Background
The neuropeptide substance P (SP) and its selective receptor, the neurokinin-1 receptor (NK-1R), is expressed at high levels within the central nervous system (CNS) (as reviewed in [
1,
2]). In addition to its functions as a neurotransmitter in the perception of pain and its essential role in gut motility, this tachykinin is now recognized to exacerbate inflammation at peripheral sites including the skin, lung, and gastrointestinal and urogenital tracts. Indeed, this neuropeptide appears to contribute to disease pathology for some infectious agents. For example, SP increases the bronchoconstriction and damaging cardiac inflammation following infection with respiratory syncytial virus and encephalomyocarditis virus, respectively [
3,
4]. Likewise, SP contributes to the severity of inflammation associated with
Trypanosoma brucei brucei infection and inflammation and granuloma size in a mouse model of
Taenia solium cysticercosis [
5‐
7].
Recently, a number of studies have identified a similar role for SP and NK-1R interactions in neuroinflammation (as discussed in [
1,
2]), and our data suggests that SP exacerbates damaging inflammation within the CNS in animal models in response to disparate bacterial pathogens. We determined that the absence of SP/NK-1R interactions in SP receptor-deficient mice or prophylactic pharmacological NK-1R inhibition in wild type animals significantly reduces bacteria-induced neuroinflammation and resultant CNS damage [
8,
9]. NK-1R null mice and mice treated with an NK-1R antagonist showed reduced inflammatory and maintained immunosuppressive cytokine production, as well as decreased astrogliosis, cellularity, and demyelination following intracerebral administration of the Gram-negative bacterial pathogens
Neisseria meningiditis and
Borrelia burgdorferi, or the Gram-positive bacterium
Streptococcus pneumoniae [
8,
9]. More recently, we have demonstrated that the specific NK-1R antagonist, aprepitant, limits inflammatory nervous system immune responses in a nonhuman primate (NHP) model of Lyme neuroborreliosis [
10]. These animal studies therefore indicate that SP/NK-1R interactions are essential for the progression of damaging inflammation following bacterial CNS infection and raise the intriguing possibility that targeting the NK-1R could be useful as an adjunctive therapy for such conditions.
We have previously demonstrated that murine glial cells functionally express the NK-1R [
11]. Importantly, we have shown that SP can exacerbate the inflammatory responses of both murine microglia and astrocytes to
N. meningiditis and
B. burgdorferi [
9]. In the present study, we report that primary human glia and immortalized human glial cell lines, as well as NHP brain tissue, constitutively express robust levels of full-length NK-1R. Furthermore, we show that SP can augment the inflammatory and/or neurotoxic responses of human microglia and astrocytes to disparate and clinically relevant bacterial pathogens. Taken together, these results are consistent with our animal model studies and indicate that SP/NK-1R interactions could play a significant role in the initiation and/or progression of damaging inflammation in humans following bacterial CNS infection.
Methods
Bacterial propagation
First passage
B. burgdorferi strain B31 clone 5A19 spirochetes, isolated from an ear biopsy of a previously infected mouse, were grown in Barbour-Stoenner-Kelly-H medium supplemented with 6% rabbit serum and antibiotics (rifampicin at 45.4 μg/mL, phosphomycin at 193 μg/mL, and amphotericin at 0.25 μg/mL; Sigma-Aldrich, St. Louis, MO) to late logarithmic phase under microaerophilic conditions. An inoculum containing 1 × 10
7 spirochetes/mL in RPMI 1640 medium (Invitrogen, USA) was prepared for use in in vitro studies and to infect ex vivo NHP brain tissue as previously described [
12]. For in vitro human glia infection studies,
Neisseria meningitidis strain MC58 was cultured in Columbia broth on an orbital shaker at 37 °C with 5% CO
2 [
9].
Streptococcus pneumoniae strain CDC CS109, an isolate from a patient with meningitis, was grown from frozen stock on tryptic soy agar with 5% defibrinated sheep blood and subsequently cultured in Todd-Hewitt broth at 37 °C as previously described by our laboratory [
8].
Staphylococcus aureus strain UAMS-1 was grown in lysogeny broth (LB) on an orbital shaker at 37 °C with 5% CO
2 overnight.
Nonhuman primate frontal cortex brain slice isolation and ex vivo infection
Freshly harvested frontal cortex tissues were collected at necropsy from four rhesus macaques (
Macaca mulatta) that were scheduled for euthanasia due to chronic idiopathic diarrhea or had undergone trauma. Animals were euthanized in accordance with the recommendations of the American Veterinary Medical Association’s Panel on Euthanasia. The frontal cortex was sliced into 2-mm sections, and each section was placed in separate wells of 12-well plates. Each well contained 2 mL of RPMI 1640 medium (BioWhittaker, Walkersville, MD) supplemented with 10% FBS, as previously described [
13]. Tissue sections were exposed to medium alone or to medium containing
B. burgdorferi (1 × 10
7 bacteria/mL) and were processed for analysis at the indicated time points.
Source and propagation of human glial cell lines and primary cells
U87-MG, an immortalized human astrocytic cell line, was obtained from the ATCC (HTB-14). Cells were maintained in Eagle’s Minimum Essential Medium supplemented with 10% FBS and penicillin/streptomycin. The human microglial cell line, hμglia, was a kind gift from Dr. Jonathan Karn (Case Western Reserve University). These cells were derived from primary human cells transformed with lentiviral vectors expressing SV40 T antigen and hTERT and have been classified as microglia due to their microglia-like morphology; migratory and phagocytic activities; presence of the microglial cell surface markers CD11b, TGFβR, and P2RY12; and characteristic microglial RNA expression profile [
14]. This cell line was maintained in Dulbecco’s modified Eagle medium supplemented with 5% FBS and penicillin/streptomycin. Primary human astrocytes and microglia were purchased from ScienCell Research Laboratories (Carlsbad, CA) and were cultured in medium supplied by the vendor.
In vitro infection of human microglia and astrocytes and exposure to bacterial components
Cells (1.5 × 105) seeded in 12-well flat-bottom tissue culture plates were infected with bacteria at the indicated multiplicities of infection (MOI) in antibiotic-free culture medium for 2 h prior to washing and addition of complete culture medium. Alternatively, human glial cells were exposed to Pam3Cys, polyinosinic:polycytidylic acid (poly I:C) sodium salt, bacterial lipopolysaccharide (LPS), and/or flagellin, ligands for TLR2, TLR3, TLR4, and TLR5, respectively. Pam3Cys was purchased from InvivoGen (San Diego, CA). Flagellin (isolated from Salmonella typhimurium strain 14028) was purchased from Enzolife sciences (Farmingdale, NY). LPS (isolate from Escherichia coli) and poly I:C were purchased from Sigma-Aldrich (St. Louis, MO). Following infection or exposure to bacterial products, cells were then cultured in the presence or absence of SP (Sigma-Aldrich) at a concentration of 5 or 10 nM. At the indicated time points, whole-cell protein isolates were collected and RNA was isolated for immunoblot analysis and semi-quantitative RT-PCR, respectively.
RNA extraction and semi-quantitative reverse transcription PCR
Total RNA was isolated from cultured glial cells using Trizol Reagent (Thermo Fisher Scientific) according to the manufacturer’s instructions and quantified using a Nanodrop ND-1000 spectrophotometer. Prior to reverse transcription, RNA was treated with amplification grade DNase (Sigma-Aldrich) to remove genomic DNA. All RNA samples were diluted to the same concentration and reverse transcribed in the presence of random hexamers using 200 U of RNase H minus Moloney leukemia virus reverse transcriptase (Promega, Madison, WI) in the buffer supplied by the manufacturer. Semi-quantitative RT-PCR was performed on 5% of the reverse-transcribed cDNA product to assess the relative levels of expression of mRNA encoding NK-1R and the housekeeping gene product glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Positive and negative strand PCR primers used, respectively, were AACCCAAGTTCGAACCAG and ATGTACCTATCAAAGGCCACAGCC to amplify mRNA encoding total NK-1R, TCTTCTTCCTCCTGCCCTACATC and AGCACCGGAAGGCATGCTTGAAGCCCA to amplify mRNA encoding full-length NK-1R, ATCCTGGTGGCGTTGGCAGTC and GAGAGATCTGGCCATGTCCATAAAGA to amplify mRNA encoding preprotachykinin (PPT), and CCATCACCATCTTCCAGGAGCGAG and CACAGTCTTCTGGGTGGCAGTGAT to amplify mRNA encoding GAPDH.
Immunoblot analysis
Homogenates from NHP frontal cortical tissue and whole-cell protein isolates from human cell cultures were subjected to immunoblot analysis as we have previously described [
11,
15] using a mouse monoclonal antibody directed against human NK-1R (ThermoFisher Scientific; clone ZN003). In some experiments, nuclear protein extracts were obtained from hμglia cells as follows. Cells were suspended in a pH 7.9 lysis buffer containing 10 mM HEPES, 1.5 mM MgCl
2, 10 mM KCl, 0.5 mM DTT, 0.05% NP40, and protease inhibitor cocktail for 10 min at 4 °C. The nuclei and other fragments were pelleted by centrifugation and supernatants were retained as cytoplasmic fractions. Nuclei were lysed by exposure to pH 7.9 high salt buffer containing 5 mM HEPES, 1.5 mM MgCl
2, 0.2 mM EDTA, 0.5 mM DTT, 26% glycerol, and 300 mM NaCl for 30 min at 4 °C. Samples were cleared of cellular debris by centrifugation, and supernatants containing the nuclear fraction were subjected to immunoblot analysis using a mouse polyclonal antibody directed against the p65 (RelA) subunit of NF-kB (Millipore, Billerica, MA). Protein bands corresponding to NK-1R or RelA were detected using a Bio-Rad ChemiDoc imaging system, and quantification was performed using ImageLab software (Bio-Rad) normalized to the expression of the housekeeping gene product β-actin. NK-1R and p65 (RelA) protein expression is presented graphically as relative levels adjusted to β-actin expression, and the immunoblots shown are representative of at least three separate experiments.
ELISA quantification of SP levels and IL-6 production
Levels of SP in ex vivo NHP cortical supernatants were determined using a commercially available ELISA kit according to the directions provided by the manufacturer (R&D Systems). IL-6 production by human glial cultures was assessed by specific capture ELISA using a rat anti-human IL-6 capture antibody and a biotinylated rat anti-human IL-6 detection antibody (BD Pharmingen). Bound antibody was detected by streptavidin-horseradish peroxidase (BD Biosciences) followed by the addition of tetramethylbenzidine (TMB) substrate. H2SO4 was used to stop the reaction, and absorbance was measured at 450 nm. Dilution of recombinant IL-6 (BD PharMingen) was used to generate a standard curve, and the IL-6 concentration in each supernatant was determined by extrapolation of absorbances to the standard curve.
Fluorescent immunohistochemical analysis
Hμglia cells (1.5 × 105) were plated on acid-washed glass coverslips coated with poly-d-lysine. Cells were fixed (2% PFA), permeabilized (with 50% acetone 50% methanol solution), and blocked (5% goat serum). Cells were stained with a monoclonal mouse antibody directed against NK-1R (clone ZN003, Thermo Scientific, Rockford, IL) and a polyclonal goat antibody directed against the microglial marker Iba1 (Abcam, Cambridge, MA) prior to incubation with secondary antibodies coupled to Alexa Fluor 488 or Alexa Fluor 594. Samples were mounted with Prolong Gold containing DAPI (Invitrogen) and imaged using an Olympus 1X71 inverted microscope and an Olympus DP70 digital camera.
Flow cytometric analysis
U87-MG cells, primary human astrocytes, or hμglia cells, seeded in 12-well plates (1.5 × 105) were unstimulated or exposed to bacterial products for 2 h prior to addition of an enzyme free dissociation buffer (ThermoFisher Scientific), washing, and blocking (5% normal goat serum). Cells were then stained with a monoclonal mouse antibody directed against NK-1R (clone ZN003, Thermo Scientific) followed by incubation with a secondary antibody coupled to either Alexa Fluor 488 or Alexa Fluor 594, prior to flow cytometric analysis using an Accuri C6 cytometer (BD Biosciences, Franklin Lakes, NJ).
Assessment of soluble neurotoxic mediator production by infected human glia
Primary human astrocytes were uninfected or infected with B. burgdorferi or S. pneumoniae in the absence or presence of SP (5 nM). At 24 h following infection, conditioned medium was collected and placed on HCN-1A neuronal cells. At 24 h following addition of the conditioned medium, the viability of the HCN-1A cells was assessed by trypan blue exclusion in ten microscopy fields.
Statistical analysis
Data are presented as the mean ± standard error of the mean (SEM). Statistical analyses were performed using Student’s two-tailed t test or a one-way analysis of variance (ANOVA) with Bonferroni’s or Tukey’s post hoc tests as appropriate using commercially available software (GraphPad Prism, GraphPad Software, La Jolla, CA). In all experiments, results were considered statistically significant when a P value of less than 0.05 was obtained.
Discussion
Bacterial infections of the CNS constitute a group of highly damaging and often life-threatening diseases. What makes the etiology of these diseases so perplexing is that severe CNS inflammation can be initiated by bacterial species that are generally regarded to be of low virulence [
17]. While such responses may be protective, inflammation elicited by infectious agents often results in progressive CNS damage. Indeed, we have recently demonstrated that inflammation plays a key role in pathogenesis in a NHP model of acute Lyme neuroborreliosis [
18]. A hallmark of developing inflammation is the synergistic interaction between cells and their products that can amplify the response. It is now widely accepted that SP, the most abundant tachykinin in the CNS, can exacerbate the inflammatory responses of both leukocytes and resident glial cells via the high affinity full-length NK-1R isoform (as reviewed in [
1,
2]). Importantly, we have demonstrated that SP can augment proinflammatory mediator production by murine glia in response to bacterial challenge [
9]. Consistent with this finding, we have reported that endogenous SP/NK-1R interactions are required for maximal proinflammatory cytokine expression in vivo following direct CNS administration of
N. meningitidis or
B. burgdorferi in mice [
9]. In addition, we have shown that an NK-1R antagonist can attenuate the neuronal and glial production of inflammatory mediators including CCL2 and IL-6 in rhesus macaque frontal cortex explants and isolated DRG cells following
B. burgdorferi challenge [
19]. Furthermore, we have recently demonstrated that NK-1R antagonist treatment can attenuate aspects of the bacteria-induced inflammatory responses in CNS tissues in an in vivo NHP model of Lyme neuroborreliosis [
10].
In the present study, we have confirmed the robust expression of fNK-1R in NHP cortical brain tissue, with negligible expression of the truncated low affinity isoform (as described in [
20]) that has been reported to lack the ability to elicit proinflammatory responses in other cell types [
21,
22]. In contrast to our studies in the NHP brain cortex at 2 weeks following in vivo
B. burgdorferi infection [
10] and a report in the rat spine following chronic stress [
23], we have shown that acute ex vivo challenge with
B. burgdorferi fails to elicit significant changes in NK-1R mRNA or protein expression above constitutive levels. However,
B. burgdorferi infection did elicit a statistically significant elevation in SP protein levels within brain tissue, indicating that the expression of neurokinin signaling components can be modulated in situ in response to bacterial challenge.
We have previously documented the functional expression of NK-1R by peripheral myeloid immune cell types including macrophages and dendritic cells [
24,
25]. However, the expression of the SP receptor by microglia has been more contentious. Early findings indicated the absence of NK-1R expression by rat microglia based upon SP binding studies [
26] while another group reported the lack of NK-1R expression by activated rat microglia following cerebral ischemia [
27]. In contrast, one study reported the presence of NK-1R in human fetal microglia [
28] and we have previously shown the functional expression of NK-1R by primary murine microglia [
11]. In the present study, we have demonstrated the constitutive expression of full-length NK-1R protein by both a human microglial cell line and primary human microglia as determined by immunoblot analysis, immunohistochemical staining, and flow cytometry, at robust levels that could not be further elevated by exposure to bacterial ligands for TLR2, TLR4, or TLR5, either alone or in combination with SP treatment. In agreement with our results in ex vivo NHP cortical brain tissue, we were not able to detect significant levels of the truncated NK-1R isoform in human microglial cells. Furthermore, we have proved that fNK-1R is functionally expressed by human microglial cells with the demonstration that SP can elicit the activation of the critical proinflammatory transcription factor NF-kB, which is consistent with our prior studies in murine macrophages, dendritic cells, and microglia [
11,
16].
In contrast to microglia, the expression of NK-1R by astrocytes has been more clearly established with the demonstration of this receptor in primary cortical mouse and rat astrocytes [
29‐
31]. Furthermore, human brain astrocytes have been reported to express NK-1R, albeit at markedly lower levels than that seen in spinal cord cells [
32]. However, it should be noted that the NK-1R isoform expressed was not defined in these studies, and at least one group has failed to detect the presence of this receptor in activated rat astrocytes following an ischemic insult [
27]. Here, we show that both U87-MG human astrocytic cells and primary human cortical astrocytes express NK-1R mRNA and the full-length isoform protein as determined by immunoblot analysis and flow cytometry. Interestingly, we have found that exposure to bacterial components that serve as ligands for TLRs can elevate NK-1R mRNA expression and cell surface protein expression by U87-MG cells. Furthermore, challenge with bacteria or their products can elevate total cellular and cell surface NK-1R protein levels by primary human astrocytes. An elevation in NK-1R expression by astrocytes following exposure to activating stimuli is consistent with the documented ability of inflammatory mediators to increase NK-1R levels in U87-MG cells and primary rat astrocytes [
33] and leukocytes [
24,
34].
In accord with previous studies in human spinal astrocytes and primary rat astrocytes [
26,
32], SP failed to induce significant IL-6 production by either U87-MG cells or primary human astrocytes when used as the sole stimulus. However, SP significantly augmented cytokine responses by both cell types following exposure to bacterial TLR ligands. This finding is in agreement with the work of Luber-Narod and colleagues [
26] in rat astrocytes, but contrasts with another early report that SP does not affect the responses of human cortical astrocytes [
32]. Importantly, we have shown that SP can significantly elevate the production of IL-6 or soluble neurotoxic mediators induced by disparate Gram-negative and Gram-positive bacterial pathogens of the CNS, including
B. burgdorferi,
N. meningitidis,
S. pneumoniae and, to a lesser extent,
S. aureus.
Taken together, the robust constitutive and functional expression of the full-length NK-1R isoform by human microglia and astrocytes, and the ability of SP to augment inflammatory signaling pathways and mediator production by these cells, support the contention that SP/NK-1R interactions play a significant role in the damaging neuroinflammation and neurological sequelae associated with bacterial infections of the CNS in human subjects. Furthermore, given the available data that SP/NK-1R interactions also augment detrimental inflammation during parasitic CNS infections and perhaps multiple sclerosis, while contributing to neuroprotection during some degenerative CNS disorders and intracellular viral/bacterial infections (as discussed in [
2]), the functional expression of NK-1R by human glial cells may have broader implications. Clearly, further investigation of the ability of SP to augment CNS inflammation following infection and the benefits of targeting NK-1R in such clinical conditions is warranted.