin G (IgG; 1.8 mg·kg−1) or A-FABP neutralizing mAb 6H2 (1.8 or 3.6 mg·kg−1) 1 h and continuously every 3 days after MCAO surgery or sham operation. Brain infarct volume, brain water content and inflammatory cytokines expression levels were measured at 24 h after MCAO. Neurological scores and survival rate were monitored every day for 7 consecutive days. (a) Representative photograph of coronal brain sections of mice treated with IgG (1.8 mg·kg−1) or 6H2 (1.8 or 3.6 mg·kg−1) after MCAO surgery stained with TTC at 24 h (n = 5) and the relative infarct volume. (b) Dynamic change of serum A-FABP levels in mice treated with IgG (1.8 mg·kg−1) or 6H2 (1.8 mg·kg−1) after MCAO surgery (n = 8). (c–e) Percentage of brain water content (c) (n = 5), neurological score (d) (n = 8) and survival rate (e) (n = 9) of the mice subjected to MCAO surgery treated with IgG (1.8 mg·kg−1) or 6H2 (1.8 mg·kg−1) after MCAO surgery. (F) Relative mRNA levels of inflammatory cytokines in the brain of mice treated with IgG (1.8 mg·kg−1) or 6H2 (1.8 mg·kg−1) after MCAO surgery (n = 8). Data are presented as mean ± SD. *P < 0.05, significantly different as indicated.

    3.6 Treatment with 6H2 alleviates A-FABP-induced BBB disruption during the MCAO surgery

    Our previous study showed that the A-FABP-JNK-MMP-9 signalling pathway exaggerated the cerebral ischaemic injury by promoting MCAO-induced BBB leakage (Liao et al., 2020). Thus, the effect of 6H2 in rescuing the BBB integrity was evaluated. 6H2 or mouse IgG was intravenously injected into mice 1 h after MCAO surgery. Ischaemia-induced BBB leakage was decreased significantly in the 6H2-treated mice as indicated by the reduced permeability to Evans blue dye compared with the mouse IgG-treated controls (Figure 5a). Ischaemia-induced degradation of tight junction (TJ) proteins zonula occluden (ZO-1) and occludin were also significantly attenuated by the 6H2 indicating the protective effect of 6H2 in the BBB leakage (Figure 5b). Moreover, the MCAO-induced activation of JNK/c-Jun signalling pathway and the cerebral and circulating levels of MMP-9 during ischaemic stroke were also markedly inhibited by 6H2 treatment (Figure 5c,d). These data indicated that treatment of 6H2 alleviated ischaemic stroke outcome in mice, possibly through suppressing the A-FABP-JNK-MMP9 signalling pathway, thus attenuating BBB disruption.

    Details are in the caption following the image
    Treatment with 6H2 attenuates the cerebral ischaemia-induced blood brain barrier disruption. Eight-week-old C57BL/6N mice were subjected to MCAO surgery or sham operation followed with intravenous injection of non-immunoglobulin G (IgG) (1.8 mg·kg−1) or A-FABP neutralizing mAb 6H2 (1.8 mg·kg−1) 1 h after MCAO surgery or sham operation. Mice were killed after 24 h reperfusion and blood brain barrier integrity was assessed. (a) Representative photographs of the brain of mice stained with Evans blue 24 h after MCAO or sham operation (n = 5). (b) Representative immunoblots of tight junction (TJ) proteins in the brains of mice treated with IgG or 6H2 after MCAO or sham operation and the band intensity of each protein relative to GAPDH (n = 6). (c) Representative immunoblots of p-JNK (Thr183/Tyr185) and p-c-Jun (Ser 63/73) in the infarct core of brains of mice treated with IgG or 6H2 after MCAO or sham operation and the band intensity of each protein relative to their respective control protein or GAPDH; (n = 5). (d) Serum levels of MMP-9 in mice treated with IgG or 6H2 (n = 10). Data are presented as mean ± SD. *P < 0.05, significantly different as indicated.

    3.7 Long-term 6H2 treatment does not affect the systemic metabolism in healthy mice

    To explore if treatment with 6H2 leads to any side effects, mice were intravenously injected with 6H2 (1.8 mg·kg−1) or mouse IgG (1.8 mg·kg−1) every 3 days for 21 days and basic metabolic parameters were monitored (Figure 6a). No significant changes were observed in the body weight (Figure 6b), fat (Figure 6c) and lean mass (Figure 6d) and fasting glucose (Figure 6e) in 6H2-treated mice when compared with mouse IgG-treated mice. Glucose metabolism was monitored using a glucose tolerance test (GTT). Treatment with 6H2 did not alter the glucose tolerance, as demonstrated by glucose disposal curves similar to those of mice treated with IgG (Figure 6f). As A-FABP is a lipid chaperone related to lipid homeostasis (Li et al., 2021), the serum levels of total cholesterol (TC), triglyceride (TG) and free fatty acid (FFA) were measured (Figure 6g–i). No significant difference in these parameters was observed between the mouse IgG- and 6H2-treatment groups. In summary, these data supported the finding that 6H2 is a safe neutralizing mAb for the treatment of ischaemia-induced cerebral injury in mice.

    Details are in the caption following the image
    Metabolic profiles of mice treated with 6H2 for 21 days. Eight-week-old C57BL/6N mice were injected intravenously with A-FABP neutralizing mAb 6H2 (1.8 mg·kg−1) or mouse non-immunoglobulin G (IgG; 1.8 mg·kg−1) every 3 days. Body weight and body composition were measured every 5 days. Fasting glucose was tested at day 5 and day 15. Glucose tolerance test (GTT) was performed at day 15. Total cholesterol (TC), triglycerides (TG) and free fatty acid (FFA) were determined in mice before and after treatment with 6H2 or IgG at indicated time points. (a) The time course of monitoring metabolic profiles of mice treated with 6H2 or IgG. (b) Body weight of mice before and during the period of 6H2 or IgG treatment. (c, d) Fat mass and lean mass amount in mice before and during the period of 6H2 or IgG treatment. (e) Fasting glucose in mice before and after 6H2 or IgG treatment. (f) Blood glucose levels in mice during GTT performed after 15 days of 6H2 or IgG treatment. (g–i) Total cholesterol (TC), total triglyceride (TG) and free fatty acid (FFA) levels in mice before and after 6H2 or IgG treatment. Data are presented as means ± SD; n = 10.

    4 DISCUSSION

    Ischaemic stroke is one of the major causes of death and disability in the world, while there is a lack of effective treatments. Limitations and the drawback of using the only FDA-approved pharmacotherapy tPA or mechanical thrombectomy for the treatment of ischaemic stroke remain a major challenge to the global economic and health care system.

    In recent years, monoclonal antibody therapy is becoming more and more common for the treatment of various diseases such as cancers (Pento, 2017), autoimmune diseases (Arzoo et al., 2002) and neurological disorders (Sun et al., 2023), because of their better performance in terms of specificity, safety and pharmacokinetics, compared to small-molecule chemical drugs. Though a number of monoclonal antibodies were developed targeting signalling cascades and endogenous molecules modulating post-stroke inflammation, neuronal repair and regeneration, and showed beneficial effects in animal studies, none of them is approved for clinical use in stroke. The anti-ICAM-1 mAb (Enlimomab) (Investigators, 2001) which reduces leukocyte adhesion and infarct size in experimental animal also failed in clinical trials, due to adverse effects including increased primary infections and fever resulting in increased mortality (Investigators, 2001). Thus, developing a safe and effective mAb for ischaemic stroke that improves outcome and rehabilitation is warranted.

    A-FABP is an adipocyte-secreted protein which is recognized as a biomarker and the therapeutic target of various cardio-metabolic diseases (Hotamisligil et al., 1996; Makowski et al., 2001). We demonstrated that circulating A-FABP was elevated in patients with ischaemic stroke (Liao et al., 2020) and was closely associated with early death (Tso et al., 2011). Circulating A-FABP is also an independent prognostic biomarker (Tu et al., 2017) and is positively associated with the infarct size and circulating MMP-9 levels of ischaemic stroke patients (Liao et al., 2020). These data are in line with our current findings that infusion of rA-FABP in A-FABP deficient mice exaggerated ischaemic stroke injury with increased infarct size and decreased survival rate. Incubation with rA-FABP also induced MMP-9 expression in mouse peritoneal macrophages. Ischaemia-mediated elevation of A-FABP leads to the disruption of BBB and post-ischaemic inflammation in mice through inducing the production and release of MMP-9 and pro-inflammatory cytokines from circulating monocytes and cerebral macrophages (Liao et al., 2020). Notably, cerebral overexpression of A-FABP in A-FABP deficient mice led to an increased permeability of Evan blue dye in the ischaemic brain which was attenuated by treating the mice with SB-3CT, an inhibitor of MMP-9 (Liao et al., 2020). These data indicate that the detrimental effects of the circulating A-FABP in ischaemic stroke in mice are similar to those in humans, and contributes, at least, to some particular outcomes of ischaemic stroke. We also demonstrated that replenishment of circulating A-FABP exaggerated the stroke outcome in A-FABP-deficient mice in the current study. Therefore, blocking circulating A-FABP may represent an effective treatment for ischaemic stroke.

    In order to target circulating A-FABP, we generated and evaluated several anti-A-FABP mAbs and identified one candidate 6H2 with high selectivity and affinities towards A-FABP as a potential therapeutic to ameliorate ischaemic stroke injury. Single-dose treatment with 6H2 targeting elevated circulating A-FABP in WT mice subjected to MCAO showed a very similar beneficial effect as that in the A-FABP-deficient mice treated with IgG whereby both showed a similar percentage of the infarct volume ~12% and neurological deficit. Increased survival rate in 6H2-treated WT mice was also associated with improved BBB integrity, reduced MMP9 expression and post-stroke inflammation comparing to their relative controls.

    Post-stroke inflammation is known to exacerbate cell damage and cause further disruption to the BBB which results in exaggerated stroke outcome. Toll-like receptor 4 (TLR4) is significantly activated upon ischaemic stroke, contributing to post-stroke inflammation and brain tissue damage (Caso et al., 2007). A-FABP is a downstream target of TLR4 which potentiated JNK-c-Jun-mediated (Hui et al., 2010) and NF-kB-mediated (Makowski et al., 2005) cytokine production. Elevated A-FABP also associated with increased cytokine production after stroke insult (Liao et al., 2020). Thus, it is possible that 6H2 treatment may also alleviate TLR4-induced stroke outcomes.

    Though the exact mechanism on how circulating A-FABP mediates brain tissue injury following ischaemic stroke was not revealed in the current study, our findings showed that both adenovirus-mediated overexpression of A-FABP (Hui et al., 2010; Liao et al., 2020) or treatment with exogenous rA-FABP (Supplementary figure 1) activates pro-inflammatory signalling, via JNK/c-Jun, in macrophages. Our previous studies also demonstrated that rA-FABP can directly diffuse into different cells to exert its functions such as enhancing energy expenditure in brown adipocytes (Shu et al., 2017) and induces the expression of TGF-β in activated hepatic stellate cells (Wu et al., 2021). These data imply that both circulating (extracellular) and intracellular A-FABP may act through the same signalling pathway (JNK/c-Jun) to induce MMP-9 expression, thus disrupting the BBB and exaggerating ischaemic stroke outcome. Therefore, treatment with 6H2 neutralizes the activity of circulating A-FABP could also prevent it from diffusing into cells and exerting detrimental action intracellularly. As BBB integrity is compromised upon ischaemic stroke insult, it is also possible that mAb 6H2 enters the brain parenchyma and neutralizes cerebral A-FABP, thus alleviating stroke outcome.

    In addition, treatment with 6H2 every 3 days after MCAO surgery improved the long-term (7 days) survival rate in the mice. Intravenous injection of 6H2 in healthy C57BL/6N mice for 21 days did not alter any systemic metabolic parameters indicating the safety of 6H2 usage. Taken together, all these data implicate that neutralizing circulating A-FABP with 6H2 is not only an effective but also a safe therapeutic approach to increase the treatment time-window and post-stroke recovery of neurons, rescuing the stroke patients from permanent disability and death.

    Our protein–protein docking analysis revealed that binding of 6H2 to A-FABP was not likely to cause distortion in the structure of A-FABP. The peptide region of A-FABP selected for antibody generation forms hydrogen bond with amino acid residue (Asn33) of the complementarity-determining regions of 6H2. The structure of the A-FABP-6H2 complex also showed that 6H2 binds directly to the ‘lid’ of the β-barrel formed by the two consecutive α-helices. The location of the epitope recognized by 6H2 suggests its potential interference with the A-FABP lipid binding pocket (Gillilan et al., 2007). Thus, these data indicated that 6H2 exerted its therapeutic effects by neutralizing the biological activity of A-FABP, probably through hindering the interaction between A-FABP and its substrates, including free fatty acids (FFA).

    By contrast, CA33 is also a mAb targeting secretory A-FABP, which exerts anti-diabetic effects in mice by decreasing hepatic glucose production and enhancing peripheral insulin sensitivity (Burak et al., 2015). Compared with our newly developed 6H2, CA33 exhibited low affinities, in the micromolar range, towards human and mouse A-FABP and cross-reactivity to H-FABP (FABP3). Importantly, CA33 did not bind to the “lid” of the β-barrel or the hinge and did not affect the binding of lipid to A-FABP, suggesting its activity is independent of affecting A-FABP-lipid binding (Burak et al., 2015). These data suggest that the therapeutic effects of the two mAbs, 6H2 and CA33, may be different and specific, according to their target epitopes. It also suggests the detrimental effects of A-FABP in various diseases may be attributed to the function of a particular epitope of A-FABP. Further investigations are warranted to verify these possibilities.

    Circulating levels of FFA are positively associated with risk factors for stroke (Burak et al., 2015; Kim et al., 2012) and a change in the FFA profile with a sudden increase of the levels of very long n-3 and n-6 FFA including 20:4n-6, 22:4n-6, 22:5n-6 and 22:6n-3 was observed after 2 h of MCAO surgery in mice. These changes in plasma FFA profile are consistent with FFA released from brain phospholipids hydrolyzed following an ischaemic insult (Kim et al., 2012). Circulating FFA was also an independent predictor of poor functional outcome and recurrence in patients with ischaemic stroke (Niu et al., 2017). Notably, circulating A-FABP is raised 2 h after MCAO surgery (Liao et al., 2020). As 6H2 alleviates stroke outcome probably through suppressing the interaction between FFA and A-FABP, further studies are warranted to verify if the detrimental role of A-FABP in stroke is at least partly FFA-dependent and if the level of FFAs is a key determinant in the severity of ischaemic stroke outcome, through interacting with A-FABP.

    In summary, we have generated a novel mAb 6H2 with high affinity, high specificity and different target epitope, which can be a potential therapeutic agent alleviating ischaemic stroke injury by neutralizing circulating A-FABP and possibly also cerebral A-FABP. We will further focus on identifying the structural determinant of 6H2 action and its humanization. It has to be noted that t-PA is the only therapy identified by experimental setting that has been shown to improve ischaemic stroke outcome in clinical trials so far. Further investigations regarding the clinical use of the mAb 6H2, identified here, are warranted.

    AUTHOR CONTRIBUTIONS

    Ruby Lai Chong Hoo and Junlei Chang completed conception and design of this study. Boya Liao and Shilun Yang drafted the manuscript, conducted experiments and prepared the figures. Leiluo Geng, Jiuyu Zong, Zixuan Zhang, Mengxue Jiang, Xue Jiang and Simeng Li conducted experiments. Ruby Lai Hoo and Junlei Chang edited the manuscript. Aimin Xu, Ruby Lai Chong Hoo and Junlei Chang approved the final version of the manuscript.

    ACKNOWLEDGEMENTS

    This study was supported by the Science Technology and Innovation Commission of Shenzhen Municipality (SGLH20180625142404672, JCYJ20210324115800003, JCYJ20210324101401004), the Innovation and Technology Fund, Guangdong-Hong Kong Technology Cooperation Funding Scheme (GHP/079/18SZ), Area of Excellence Scheme (AOE/M-707/18) from Research Grants Council of the Hong Kong Special Administrative Region, National Natural Science Foundation of China (82104167), CAS-Croucher Funding Scheme for Joint Laboratories, International collaboration project of Chinese Academy of Sciences (172644KYSB20200045), and Guangdong Innovation Platform of Translational Research for Cerebrovascular Diseases.

      CONFLICTS OF INTEREST STATEMENT

      Please give the state of the patent application –has it been granted yet? Which country(ies) – China? USA? Europe? - are involved. Also please list the authors of this MS who are also named in the patent application.

      DECLARATION OF TRANSPARENCY AND SCIENTIFIC RIGOUR

      This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research as stated in the BJP guidelines for Design and Analysis, Immunoblotting and Immunochemistry and Animal Experimentation, and as recommended by funding agencies, publishers and other organizations engaged with supporting research.

      DATA AVAILABILITY STATEMENT

      Data openly available in a public repository that issues datasets with DOIs.