Human liver‐expressed antimicrobial peptide 2 elevation in the cerebrospinal fluid in bacterial meningitis

Abstract Objective To study the presence of liver‐expressed antimicrobial peptide 2 (LEAP2) in human cerebrospinal fluid (CSF) and to measure its concentrations in neurological disorders. Materials & Methods We identified the presence of LEAP2 in human CSF by chromatographic analysis and a LEAP2‐specific enzyme immunoassay. We measured LEAP2 concentrations in the CSF of 35 patients with neurological disorders. Results CSF LEAP2 concentrations in the bacterial meningitis group (mean ± SD, 9.32 ± 3.76 ng/ml) were significantly higher (p < .05) than those in the other four groups (psychosomatic disorder, 0.56 ± 0.15 ng/ml; peripheral autoimmune disease, 1.00 ± 0.60 ng/ml; multiple sclerosis, 0.62 ± 0.30 ng/ml; aseptic meningitis, 1.59 ± 0.69 ng/ml). Conclusions This is the first study to identify the presence of human LEAP2 in the CSF. Levels of LEAP2 were increased in the CSF of patients with bacterial meningitis. LEAP2 may have potential as a biomarker for bacterial meningitis.

system by acting on immune cells and endothelial cells to suppress the production of pro-inflammatory cytokines, suggesting its therapeutic viability in inflammatory conditions (Colldén et al., 2017;Yanagi et al., 2018). Leap2 was found to be highly expressed in the liver and intestine in rodents, and its expression was decreased in the liver of rats under fasting condition (Islam et al., 2020). LEAP2 was also demonstrated to act as an inverse agonist of GHSR1a by blocking its constitutive activity (M'Kadmi et al., 2019). LEAP2 suppressed food intake and growth hormone release and maintained viable glucose levels during chronic caloric restriction in mice (Ge et al., 2018). The six N-terminal amino acids of LEAP2 are crucial for GHSR1a binding, and its 10-amino-acid N-terminal fragment stimulated insulin release from human pancreatic islets (Hagemann et al., 2020;Wang et al., 2019). Plasma LEAP2 concentrations in humans increased with body weight gain and elevations of blood glucose and decreased with fasting and weight loss surgery (Mani et al., 2019).
Although GHSR1a is widely expressed in the central nervous system (CNS) (Al-Massadi et al., 2018), it is still unclear whether LEAP2 is involved in human neurological diseases. We demonstrated that the intracerebroventricular administration of LEAP2 to rats antagonized central ghrelin functions, including the promotion of food intake, elevation of blood glucose levels, and reduction of body temperature (Islam et al., 2020). Previous studies have not shown whether LEAP2 is present in human cerebrospinal fluid (CSF), and if so, whether its levels are altered in neurological disorders.

| Chromatographic characterization of CSF LEAP2
The N-terminus of LEAP2 is a methionine residue whose thiol molecule is susceptible to oxidization in the normal atmosphere.
To explore the presence of oxidized LEAP2 in human CSF, we oxidized synthetic LEAP2 (Peptide Institute) by the hydrogen peroxide method (Williams et al., 2010). Synthetic human LEAP2 (1 nmol) was reacted in a solution containing 1 N formic acid and 0.06% hydrogen peroxide for 30 min at 30°C. This peptide was analyzed by reverse-phase high-performance liquid chromatography (RP-HPLC) (Chromaster; Hitachi) using a TSKgel ODS-120T column (Tosoh Corporation) with a linear gradient of 10%-60% acetonitrile containing 0.1% trifluoroacetic acid (TFA) for 40 min at a flow rate of 1 ml/min. The absorbance of the eluent was monitored at 210 nm.
To identify LEAP2 in human CSF, the CSF obtained by lumbar puncture in a patient with idiopathic normal pressure hydrocephalus was first treated in a 10-kDa column (AmiconUltra 0.5 ml Centrifugation Filters; Merck Millipore) to collect substances whose molecular masses were less than 10 kDa. Next, the substances were analyzed by RP-HPLC under the conditions described above. HPLC fractions separated every 30 s were analyzed with an enzyme-linked immunosorbent assay (ELISA; LEAP2 (38-77) (Human)/LEAP2 (37-76) (Mouse))-enzyme immunoassay (EIA) Kit (Phoenix Pharmaceuticals).
This ELISA detected oxidized and nonoxidized LEAP2 with equal sensitivity on a molar basis.

| Measurements of LEAP2 in CSF and serum
Frozen 100 µl serum samples kept at − 80°C were allowed to melt at room temperature. Samples were loaded onto a Sep-Pak Vac C18 cartridge (Waters Corporation) prepared as described else- where (Mitsukawa et al., 1990). Substances were eluted with 60%

| Data analysis
Data were analyzed by one-way ANOVA with the post hoc Games-Howell test for intergroup analysis. The correlations between CSF LEAP2 and body mass index (BMI), serum glucose, C-reactive protein (CRP), CSF protein, and CSF cell count were analyzed by Spearman's correlation coefficient test. Data are expressed as means ± SD, and significance was assumed at p < .05. IBM SPSS version 25 was used for statistical tests.

| Identification of LEAP2 in human CSF
To identify the presence of LEAP2 in human CSF, we analyzed LEAP2-immunoreactive substances in the CSF by RP-HPLC. ELISA analysis of human CSF samples showed two immunoreactive peaks ( Figure 1a). Their elution positions were identical to those of oxidized and nonoxidized LEAP2. The molar ratio of oxidized LEAP2 to nonoxidized LEAP2 was 1:4.9. CSF LEAP2 concentrations were not affected by the CSF storage period, which ranged from 32 days to 2,058 days (Figure 1b).

| CSF LEAP2 concentrations in bacterial meningitis
CSF LEAP2 concentrations in the bacterial meningitis group were significantly higher than those in the other four clinical groups

| Serum LEAP2 in bacterial infection
The serum LEAP2 concentrations in the 12 healthy subjects ranged from 6.89 ng/ml to 28.84 ng/ml (mean ± SD, 18.35 ± 7.15 ng/ ml) ( Table 2). The serum LEAP2 concentration in the patient with Listeria meningitis (13.47 ng/ml) was also within this range, while that in the patient with bacteremia (31.40 ng/ml), who had no CNS infection, was higher than the range observed in the healthy subjects. The CSF LEAP2 concentration was lower in the bacteremia patient (2.53 ng/ml) than in the five patients with bacterial meningitis (ranging from 4.51 ng/ml to 13.40 ng/ml; mean ± SD, 9.32 ± 3.76 ng/ml).

| D ISCUSS I ON
In this study, we first verified the presence of LEAP2 in human CSF by RP-HPLC combined with ELISA. LEAP2 has an N-terminal methionine whose thiomethyl residue is readily oxidized under conventional conditions (Davies, 2016). We hypothesized that a portion of LEAP2 might be oxidized in the CSF. We demonstrated that both methionine-oxidized and nonoxidized LEAP2 were present in the CSF.
LEAP2 penetrates the negatively charged bacterial membrane because of its characteristic strong basicity and two disulfide bonds, causing holes or pores that eventually lead to bacterial cell death.
This behavior affects the membranes of bacteria more preferentially than those of eukaryotes (Townes et al., 2009). Serum LEAP2 in mice was shown to antagonize ghrelin action on its cognate receptor and also to suppress ghrelin production (Ge et al., 2018). By contrast, ghrelin suppressed LEAP2 expression in the liver via the ghrelin receptor (Islam et al., 2020). Ghrelin exerts a broad spectrum of physiological actions, including modulating immunity and inflammation and reducing oxidative stress (Colldén et al., 2017;Yanagi et al., 2018).
Considering LEAP2's antagonism of ghrelin, LEAP2 may also be involved in modulating the immune system. Although plasma LEAP2 concentrations in patients with rheumatoid arthritis were reported to be higher than those in healthy subjects (Francisco et al., 2020), the role of LEAP2 in inflammation has not yet been clarified.
In this study, we showed that patients with bacterial meningitis had significantly higher CSF LEAP2 concentrations than those with other neurological disorders and a control group. CSF LEAP2 con-

TA B L E 2 LEAP2 concentrations in CSF
and serum cerebellum, midbrain, and medulla oblongata, although at low levels (Islam et al., 2020). Although it is unclear whether LEAP2 enters the brain by the same mechanism as ghrelin, LEAP2 levels rise in the human CNS during bacterial infection.
Inflammation produces a variety of anorectic substances that reduce food intake (Roxburgh and McMillan, 2014). Intraventricular administration of LEAP2 suppressed ghrelin-induced feeding in rats (Islam et al., 2020). The elevated level of CSF LEAP2 in bacterial meningitis suggests that the molecule may play a role in some brain areas that regulate food intake and may mediate appetite reduction under inflammatory states. In this study, the CSF-to-serum ratio of LEAP2 in the patient with Listeria meningitis was markedly higher than that in the patient with bacteremia (0.532 versus. 0.081, respectively), suggesting the existence of a mechanism by which the LEAP2 level in CSF rises independently of that in serum.

Clinical guidelines recommend that when bacterial meningitis
is suspected, empirical therapy should be commenced before CSF tests and CNS imaging are performed (Young and Thomas, 2018).
Once antibiotics are administered, CSF cell counts or cell types may mimic those in aseptic meningitis, and cell culture results may become negative (Converse et al., 1973). Given that increased CSF LEAP2 levels are specific for bacterial meningitis, CSF LEAP2 measurement may be useful as an auxiliary diagnostic measure.
Limitations of this study include the lack of serum samples from patients with pneumococcal meningitis and the small number of CSF samples obtained during the course of routine clinical examination.
Considering LEAP2's functions as an antimicrobial peptide and an antagonist of the ghrelin receptor, elevated CSF LEAP2 concentrations in bacterial meningitis suggest that LEAP2 may act against local bacterial infection or acute inflammation. Thus, increased LEAP2 levels during bacterial infection may be applicable as a potential biomarker. Furthermore, the presence of LEAP2 in the human CSF may suggest the existence of a peptidergic innate immune system in the CNS. This study should contribute to better understanding the host defense in bacterial meningitis.