What can we learn about functional importance of human antimicrobial peptide LL-37 in the oral environment from severe congenital neutropenia (Kostmann disease)?

The human antimicrobial peptide LL-37 is produced by neutrophils and epithelial cells, and the peptide can be detected in plasma as well as saliva. LL-37 is active against both both gram-positive and gram-negative bacteria including oral pathogens such as Porphyromonas gingivalis and Streptococcus mutans. Besides its antimicrobial properties, LL-37 modulates the innate immune system, and furthermore, it also affects host cell viability. Although, both structural and functional properties of LL-37 have been extensively investigated, its physiological/pathophysiological importance in-vivo is not completely understood. In this review, Kostmann disease (morbus Kostmann) is highlighted since it may represent a LL-37 knockdown model which can provide new important information and insights about the functional role of LL-37 in the human in-vivo setting. Patients with Kostmann disease suffer from neutropenia, and although they are treated with recombinant granulocyte colony-stimulating factor (G-CSF) to normalize their levels of neutrophils, they lack or have very low levels of LL-37 in plasma, saliva and neutrophils. Interestingly, these patients suffer from severe periodontal disease, linking LL-37-deficiency to oral infections. Thus, LL-37 seems to play an important pathophysiological role in the oral environment antagonizing oral pathogens and thereby prevents oral infections.


Introduction
Human LL-37 belongs to the cathelicidin family of antimicrobial/ host defense peptides. LL-37 exerts antimicrobial and host cell cytotoxic effects, but it also influences innate immunity in a complex way [1][2][3][4][5][6][7]. Interestingly, LL-37 may have both pro-and anti-inflammatory effects through its interactions with the NF-κB pathway. Upon internalization of LL-37 and/or activation of putative plasma membrane receptors for LL-37, such as the P2X7 receptor, the peptide is thought to interact with numerous genes and proteins of human host cells probably explaining its many cellular actions and diverse effects [7,8]. Importantly, the LL-37 molecule is protonated giving it positive net charges at physiological conditions, and moreover, the peptide is amphiphilic. These structural properties of LL-37 provide the background and explain many of the interactions which can occur between LL-37 and different cellular and subcellular targets. The biological activity of LL-37 is probably very much dependent on these structural characteristics of the peptide.
LL-37 is synthesized by different types of white blood cells of which neutrophils are regarded as very important producers of the peptide, but also epithelial cells can synthesize LL-37 [2,7]. The peptide is encoded by the CAMP gene and synthesized as the pro-form hCAP18. The hCAP18 protein is released and cleaved extracellularly to biologically active LL-37 by serine protease 3 and kallikrein 5 [9,10].
In order to assess and judge the physiological/pathophysiological importance of LL-37, data obtained in animal cells and animal in-vivo models are of limited value since LL-37 is expressed only in humans, although for example mice express the LL-37 orthologue CRAMP [11]. Hence, studies in primary cells and/or cell lines which aim to assess and clarify the functional importance of LL-37 are preferably performed in human cells, and for assessment of the physiological/pathophysiological importance of LL-37 in-vivo, the human in-vivo setting is warranted. In this review, it is proposed that patients suffering from Kostmann disease (morbus Kostmann), characterized by severe neutropenia, may represent an important source of information regarding the functional significance of LL-37 in-vivo, since these individuals can represent a human LL-37 knockdown model [12]. The Kostmann disease patients, successfully treated with granulocyte colony-stimulating factor (G-CSF) to normalize their neutrophil counts, still suffer from oral disease and interestingly their plasma, saliva and neutrophils show no or very low levels of LL-37, indicating that the peptide combats oral pathogens and prevents oral infections.
2. LL-37 is active against oral bacteria LL-37 is active against both gram-positive and gram-negative bacteria [1,4,5,13]. As mentioned before the peptide is positively charged, and therefore it shows high affinity for negatively charged groups of plasma membranes. Interestingly, the bacterial cell wall is particularly rich in negatively charged groups, which probably explains why LL-37 preferably interacts with this structure, although it may also bind to host cell plasma membranes. Upon binding of LL-37 to the wall of bacteria, it forms pores and thereby permeabilizes the bacterial cell wall leading to bacterial cell lysis. LL-37 is thought to induce pore formation via either the barrel-stave or the carpet model [4]. For the barrel-stave model, LL-37 molecules adhere to the bacterial membrane and organize in oligomers which form a cylindrical pore. For the carpet model, peptides coat the membrane, and when the concentration of LL-37 reaches a critical value the membrane is either permeabilized in a detergent-like manner or curved causing the formation of toroidal pores. LL-37 has also been reported to bind the bacterial endotoxins lipopolysaccharide (LPS) produced by gram-negative bacteria and lipoteichoic acid (LTA) produced by gram-positive bacteria and thereby neutralize many types of bacteria [14,15]. Hence, LL-37 may exert its antibacterial activity both through permeabilization of the bacterial cell wall and via binding and neutralization of LPS and LTA.
The peptide shows high activity against many types of oral bacteria including the important oral pathogens Porphyromonas gingivalis and Streptococcus mutans which are involved in the pathogenesis of periodontitis and caries, respectively [16][17][18]. Interestingly, strains of Streptococcus mutans isolated from caries-active patients seem to be less sensitive to LL-37-induced anti-bacterial activity compared to Streptococcus mutans strains obtained from caries-free individuals, suggesting that Streptococcus mutans of caries-active patients can develop resistance to LL-37 [17]. Gingipain proteases secreted by the periodontitis pathogen Porphyromonas gingivalis represent a virulence factor of the bacteria that may degrade and inactivate LL-37 and thereby protect the microorganisms against LL-37-induced cytotoxicity [18,19]. Interestingly, inhibitors of gingipains can prevent degradation of LL-37 induced by gingival crevicular fluid collected at Porphyromonas gingivalis-positive sites [20]. These results suggest that inhibitors of gingipains protect LL-37 from degradation, and thus they may have beneficial effects in patients suffering from periodontitis via this mechanism [20].

LL-37 has both pro-and anti-inflammatory properties
LL-37 exerts multiple effects in innate immunity involving both proand anti-inflammatory actions, and these effects seem to depend on cell type [7]. The peptide has been reported to act as a chemoattractant for monocytes, neutrophils, T cells and eosinophils through activation of formyl-peptide receptors [21,22]. LL-37 has also been shown to stimulate the differentiation of macrophages into a pro-inflammatory macrophage phenotype, representing another LL-37-evoked pro-inflammatory mechanism of action [23]. Additionally, LL-37 seems to activate many other pro-inflammatory pathways, but importantly the peptide can also induce anti-inflammatory signaling as reviewed by Hancock et al. [7]. In fact, both LL-37-induced pro-and anti-inflammatory mechanisms are suggested to involve LL-37-evoked modulation of NF-κB signaling [7].

LL-37-induced host cell cytotoxicity
It is well-documented that high concentrations (> 1 μM) of LL-37 can reduce cell viability in many different human cell types. In Table 1, examples of studies showing LL-37-induced down-regulation of cell viability and cell number are presented. Notably, there are also other reports, in addition to those presented in Table 1, describing LL-37evoked cytotoxicity in human cells. Importantly, reduction of cell viability caused by LL-37 is not restricted to only one or a few cell types but observed in many different human cell types (Table 1). Although, many reports convincingly show that LL-37 attenuates cell viability, there are some studies in skin cells showing that LL-37 may prevent apoptosis. In this context, LL-37 has been demonstrated to antagonize apoptosis induced by the topoisomerase I inhibitor camptothecin in human keratinocytes and to reduce sodium-nitroprusside-induced apoptosis in dermal fibroblasts isolated from patients with systemic sclerosis [29,30]. Thus, although there are many studies demonstrating that high concentrations of LL-37 are cytotoxic for human cells, there are also reports showing that LL-37 may suppress apoptosis induced by pro-apoptotic agents.
The mechanisms behind LL-37-induced cytotoxicity and cell death have not been fully clarified. Treatment with synthetic LL-37 has been demonstrated to cause cellular accumulation of trypan blue and morphological changes in human vascular smooth muscle cells, osteoblastlike MG63 cells and periodontal ligament cells, such as membrane blebbing and cell shrinkage, representative for apoptosis [25,27,28]. Moreover, there are reports showing that LL-37 causes a translocation of phosphatidylserine from the inner to the outer side of plasma membrane, a process regarded as a sign of early apoptosis [3,28]. The LL-37-induced flip of phosphatidylserine is assessed by the phospholipid-binding protein Annexin V labelled with a fluorescent compound for detection [31]. LL-37 has also been shown to induce DNA fragmentation visualized by increased TUNEL staining [3,26]. There are reports convincingly showing that high cytotoxic concentrations of LL-37 do not induce activation of either caspases or PARP-cleavage, although they cause DNA fragmentation assessed by TUNEL assay, demonstrating that LL-37-induced cytotoxicity does not involve all the characteristic signs of apoptosis [3,26]. Interestingly, co-stimulation with LL-37 and Pseudomonas aeruginosa, but not treatment with either the peptide or bacteria alone, activates caspases in a human bronchial epithelial cell line, suggesting that LL-37 and Pseudomonas aeruginosa can act in synergy to promote activation of caspases [26].

LL-37 in human saliva
LL-37 has been detected in unstimulated whole saliva collected from both children and adults [32][33][34][35]. Interestingly, hCAP18 and LL-37 can be demonstrated in both isolated human parotid and submandibular/ sublingual saliva, suggesting that salivary LL-37 indeed is produced by cells of the major salivary glands [36]. As depicted in Fig. 1, strong immunoreactivity for LL-37 is detected in neutrophils of blood vessels in human parotid and submandibular glands, suggesting that glandular neutrophils represent an important source of salivary LL-37 [36]. However, LL-37 immunoreactivity has also been observed in ductal epithelial cells of human parotid glands, indicating that also these cells contribute to salivary LL-37 [37]. Furthermore, acinary cells of mouse submandibular glands show an immunoreactive signal for the mouse LL-37 ortholog CRAMP [32]. Taken together, these findings suggest that many cell types of salivary glands may contribute to salivary LL-37. Importantly, Yang et al. [38] have demonstrated that human whole saliva contains proteinase 3 (serine protease 3), an enzyme responsible for extracellular processing of hCAP to LL-37, indicating that salivary hCAP18 indeed can be cleaved to biologically active LL-37. Hence, salivary gland cells may release the pro-form of LL-37, hCAP18, which is then processed to LL-37 by salivary proteinase 3 in the salivary ducts and/or more distally when saliva reaches the oral cavity.
Neutrophil extracellular traps, which show high capacity to both bind and kill bacteria, have been demonstrated in human saliva [39]. The salivary neutrophil extracellular traps possess bound hCAP18/LL-37, indicating that LL-37 is responsible for at least some of their antimicrobial activity [39]. It seems reasonable to suggest that salivary LL-37 can neutralize oral bacteria both through a direct effect independent of neutrophil extracellular traps, and via an indirect effect involving initial trapping of bacteria by neutrophil extracellular traps followed by lysis of the bacteria induced by LL-37 bound to these traps. Importantly, patients with Kostmann disease lack salivary LL-37, and these patients suffer from oral infectious/inflammatory diseases, suggesting that salivary LL-37 indeed is important for protection against oral pathogens and prevention of oral disease [12,40]. Interestingly, the Kostmann disease patients have no or very few neutrophils, and if these patients are treated with G-CSF to normalize their neutrophil levels, their new neutrophils show no LL-37 expression, indicating that LL-37 produced by functionally intact neutrophils represents an important source of salivary LL-37 [12]. Notably, there are, to the best of my knowledge, no studies linking LL-37-deficiency in the saliva to lack of salivary gland expression of hCAP18/LL-37 in Kostmann disease patients. Lack of LL-37 in saliva of patients with Kostmann disease is probably not due to lack of proteolytic cleavage of hCAP18 to LL-37, since neither hCAP18 nor LL-37 is detected in the saliva of Kostmann disease patients [12].

Kostmann disease
Kostmann disease, also named severe congenital neutropenia, is an autosomal recessive hereditary disease characterized by severe neutropenia and named after the Swedish pediatrician Dr. Rolf Kostmann. He came across the original Kostmann families during his early career as a medical doctor in the northern parts of Sweden. Dr. Kostmann described the symptoms and the neutropenia of the Kostmann patients,  defined their family bonds, and showed the heredity of the disease in his PhD thesis presented 1956 [41]. He noticed that many children within the Kostmann families suffered from infectious diseases and died from banal infections such as otitis media at very young age. Mutations in many different genes have been implicated in the etiology of Kostmann syndrome. For example, homozygous mutations in the HAX1 gene is considered to be associated with the disease [42]. HAX1 encodes for the protein HAX-1, also named HCLS1-associated X1, a protein coupled to mitochondrial function and acting as an inhibitor of apoptosis in myeloid cells [42]. Notably, Kostmann disease (severe congenital neutropenia) is a rare disease with a reported prevalence of 1-2 cases per one million individuals [43,44].

Kostmann disease patients are LL-37-deficient and suffer from oral infections
The study by Pütsep et al. [12] represents an important eye-opener high-lightning the association between Kostmann disease and LL-37 deficiency. Here, the authors show that lack of LL-37 in body fluids and neutrophils of patients with Kostmann disease is associated with chronic periodontal disease (Fig. 2). Although, the Kostmann disease patients were treated with G-CSF and thereby showed normal neutrophil counts, they had extremely low levels of both LL-37 and its proform hCAP18 in plasma, saliva and neutrophils. Interestingly, one of the Kostmann disease patients included in the study by Pütsep et al. [12] received a bone-marrow transplant. This patient had almost normal levels of LL-37 in plasma, saliva and neutrophils and furthermore no periodontal disease, indicating that functionally intact neutrophils are critically important for normal LL-37 levels in plasma and saliva and for the prevention/protection against periodontal disease. In a case study, Hakki et al. [45] reported that two siblings, a girl 3 years of age and a boy 6 years of age, diagnosed with Kostmann disease, suffered from recurrent severe oral infections and periodontal disease with loss of attachment and mobility of their teeth, confirming that Kostmann disease patients suffer from periodontal disease as reported previously by Pütsep et al. [12]. The boy lost all his deciduous teeth within 6 months, but then the dentists managed to stabilize the oral health situation of these two children through intensive therapy, and thereby prevented new oral infections [45]. Although, Hakki et al. [45] did not determine the LL-37 levels in plasma, saliva and neutrophils of these two siblings, it can probably be assumed that both of them were LL-37-deficient. Zetterström [46] suggests that the oral manifestations observed in patients with Kostmann disease are coupled to lack of salivary LL-37, but it is difficult to completely rule out the possibility that also their systemic deficiency of LL-37 in plasma and neutrophils influences the susceptibility for oral disease.
Recently, it has been demonstrated that Kostmann disease patients have higher periodontal pocket depths, higher bleeding-on-probing scores and more caries than healthy control subjects [40]. These authors showed that signs of oral infection/inflammation in Kostmann disease patients were associated with less diversity of the oral bacterial flora. Their analysis of the oral microbiome demonstrated that the genus Porphyromonas was observed in the core microbiome of patients with Kostmann disease but not in that of healthy controls [40]. Moreover, Firmicutes was more abundant in saliva of Kostmann disease patients than that of healthy controls [40]. Hence, it is possible that lack of LL-37, as observed in Kostmann disease patients, changes the oral microbiome in a less favorable way promoting oral infections.
Salivary flow rate and the buffer capacity of saliva are important factors protecting against oral disease. Lack of salivary LL-37 probably has direct detrimental effects on the protection against oral pathogens, but it may also have indirect effects affecting the composition of saliva and/or properties of salivary proteins which in turn can have a negative impact on salivary flow rate and buffer capacity.
The connection between oral infections, such as periodontitis, and systemic inflammatory diseases has been intensively studied during recent time, and many studies show positive correlation between periodontitis and cardiovascular disease. In a large study on patients with or without myocardial infarction, periodontitis was more common in patients with myocardial infarction than in the control subjects, and there was an increased risk of developing myocardial infarction in patients suffering from periodontitis [47]. Thus, it is possible that periodontitis associated with lack of LL-37, as observed in Kostmann disease, may secondarily be associated with systemic cardiovascular diseases such as myocardial infarction.

Conclusions
Besides its antimicrobial properties, LL-37 has a wide-range of effects including modulation of innate immunity and down-regulation of host cell viability, and thus the peptide shows a rather complex functional picture making it difficult to draw conclusions about its physiological/pathophysiological importance in-vivo. The Kostmann disease patients have no or very low levels of LL-37 in plasma, saliva and neutrophils, and they suffer from oral infections/inflammatory diseases, although they are successfully treated with G-CSF to normalize their neutrophil counts. The phenotype of these patients, representing a human LL-37 knockdown model, suggests that LL-37 possesses activity against oral pathogens, and thus the peptide seems to play an important physiological/pathophysiological role in the oral environment. However, it has to be taken into consideration that Kostmann disease is a rare disease, and thus the numbers of documented cases showing negative correlation between the levels of LL-37 and oral disease are relatively few.

Conflict of interest
The author declares no conflict of interest.