Ankylosing spondylitis PET imaging and quantifications via P2X7 receptor-targeting radioligand [18F]GSK1482160

Ankylosing spondylitis (AS) is a chronic inflammatory disease of the axial spine; however, the quantitative detection of inflammation in AS remains a challenge in clinical settings. We aimed to investigate the feasibility of using a specific P2X7R-targeting 18F-labeled tracer [18F]GSK1482160 for positron emission tomography (PET) imaging and the quantification of AS. The radioligand [18F]GSK1482160 was obtained based on nucleophilic aliphatic substitution. Dynamic [18F]GSK1482160 and [18F]FDG micro-PET/CT imaging were performed on AS mice (n = 8) and age-matched controls (n = 8). Tracer kinetics modeling was performed using Logan’s graphical arterial input function analysis to quantify the in vivo expression of P2X7R. The post-PET tissues were collected for hematoxylin–eosin (H&E), immunohistochemical (IHC), and immunofluorescence (IF) staining. [18F]GSK1482160 PET/CT imaging revealed that the specific binding in the ankle joint and sacroiliac joint (SIJ) of the AS at 8 weeks group (BPNDankle−AS−8W (non-displaceable binding potential of the ankle) 3.931 ± 0.74; BPNDSIJ−AS−8W (BPBD of the SIJ) 4.225 ± 0.84) were significantly higher than the controls at 8 weeks group (BPNDankle−Ctr−8W 0.325 ± 0.15, BPNDSJJ−Ctr−8W 0.319 ± 0.17) respectively, and the AS at 14 weeks group (BPNDankle−AS−14W 12.212 ± 2.25; BPNDSJJ−AS−14W 13.389 ± 3.60) were significantly higher than the controls at 14 weeks group (BPNDankle−Ctr−14W 0.204 ± 0.16, BPNDSJJ−Ctr−14W 0.655 ± 0.35) respectively. The four groups had no significant difference in the [18F]FDG uptake of ankle and SIJ. IHC and IF staining revealed that the overexpression of P2X7R was colocalized with activated macrophages from the ankle synovium and spinal endplate in mice with AS, indicating that quantification of P2X7R may contribute to the understanding of the pathogenesis of inflammation in human AS. This study developed a novel P2X7R-targeting PET tracer [18F]GSK1482160 to detect the expression of P2X7R in AS mouse models and provided powerful non-invasive PET imaging and quantification for AS.


Introduction
Ankylosing spondylitis (AS) is an insidiously progressive and debilitating form of arthritis involving the axial skeleton, particularly the sacroiliac joints (SIJ) [1].AS is a disease common in young adults, with a peak incidence between the ages of 20 and 30.Approximately 23,000,000 people in the world are affected by AS and the trend continues to increase [2].SIJ are most often involved in the early stage of AS, and axial bone fusion can occur in the late stage, resulting in disability and deformity, thereby causing a heavy economic burden on the family and society [3].Although the presence of the major histocompatibility complex class I allele human Shiyanjin Zhang, Yifan Qiu and Lihua Huang equally contributed to this work.leukocyte antigen B27 (HLA-B27) accounts for the majority of genetic risk, evidence from recent studies shows that it accounts for only 20-25% of total heritability and 40% of genetic risk [4].Less than 5% of HLA-B27 carriers in the general population suffer from AS [5].
The accurate diagnosis of AS remains a global problem.At present, in clinical practice, the diagnosis of AS is based on evidence from multiple experimental setups, including blood laboratory testing (HLA-B27 and erythrocyte sedimentation rate) and detection of anatomic abnormalities through X-rays, computed tomography (CT), and magnetic resonance imaging (MRI).These need to be combined with a series of severe clinical symptoms of the patient, such as back pain, spinal deformity, and disability for a comprehensive diagnosis [6].Existing diagnostic methods have limitations, such as a long diagnostic cycle, low accuracy, and inability to quantify.Timely and accurate diagnosis is of great significance to guide the clinical treatment of AS and determine whether the disease process can be delayed with drugs such as non-steroidal anti-inflammatory drugs, disease-modifying anti-rheumatic drugs and targeted biological agents [7].Accurate localization and quantitative detection of AS lesions will help to determine whether AS is in the active or stationary phase and whether spinal osteotomy and orthopedic surgery are suitable.Therefore, a new, sensitive, and non-invasive examination method for AS diagnosis is urgently needed.Accurate localization and quantitative detection of AS are of great significance to evaluate the severity and prognosis of the disease.
Positron emission tomography (PET) is an emerging detection technology, and it is now widely used in the early diagnosis, staging, and restaging of various diseases [8].At present, [ 18 F]fluorodeoxyglucose ([ 18 F]FDG) PET is the most commonly used PET/CT imaging agent in clinics.However, no abnormal [ 18 F]FDG uptake was observed in the inflammatory lesions of AS [9,10].Therefore, it is important to develop more specific molecular PET/CT imaging agents, which not only better understand the biological characteristics of AS, but also have a potential impact on evaluating new clinical drugs against AS, noninvasively and longitudinally.The P2X7 receptor (P2X7R) is an ATP-gated ion channel that can be expressed by various immune cells, including monocyte-derived cell lines, T and B lymphocytes, and macrophages [11].When the extracellular ATP concentration is high, the P2X7R is activated, which in turn activates the downstream NLRP3 inflammasome, which activates caspase-1, promotes the maturation and secretion of IL-1β, and produces corresponding mature cytokines [12].Studies have revealed an association between genetic variations in the P2X7R gene and AS susceptibility [13], and that overexpression of the P2X7R may be involved in the progression of AS.In this study, we successfully synthesized a P2X7R-targeting PET probe [ 18 F]GSK1482160 by nucleophilic aliphatic substitution, with significantly improved radiochemical yield (RCY) and radiometabolism stability compared with previous methods [14][15][16].A mouse spinal arthritis model was constructed with a proteoglycan and dimethyldioctadecylammonium (DDA) adjuvant, and [ 18 F]GSK1482160 micro PET/CT imaging was performed to investigate the feasibility of the P2X7R for AS.

In vitro and in vivo stability
The in vitro stability of [ 18 F]GSK1482160 was assessed in both saline and fetal bovine serum (FBS).[ 18 F]GSK1482160 (1.34 − 3.35 MBq) was incubated in 0.5 mL of saline and FBS at 37 °C separately for different time points (0.5, 1, 1.5, 2, and 2.5 h).After incubation, 50 μL aliquots of the saline samples were.measured by radio-HPLC directly.Correspondingly, 50 μL aliquots of the FBS sample were removed and mixed with 100 μL of ethanol to precipitate the serum protein at each time point and centrifuged at 10 000 g for 2 min.After centrifugation, the supernatant was filtrated through 0.22 μm filters (Jinteng Co., Ltd., China).Then, the filtrated samples were analyzed using radio-HPLC (Fig. 1c, d).The in vivo stability was performed with radio metabolite analyses of the blood and liver of healthy male Sprague-Dawley (SD) rats.Healthy male SD rats (150 − 250 g) were anesthetized with 2 − 3% isoflurane/oxygen, and [ 18 F]GSK1482160 (37 − 74 MBq) was administered via an intravenous tail-vein injection.Healthy SD rats (n = 3) were sacrificed after tracer injection [ 18 F]GSK1482160 (at 5, 30, and 60 min) tail intravenous, and their blood and liver were collected for tissue homogenization for the radiometabolism HPLC and fractions were counted with a γ counter following a previously described procedure [15,16] (Fig. 1e,f).

P2X7R cell binding assay
The in vitro fluorescence assay based on ethidium bromide measurement was used to determine the P2X7R inhibitory activities of GSK1482160 (including R and S-enantiomer, confirmed with optical rotation tests) was synthesized following a previous protocol [17].Briefly, approximately 2 × 10 5 human HEK293-hP2X7R cells per well were placed in a 96-well poly-lysine-coated black microplate (Corning Inc., Kennebunk, ME) overnight.The HEK293-P2X7R cells were washed three times with cold phosphate-buffered saline (PBS) before performing the assay.Each well was replaced with 90 μL of assay solution containing 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (10 mM), N-methyl-D-glucamine (5 mM), KCl (5.6 mM), D-glucose (10 mM), and CaCl 2 (0.5 mM) (pH 7.4) and supplemented with 280 mM sucrose.Next, 10 µL of 60 μM BzATP (final 6 µM) and 10 µL of serially diluted compounds (final concentration 0.3-3 µM) were added to each well.After incubation for 15 min, ethidium bromide uptake was detected by measuring the fluorescence with a microplate reader (BioTek Instruments Inc., Winooski, VT), using an excitation wavelength of 540 nm and emission wavelength of 615 nm.The results were expressed relative to the maximum accumulation of ethidium bromide when stimulated with BzATP only, and the IC 50 values were determined using nonlinear regression by a 3-parameter inhibitor-effect equation in Prism (GraphPad Software, Inc., San Diego, CA).

Animal model of AS
All animal studies were performed in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals.Thirty-two 32-week-old female BALB/c mice were randomly divided into two groups: a control group (n = 16), and AS modeling group (n = 16).The AS mouse model was constructed as described in previous studies [18,19].Briefly, 100 µg proteoglycan lyophilized powder (Sigma, St. Louis, MO, USA) and 20 mg DDA adjuvant (Sigma) were mixed with 200 µL of normal saline to obtain an emulsion.Next, a group of AS mice, each received 200 µL intraperitoneal injections of this emulsion at 0, 21, and 42 days.The group of control mice each received 200 µL intraperitoneal injections of saline.

Micro-PET/CT procedures for AS
PET/CT imaging studies were performed on the control (n = 8) and the AS group (n = 8) after 8 and 14 weeks of modeling separately.The interval between [ 18 F]FDG and [ 18 F]GSK1482160 imaging was one day, and the mice fasted for more than 6 h before [ 18 F]FDG imaging.The mice were anesthetized with 1.5-2% isoflurane inhalation and intravenously injected with 200 μL 3.2 ± 0.32 MBq/mouse [ 18 F]FDG or [ 18 F]GSK1482160.A dynamic 0-30 min micro-PET/CT (Mediso nanoScan PET/CT imaging system, Mediso Inc, Hungary) scan protocol was performed immediately after [ 18 F]FDG or [ 18 F]GSK1482160 injection.Dynamic PET images consisting of 27 frames (10 × 3 s, 3 × 10 s, 4 × 60 s, and 10 × 150 s) were reconstructed with CT attenuation correction and isotope decay correction.

PET data analysis
Original [ 18 F]GSK1482160 and [ 18 F]FDG PET/CT images were analyzed using the Carimas 2.10 software (Turku PET center, Finland) to obtain the time activity curve (TAC) for regions of interest (ROIs) based on both PET and CT images.The ankle joints, spine, and muscle regions were hand-drawn using the combined PET/CT image at 30 min.PET data in target regions were expressed as standardized uptake values (SUVs).The volume of interest (VOI) consisting of at least 3 ROIs on the target area was obtained.The VOI was plotted on the target lesion and scanned for 30 min for static PET/CT to obtain SUV max values, and over 30 min for dynamic PET/CT to obtain the TAC.Logan graphical arterial input function analysis (LoganAIF) [20,21] was used to estimate the binding potential (BP) of [ 18 F] GSK1482160 across the ankle and spine.TACs of the left ventricle and muscle were obtained from the original micro-PET data and used for arterial or reference input function.Three-dimensional ROIs and corresponding TACs of the ankle joints and spine were obtained and used for output function.Patlak modeling [22] was used to estimate the influx rate of [ 18 F]FDG in the ankle and spine.The TAC of the left ventricle was obtained from the original micro-PET data and used for arterial input function.The TAC of the left ventricle, ankle, and spine were used to fit standard Patlak modeling to assess [ 18 F]FDG tissue kinetics via the least square regression method using Matlab (MathWorks.Inc) program as we previously reported [23].The goodness-offit was evaluated with R-squared.SUV max was calculated as the activity concentration within the VOI (Bq/mL) divided by injected activity (Bq) and multiplied by body weight (g).
joints were carefully dissected to preserve their integrity.Blocks of spine and ankle joint tissues were sectioned; samples were dehydrated in a graded ethanol series, embedded in paraffin, serially sectioned to generate 5 μm slices, and stained for hematoxylin-eosin (H&E), immunohistochemistry (IHC), and immunofluorescence (IF).Images of IHC were acquired and edited using a pathological section scanner (3DHISTECH, Budapest, Hungary).Double IF labeling was performed using a rabbit primary antibody for the murine P2X7R (1:100-fold dilution, ALOMONE LABS Inc., Israel) and rat anti-mCD68 (1:50-fold dilution, Abcam Inc., Waltham, MA).Alexa Fluor 488-conjugated goat anti-rabbit (1:400) and cyanine dye 3-conjugated goat anti-rat secondary antibodies (1:300) were used to visualize these primary antibodies.Images were acquired using a fluorescence confocal system equipped with a Nikon Eclipse E600 microscope and multiple laser diode modules.

Human specimen collection
Specimens from an AS patient who underwent osteotomy and orthopedic surgery and another patient with intervertebral disc herniation were obtained from the Fifth Affiliated Hospital of Sun Yat-Sen University (FAHSYSU).All subjects were recruited and samples were collected under Institutional Review Board (IRB)-approved protocols (ZDWY.JZWK.002) at FAHSYSU.We complied with all applicable requirements of the local and national regulations and obtained informed consent from each subject before the study began.These protocols were reviewed by the IRB of FAHSYSU before transferring and analyzing the samples.The posterior longitudinal ligament tissue was collected aseptically post-operation.After carefully removing the surrounding adipose tissue, bone, blood vessels, etc., the posterior longitudinal ligament tissue was placed in a clean and sterile 50 mL centrifuge tube on ice, washed with PBS 3 times, and frozen in a liquid nitrogen tank for subsequent experiments.Tissue histology, H&E immunohistochemistry, and immunofluorescence staining steps are the same as in the previous section.

Statistical analysis
Statistical analysis was performed using SPSS 22.0 software (IBM, USA) and GraphPad Prism 8.0 software (GraphPad Software, Inc., USA).Data are presented as mean ± standard deviation (SD) or mean ± standard error of the mean (SEM).Normally distributed data were analyzed by Student t-test, one-way ANOVA, or Mann-Whitney U parameter analysis and p < 0.05 was considered statistically significant.

In vitro and in vivo stability
At 37 °C incubation [ 18 F]GSK1482160 exhibited good in vitro stability in saline and FBS, and there was no significant falloff after approximately 2.5 h of labeling (Fig. 1b,c, and d).The in vivo stability of the probe was investigated after the tracer was tail injected into 3 healthy SD rats.According to the in vivo radiometabolic HPLC, there was only one major metabolite peak, along with the parental [ 18 F]GSK1482160 in all tested samples, which was consistent with previous preliminary results [15].At 60 min tail post-iv injection, [ 18 F]GSK1482160 retained 62.5 ± 5.56% in blood samples, and 63.85 ± 7.57% in liver tissues (Fig. 1e, f, and Table 1) indicating its good in vivo stability.

P2X7R cell binding assay
The IC 50 value of (S)-GSK1482160 (2.12 ± 0.06 nM) was consistent with previous reports and was lower than that of (R)-GSK1482160 (15.92 ± 0.47 nM).In consistent with previous reports, S-enantiomers showed better binding affinity than R-enantiomers (Fig. 1g and h).

AS mouse model and behavioral scoring
The AS mouse model was evaluated by joint swelling, arthritis score, and histopathology.Eight weeks after constructing the AS mouse model, obvious symptoms of lower extremity joint swelling began to appear, and with the extension of the modeling time, the degree of lower extremity joint swelling gradually increased and peaked at the 14 th week (Fig. 2b).Lower extremity joint inflammation was also scored with a similar protocol (Fig. 2c).The control group did not show the aforementioned changes.We used H&E staining to evaluate the histological features of the ankle, spine, and sacroiliac joints in each group.Pathological changes were recorded in the AS group, comprising synovial hyperplasia and inflammatory infiltration of the ankle joint (Fig. 2d), inflammatory infiltration, and damage to the cartilage endplate of the intervertebral disc (Fig. 2e).However, no obvious signs of inflammation were observed in the control group.

Micro-PET/CT imaging of ankle joints of the AS
Compared with the control groups, significant radioactive uptake was observed in the ankle joints in AS groups after [ 18 F]GSK1482160 injection; however, [ 18 F]FDG uptake was not significantly different between these four groups (Fig. 3a).The [ 18 F]GSK1482160 TACs of the ankle joints in the AS 8W and AS 14W groups were significantly higher than those in the control 8W and control 14W groups, respectively (Fig. 3b).There was no significant difference in the [ 18 F]FDG TACs among these four groups (Fig. 3c).At weeks 8 and 14, the SUVmax values of [ 18 F]GSK1482160 in ankle joints of AS groups were significantly higher than that in control groups at each time point (Fig. 3d).However, there was no significant difference in the SUVmax values of [ 18 F]FDG among these four groups (Fig. 3e).The trend of BP ND ankle values of [ 18 F] GSK1482160 and K i ankle of [ 18 F]FDG was roughly the same as that of the SUVmax values (Fig. 3f and g).Quantitative analysis revealed a high correlation between [ 18 F] GSK1482160 uptake in the ankle joints and arthritis scores (Fig. 3h); however, [ 18 F]FDG uptake was not associated with arthritis scores (Fig. 3i).

Micro-PET/CT imaging of spinal joints of the AS
Compared with the control groups, significant radioactive uptake was observed in the sacroiliac joints (SIJ) in AS groups after [ 18 F]GSK1482160 injection; however, [ 18 F] FDG uptake was not significantly different between these four groups (Fig. 4a).The [ 18 F]GSK1482160 time activity curves of the SIJ in the AS 8W and AS 14W groups were significantly higher than those in the control 8W and 14W groups, respectively (Fig. 4b).There was no significant difference in the [ 18 F]FDG time activity curves among these four groups (Fig. 4c).At weeks 8 and 14, the SUVmax values of [ 18 F]GSK1482160 in the SIJ of AS groups were significantly higher than that in control groups at each time point (Fig. 4d).However, there was no significant difference in the SUVmax values of [ 18 F]FDG among these four groups (Fig. 4e).By comparing the BP ND values of [ 18 F] GSK1482160 in various parts of the spine, we observed significant differences among these four groups in terms of caudate vertebra (CV), sacroiliac joint (SIJ), lumbar 6 (L6), and lumbar 5 (L5) after 14 weeks of modeling.However, after 8 weeks of modeling, there were only differences in CV, SIJ, and L6 (Fig. 4f, h).The K i of [ 18 F]FDG was not significantly different among the four groups (Fig. 4g, i).

Tracer kinetics
To further quantify the in vivo PET results, 0 -30 min kinetic data were characterized for the tracer kinetics of [ 18 F]GSK1482160 in the ankle and SIJ by LoganAIF modeling (Table 2).In the AS 14W g roup, BP ND ankle−AS−14W = 12.212 ± 2.25, BP ND SIJ−AS−14W = 13.389± 3.60 were significantly higher than that in the Ctrl 14W group (Fig. 5).In the AS 8W group, BP ND ankle−AS−8W = 3.931 ± 0.74, BP ND SIJ−AS−8W = 4.225 ± 0.84 were significantly higher than that in the Ctrl 8W group as well (Fig. 5).The 0 -30 min kinetic data for [ 18 F]FDG were characterized by Patlak modeling (Table 3).There was no significant difference in the influx rates (K i ) of [ 18 F]FDG among the four groups (p > 0.05).

Histologic analysis from AS mice
The colocalizations of P2X7R (green), CD68 (macrophages, red), and DAPI (blue) were indicated only in AS group.Histological analysis of mouse ankle joints in [ 18 F] GSK1482160-positive regions showed an enrichment of immune cells in the ankle and metacarpophalangeal joint regions, and P2X7R (green) was expressed on the surface of CD68-positive (red) macrophages in immune cell enrichment regions (Fig. 6).

Histologic analysis from human specimen
We collected supraspinous ligament specimens from the patient with AS and that disc herniation for histological examination.Inflammatory infiltration of the supraspinous ligaments (green arrows) was the pathological feature of the AS group; however, there was no corresponding change in the control group.Consistent with what was observed in animal models of AS, colocalizations of P2X7R (green), CD68 (macrophages, red), and DAPI (blue) were indicated only in the patient with AS, whereas there was no corresponding change in the control group (Fig. 7).

Discussion
AS is an inflammation-related disease with which pathological processes are difficult to quantify.Inflammation, bone erosion, and syndesmophyte formation may be ongoing for decades.We used a proteoglycan and DDA adjuvant to create an inflammation-associated AS model that mimics the pathological process of AS in patients.The condition began with peripheral joint symptoms, including ankle synovitis, synovial hyperplasia, cartilage erosion, and advanced spinal destruction.Several methods, including clinicalgrade analysis and histological techniques, were applied to detect the inflammatory status of AS, which involves the ankle, SIJ, and spine.H&E staining confirmed that the pathological changes in the ankle joint and spine of the mouse AS model were inflammatory infiltration.At the  F]GSK1482160 (f) and the influx rate constant K i ankle of [ 18 F] FDG (g) in four groups (n = 8).Quantitative analysis revealed a high correlation between [ 18 F]GSK1482160 uptake in ankle joints and arthritis scores (h); however, [ 18 F]FDG uptake was not associated with arthritis scores (i).*p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant same time, micro-PET/CT imaging showed high uptake of [ 18 F]GSK1482160 in the ankle and spine of the AS model, which indicated that the P2X7R is highly expressed in AS.Similarly, IF staining indicated that the highly expressed P2X7R is colocalized with macrophages.However, the influx rates of [ 18 F]FDG into the ankle and spine were not significantly different between the AS and the control groups.
A recent study showed that the P2X7R abundantly accumulates in activated macrophages and the aggregation of macrophages was shown significantly associated with disease progression in AS [25], which was consistent with our results.At present, a molecular probe targeting the P2X7R has been used in nerve and vascular PET/CT imaging studies [26]; however, there have been no reports regarding the detection of AS.We successfully synthesized the [ 18 F]GSK1482160, one of the P2X7R-specific radioligands.It is noteworthy that the total RCY and stability of [ 18 F] GSK1482160 were superior to those of [ 18 F]IUR-1601 and [ 18 F]IUR-1602 as reported by our recent study and others [15,27,28].It is synthesized with a novel precursor from ArCF 2 SO 2 CF 3 to directly obtain fluorine-18 radiosynthesis Therefore, the P2X7R-specific radioligand [ 18 F] GSK1482160 was successfully used for PET/CT imaging of AS in this study.As for the performance of [ 18 F] GSK1482160 in AS imaging studies, we observed that the high uptake of the probe by the ankle can be significantly observed within 15-30 min after injection.Compared with [ 18 F]FDG, [ 18 F]GSK1482160 has significantly higher binding in AS model.Similarly, we observed that the P2X7R is also highly expressed in the SIJ of the spine with a gradual decline upward along the SIJ, which is highly similar to the pathological process of AS.This proved once again that our AS model highly simulates AS in clinical settings.
We successfully used [ 18 F]GSK1482160 and PET/CT in vivo dynamic imaging to accurately locate and quantitatively analyze the systemic lesions of AS, thus providing a new idea for clinical classification and diagnosis of AS.In addition, we used the Logan graphic modeling method to analyze the tracer dynamics of [ 18 F]GSK1482160 and [ 18 F]FDG.The specificity of [ 18 F]GSK1482160 in AS detection was much higher than that of [ 18 F]FDG; this provided potent medical imaging and quantification for future clinical applications.Quantitative analysis revealed a high correlation between [ 18 F]GSK1482160 uptake in ankle joints and arthritis scores (Fig. 3h); however, [ 18 F]FDG uptake was not associated with arthritis scores (n = 8) (Fig. 3i).Clinical studies demonstrated that high uptake of [ 18 F]FDG has not been found in the lesion site of patients with AS [9,10].This may be due to the limitations of [ 18 F]FDG PET/CT in the diagnosis of AS, which is also the significance of our development of [ 18 F]GSK1482160.All these results suggest that [ 18 F]GSK1482160 can reflect the degree of inflammation of an AS model by monitoring the level of P2X7R expressed on macrophages in the ankle and spine.
Additional studies have reported the use of [ 18 F]NaF PET/CT for the diagnosis and treatment of AS [29].However, the tracer [ 18 F]NaF has strong nonspecific binding in bones which limits its application for AS diagnosis.Our study provided a novel potential molecular biomarker for AS diagnosis and assessment, with the potential to achieve precise localization.Moreover, we found that the spine and ankle lesions could be detected by PET/CT at the 8 th week of modeling, which meant that using [ 18 F] GSK1482160 PET/CT imaging could have the potential for early diagnosis of AS.However, there are still some limitations in our study, such as the small number of human clinical samples and the lack of human translation research.Although the clinical samples of patients with AS are very precious and difficult to collect, more specimens may need to be collected in the future to verify our conclusions.We believe that if this study is applied to clinical research, the precise positioning and quantitative evaluation technology of AS will likely greatly promote the clinical diagnosis and staging of AS.It will also provide a reliable theoretical basis for the formulation of a diagnosis and treatment plan for AS.In the future, our research group will continue to actively promote the clinical translation of [ 18 F]GSK1482160 and strive to further verify our conclusions on patients with AS.If this assumption is established, it will hopefully help solve the diagnostic problem of AS and may provide new methods for the treatment of AS.

Conclusion
We successfully developed a novel [ 18 F]GSK1482160 tracer and PET/CT imaging to accurately locate and quantitatively analyze the systemic lesions in mouse AS models, thereby providing a novel clinical diagnostic method for AS.The Logan graphic modeling indicated that the specificity of [ 18 F]GSK1482160 in detecting AS was much higher than that of [ 18 F]FDG and that [ 18 F]GSK1482160 can reflect the degree of inflammation in AS by monitoring the level of P2X7R expressed on macrophages in the ankle and spine.This provides potent medical imaging and quantification for clinical application in AS.Our research is of great clinical importance to further elucidate the pathogenesis of AS and potential molecular biomarkers.

Fig. 1
Fig. 1 Probe preparation and biochemical characteristics (a) Radiosynthesis scheme for [ 18 F] GSK1482160.(b) Radioactive HPLC profile of [ 18 F] GSK1482160.(c and d) In vitro stability of [ 18 F]GSK1482160 after incubation in saline and fetal bovine serum from 0.5 h to 2.5 h.(e and f) radio metabolite analysis of rat's blood samples and liver tissues post injection of [ 18 F]GSK1482160.(g and h) Saturation binding curve of (S)-GSK1482160 and (R)-GSK1482160

Fig. 2
Fig. 2 Experimental procedure and pathological process of AS.(a) Flow chart of the study.(b) Clinical manifestations of representative hind paws in each group.(c) Arthritis scores for control and AS groups during 0-14 weeks.(d) Histological changes in the ankle joint (H&E staining).The pathological features of the AS group were inflammatory infiltration of the ankle joint (green arrows) and syno-

Fig. 3 [
Fig. 3 [ 18 F]GSK1482160 and [ 18 F]FDG Micro-PET/CT dynamic imaging in the ankle joint.(a) Representative sagittal, coronal, and transverse sectional micro-PET/CT images showing the uptake of [ 18 F]GSK1482160 and [ 18 F]FDG in ankle joints in 4 different groups (Ctrl 8W, AS 8W, Ctrl 14W, and AS 14W, n = 8 for each).TAC of the ankle joints in four groups [ 18 F]GSK1482160 (b) and [ 18 F] FDG (c).The SUV max values of the ankle joints in four groups [ 18 F] GSK1482160 (d) and [ 18 F]FDG (e) at different time points during the

Fig. 4 [
Fig. 4 [ 18 F]GSK1482160 and [ 18 F]FDG Micro-PET/CT dynamic imaging in the spine.(a) Representative sagittal, coronal, and transverse sectional micro-PET/CT images showing the uptake of [ 18 F] GSK1482160 and [ 18 F]FDG in the sacroiliac joints of spines in 4 different groups (Ctrl 8W, AS 8W, Ctrl 14W and AS 14W, n = 8 for each).Time activity curve of [ 18 F]GSK1482160 (b) and [ 18 F]FDG (c) in the sacroiliac joints for four groups (n = 8).The SUV max values of [ 18 F]GSK1482160 (d) and [ 18 F]FDG (e) at different time points in the

Fig. 6
Fig.6 Immunofluorescence staining for P2X7R and CD68 in the ankle joints in the control and ankylosing spondylitis groups.The colocalizations of P2X7R (green), CD68 (macrophages, red), and DAPI (blue) were indicated only in the AS group