Introduction
Fingolimod (FTY720, chemical name 2-amino-2[2-(4-octylphenyl)ethyl]-1-3-propanediol hydrochloride) is an analogue of sphingosine and is an FDA-approved oral immunomodulatory medication used to treat multiple sclerosis (MS). Like sphingosine, fingolimod undergoes phosphorylation to an active metabolite, fingolimod phosphate (FTY720-P), which is structurally analogous to sphingosine-1-phosphate (S1P) (David et al.
2012). FTY720-P functions as an agonist for four of the five isoform receptors of S1P, i.e., S1PR1, S1PR3, S1PR4, and S1PR5 (Don et al.
2007; Hogenauer et al.
2008; Zemann et al.
2006). Yet despite its immediate agonistic action, FTY720-P promotes endocytosis and degradation of the S1P receptors, functioning effectively as an antagonist. FTY720 can distribute across the blood–brain barrier, access the central nervous system (CNS), and localize to brain white matter, where it can alter neuroinflammatory responses of auto-reactive T cells and glial cells (Foster et al.
2007). In the rodent model of MS, experimental autoimmune encephalomyelitis (EAE), FTY720 ameliorated neurological symptoms (Fujino et al.
2003) in a manner that depended on astrocyte S1P1 receptor modulation (Choi et al.
2011). The preclinical data with EAE led to the successful human clinical trials of fingolimod to reduce relapses in MS; and fingolimod is now being explored in other neuroinflammatory and/or neurodegenerative disorders (O'Sullivan and Dev,
2017).
Fingolimod is thought to reduce relapse rates in MS primarily by altering lymphocyte trafficking, that is, by inhibiting egress of lymphocytes from the lymph nodes, preventing lymphocyte migration to the CNS, and thus reducing CNS inflammation. But clinical trials also revealed a potentially beneficial effect of fingolimod on brain volume; study patients on drug had reduced annual rates of brain volume loss (BVL) by about one-third relative to placebo (Calabresi et al.
2014; Kappos et al.
2010). This implies that fingolimod can be neuroprotective in CNS inflammatory disease states. Indeed, for over a decade, emerging lines of research have indicated that fingolimod may act directly on neuroepithelial-derived cell types through S1P receptor-mediated cell signaling events. Fingolimod has been demonstrated to provide neuroprotection against excitotoxic neuronal cell death in vitro and appears to do so via direct action on neuronal S1P1 receptors, independent of astrocytes, and using intracellular pathways that support the survival of neurons (Di Menna et al.
2013). In rodent neuronal cell cultures, fingolimod elicited a neuronal gene expression response that modulated the morphology of actin-rich neuronal growth cones and stimulated neurite growth in vitro (Anastasiadou and Knoll,
2016). There is some evidence in favor of neurotrophic activity (Groves et al.
2013) and reduction of amyloid-β accumulation in Alzheimer’s disease models (Asle-Rousta et al.
2013; Hemmati et al.
2013). But, there remain questions as to the mechanism and scope of fingolimod’s neuroprotective potential during chronic CNS inflammatory diseases. Moreover, there is little information on whether fingolimod can protect neurons in the neuroinflammatory and neurodegenerative microenvironment found in chronic HIV-1 infection of the aging brain and HIV-associated neurocognitive disorders (HAND) (Canizares et al.
2014).
The objective of this study is to identify potential neuroprotective pathways of fingolimod action by examining neuronal gene expression in an inflammatory microenvironment that mimics chronic HIV infection of the CNS. Our previous studies showed that when human neuroepithelial progenitor cells (NEP) are allowed to differentiate in vitro into a mixed population of astrocytes and neurons in the presence of HIV, neurons “fail to thrive” (Geffin et al.
2013; Martinez et al.
2012; McCarthy et al.
2006). There are lower total neurite lengths per cell and moderately decreased amounts of the neurofilament light protein (NF-L) in HIV-exposed neurons. Moreover, these effects occurred even without apparent evidence of productive viral infection (McCarthy et al.
2006). While our previous studies used a mixed glial and neuronal cell population differentiated from human fetal NEP (Geffin et al.
2013; Martinez et al.
2012; McCarthy et al.
2006), for this study, we have adapted a human neural progenitor cell line, hNP1, to emphasize neuronal-specific responses to HIV exposure and fingolimod treatment. The human neural progenitor cell line hNP1 was derived from the NIH-registered human embryonic stem cell (hESC) line WA09 and established as a neural progenitor line that can be differentiated into an enriched population of human post-mitotic neurons. The neurons have functional properties such as ion channels and ionotropic receptors (Dhara et al.
2008; Guo et al.
2013; Young et al.
2011). The hNP1 cells were exposed to HIV
SF2 MC or Mock-infected culture supernatants during differentiation conditions that produced a population of neuronal lineage cells enriched for post-mitotic neurons and lacking astroglia (Geffin et al.
2017). This directly exposed the neurons to viral proteins and associated inflammatory factors (Martinez et al.
2012), a microenvironment that can cause neuronal toxicity and even neuronal cell death (Hesselgesser et al.
1998; Maragos et al.
2003; van Marle et al.
2004). Fingolimod phosphate (FTY720-P) was added to the differentiating hNP1 cultures with or without HIV exposure to assess the effects of HIV and fingolimod exposure separately and in conjunction. Because the hNP1-derived neuronal cultures lack a glial fibrillary acidic protein (GFAP)-expressing astrocyte subpopulation, the observed fingolimod effects and differential gene expression profiles represent neuronal-specific responses. This study provides evidence that fingolimod can have downstream neuroprotective benefit through differential gene expression affecting cell signaling, neuronal glucose metabolism, and neuroinflammation.
Materials and methods
Culture and maintenance of hNP1 cells
Differentiation and culture treatment
Twenty-four hours after the hNP1 cultures were changed to differentiation medium, which was designated experimental day 0, the differentiating cultures were sorted into two groups for incubation with added fingolimod phosphate (FTY720-P) or no added FTY720-P. Each group was further divided into triplicate samples for three culture treatments: HIV-exposed, Mock-exposed, or untreated (Supplementary Fig. S
1). Thus, each culture condition was tested in triplicate with or without added FTY720-P. For all culture treatments, with and without FTY720-P, culture media with additives were fully replenished every 3 days.
Fingolimod treatment
Fingolimod phosphate (FTY720-P) was received as lyophilized powder from Dr. Paul Smith, Novartis Institutes for Biomedical Research, Basel, Switzerland. Powder was initially dissolved in acidified DMSO to a final concentration of 8 mM. This was then diluted to stock solutions of 1 mM and again to 10 and 1 μM in sterile phosphate-buffered saline (PBS). Dose–response curves were performed to assess cell viability. A final concentration of 10 nM FTY720-P was optimal and was used for the experiments described herein (see “
Results”). In experimental cultures, FTY720-P was diluted 1:100 from the stock 1 μM solution into culture medium and replenished daily. In controls not cultured with FTY720-P, the solvent for FTY720-P (acidified DMSO) was diluted 1:8000 in PBS, then 1:100 into culture medium, and also replenished daily. For each culture treatment (untreated, Mock-exposed, and HIV-exposed), separate triplicate cultures were treated with 10 nM FTY720-P or the control diluted solvent solution added to differentiation medium.
After 12 days of incubation, replicate monolayer cultures were harvested for either quantitative immunoblotting or RNA extraction.
Immunoblotting
The total protein content of whole cell lysates from cell cultures was determined by BCA protein assay (Pierce, Rockford, IL) and between 5 and 10 μg total protein was loaded onto 10% polyacrylamide gels. Quantitative immunoblotting was performed basically as described (McCarthy et al.
2006) with chemiluminescent signal detection to assess amyloid precursor protein (APP) expression. APP was detected with a 1:1000 dilution of rabbit polyclonal antibody (Cell Signaling Technology, Danvers, MA) and a secondary sheep anti rabbit IgG (GE Healthcare, UK) diluted to 1:2000. β-actin was used as an internal control for potential loading errors in each lane; it was detected with an anti-β-actin mouse monoclonal antibody (Sigma, St. Louis, MO). Immunoblots were developed with SuperSignalWest Femto Maximum Sensitivity Substrate (Thermo Scientific, Rockford, IL). Signals were viewed and quantitated using a Bio-Rad Chemidoc XRS+ Molecular Imager; the APP signal density was normalized to β-actin. Quantification of APP and β-actin signals used the average of signals from three independent immunoblot experiments.
RNA preparation and gene expression microarray generation
hNP1 cultures were incubated in differentiation medium and in the specified culture treatment conditions for 12 days, with replenishment of culture medium and additives at day 11. At day 12, the incubations were terminated and cultures were lysed for preparation of RNA for gene expression determinations. Total RNA was extracted from differentiating hNP1 cells using the Illustra RNAspin Mini kit from GE Healthcare (Pittsburgh, PA). RNA concentrations, 260/280, and 260/230 absorbance ratios were determined using a Nanodrop from Thermo Scientific (Wilmington, DE). Further RNA analyses and microarray data generation were performed by the technical staff of the Oncogenomics Core Laboratory, Sylvester Comprehensive Cancer Center (SCCC), Miller School of Medicine, and the University of Miami.
Total RNA was quantified with a Nanodrop 8000 Spectrophotometer (Thermo Scientific, Wilmington), and its quality was examined with a Bioanalyzer 2100 using the RNA 6000 Nano kit (Agilent, Santa Clara, CA). Biotinylated cRNA was prepared using the Illumina TotalPrep RNA Amplification kit (Ambion, Inc., Austin, TX) according to the manufacturer’s instructions starting with 400 ng total RNA. Successful cRNA generation was checked using the Bioanalyzer 2100. Samples were added to the Beadchip after randomization using the randomized block design to reduce batch effects. Hybridization to the Sentrix Human-HT12 Expression BeadChip (Illumina, Inc., San Diego, CA), washing, and scanning were performed according to the Illumina BeadStation 500 manual (revision C). The resulting microarray data was analyzed using Illumina GenomeStudio software.
Gene expression microarray analysis
Raw expression data from microarray Human HT-12_v4_0_R1 Illumina platform was obtained for day 12 and was loaded on Bioconductor/R software and a probe level analysis was performed on them. The raw intensity data from the microarray was normalized using the Quantile normalization method, log (base 2) transformed, and scaled the median to 0 of all samples. An unpaired Student’s t test for two class comparisons was used to identify the significantly differentially expressed genes between the tested conditions. Two-sided Student’s t test was performed with confidence level = 0.95 with two variances being treated as unequal. Multiple testing corrections were not performed on the nominal p values. Probes for different genes were considered significantly differentially expressed for those with absolute fold change value (|FC|) ≥ 1.5 and nominal p value ≤ 0.05. Functional enrichment analysis was performed on significant genes using the MetaCore module on GeneGO web portal program from Thomson Reuters™ (NY).
Gene expression validation by quantitative RT-PCR
hNP1 monolayers were incubated in differentiation medium for 12 days, then washed once with PBS, and lysed using the Illustra RNAspin Mini kit from GE Healthcare (Pittsburgh, PA) for RNA extraction. RNA concentrations, 260/280, and 260/230 absorbance ratios were determined using a Nanodrop from Thermo Scientific (Wilmington, DE). Samples were diluted to 10 ng/μl in nuclease-free water. One-step RT-PCR reactions were carried out using a Bio-Rad iTaq Universal SYBR Green 1-step kit from Bio-Rad with 50 ng of RNA per 50 μl reaction and 400 nM of all oligonucleotide primers except for those specific to Adaptor Related Protein Complex 2 Beta 1 Subunit component (AP2B1). Primers were designed using the NCBI primer-BLAST tool (
http://www.ncbi.nlm.nih.gov/tools/primer-blast/) with sequences found at the NCBI Refseq website (
https://www.ncbi.nlm.nih.gov/refseq/). Primers were designed to anneal to mRNA exons at the 5′ or 3′ side of the junction to avoid amplification of DNA sequences instead of RNA sequences. All primers were synthesized by Life Technologies (Carlsbad, CA). The qRT-PCR samples were run in a Stratagene Mx3005P thermocycler (Agilent Technologies, Santa Clara, CA). The following amplification conditions were used for all genes except AP2B1: 51 °C for 20 min for reverse transcription, an initial denaturing step at 95 °C for 10 min, and then 40 cycles of 95 °C for 10 s and 57 °C for 30 s for annealing and elongation. Melting curves were performed with an initial denaturing step at 95 °C for 1 min and annealing step at 55 °C for 1 min, then 95 °C for 30 s. RT-PCR conditions for amplification of the AP2B1 gene were slightly modified: the concentration of the reverse oligonucleotide primer was 600 nM. The amplification conditions were changed to 51 °C for 20 min for reverse transcription, an initial denaturing step at 95 °C for 10 min, and then 40 cycles of 95 °C for 15 s and 60 °C for 30 s for annealing and elongation. Data was analyzed using the MxPro software (Agilent Technologies, Santa Clara, CA).
Target gene expression was normalized to the housekeeping control gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) amplified from the same sample. Gene expression was calculated using the ddCT method (Vandesompele et al.
2002). For each culture condition, three independent experimental replicates, i.e., three independent cultures, were used to determine the expression of a gene by RT-PCR. First, triplicate samples from a single experimental replicate were assayed by one-cycle RT-PCR assay, and the mean of the normalized gene expression for that experimental replicate was calculated from the three sample replicates. Fold change in gene expression (FTY720-P versus no FTY720-P) or (Mock versus untreated) or (HIV versus untreated) or (HIV versus Mock) was calculated as a direct ratio of expression values for each individual experimental replicate.
Representative genes chosen for validation by quantitative RT-PCR were those with significant changes in gene expression when comparing added FTY720-P versus no added FTY720-P or when comparing culture treatments (e.g., Mock-exposed versus untreated) with or without added FTY720-P. After validation of the oligonucleotide primers designed for each representative gene, quantitative RT-PCR was performed using the same cellular RNA specimens that were used in the microarray analyses.
Oligonucleotide primers used for this study had the following sequences: annexin 1 (ANXA1) forward 5′GTGTGGCTTCCTTTAAAATC3′ and reverse 5′CATTTTCAATAAACCAGGCC3′; B2M forward 5′AGATAGTTAAGTGGGATCG3′ and reverse 5′AAAGTGTAAGTGTATAAGCAT3′; APP forward 5′CAAAACCTGCATTGATACCA3' and reverse 5′CATCACTTACAAACTCACCA3′; STAT1 forward 5′CAGTAAAGTCAGAAATGTG3′ and reverse 5′TTCATCTTGTAAATCTTCCA3′; TAF15 forward 5′CAGCAAAACATGGAATCATC3′ and reverse 5′CATATGAGCCTTGATGTTGA3′; RBM39 forward 5′AGCATCAAATTAAGACGACG3′ and reverse 5′CTTGCATCTCTTTCCTCAG3′; ID2 forward 5′CACCCTCAACACGGATATC3′ and reverse 5′TTCAGCACTTAAAAGATTCC3′; and AP2B1 forward 5′GGAAAACAGCAGCAGTCT3′ and reverse 5′ATTTCAGATAATGCCGCTACG3′.
Discussion
In this in vitro study, functional genomic analysis was applied to HIV-exposed human neurons and Mock-exposed controls in order to discover potential neuroprotective benefits of fingolimod phosphate (FTY720-P), an analogue of sphingosine-1-phosphate and a current immunomodulatory therapy for multiple sclerosis. In the absence of fingolimod, a similar number of genes were differentially expressed by Mock- and HIV-exposed as compared to untreated cultures. However, among the genes differentially upregulated, the proportion of genes associated with immune responses was almost double in the HIV-exposed cultures (17/30, 57%) compared to Mock-exposed cultures (10/33, 30%). This HIV effect was reinforced by data from the pairwise comparison between HIV-exposed versus Mock-exposed gene expression. All nine genes that were upregulated (FC ≥ 1.5) were immune response genes. When FTY720-P was added daily to the differentiation medium, the number of genes that were differentially expressed increased by 33% in HIV-exposed and 40% in Mock-exposed cultures as compared to untreated cultures. However, fingolimod affected the gene expression differently, depending on the culture treatment. Genes involved in glycolysis were upregulated in the presence or absence of FTY720-P in Mock-exposed versus untreated cultures. However, in HIV-exposed versus untreated cultures, upregulation of glycolysis-related genes was only significant in the presence of FTY720-P, not in its absence. Expression of genes related to immune responses increased in Mock-exposed and even more so in HIV-exposed neurons compared to untreated cultures. Fingolimod depressed the magnitude of this response but only in HIV-exposed neurons. These observations from the differentially expressed gene sets were reinforced by functional enrichment analysis. The top canonical pathways and GO processes were dominated by glycolysis and immune response functions in either Mock-exposed or HIV-exposed cultures. The number of genes enriching these pathways or processes was increased in the FTY720-P-treated cultures. When examining the direct effect of fingolimod on cellular pathways and processes within a single culture treatment, i.e., comparing FTY720-P versus no FTY720-P treatment, four canonical pathways were found to be significantly affected by fingolimod in all three culture treatments. The only gene enriching the pathways was FST, the inhibitor of activin-mediated signaling, which was downregulated by FTY720-P in all culture treatments. GO processes related to signaling or immune responses were significantly affected by fingolimod in all three culture treatments. But distinct genes enriched these processes for each culture treatment: among others, HLA-A, HLA-B, APP, and AP2B1 for HIV cultures; ANXA1 and FST for Mock-exposed cultures; and annexin I and the transcriptional regulators ID1, ID2, and ID3 for untreated cultures.
Clinical trial study extensions have shown that a reduced rate of brain volume loss is sustained in patients with relapsing MS receiving fingolimod continuously (De Stefano et al.
2017). This clinical result may reflect a direct neuroprotective effect of fingolimod in the CNS. Neuroprotection could be enabled by the cellular S1P receptors present on astrocytes, microglia, and neurons, as well as the drug’s accessibility through the blood–brain barrier (Foster et al.
2007). A role for astrocytes has been demonstrated in studies using both animal and cell culture models. In the rodent model of MS, experimental autoimmune encephalomyelitis (EAE), conditional null mouse mutants lacking S1P1 on GFAP-expressing astrocytes but not on neurons showed attenuation of both EAE disease and FTY720-P efficacy (Choi et al.
2011). In an astroglial cell culture model, addition of tumor necrosis factor-alpha (TNF-α) and FTY720-P to human primary embryonic astrocytes and astrocytoma cultures was used to mimic an inflammatory milieu in the CNS. Fingolimod treatment resulted in changes in astroglial gene expression, specifically, the induction of mRNA for neuroprotective factors leukemia inhibitory factor (LIF), interleukin 11 (IL11), and heparin-binding EGF-like growth factor (HBEGF), and the inhibition of TNF-induced inflammatory genes (CXCL10, BAFF, MX1, and OAS2) (Hoffmann et al.
2015). Our study also provides evidence that FTY720-P treatment might be directly neuroprotective in inflammatory milieus that may be toxic or detrimental to neurons. In the neuronal lineage-specific cell culture used in this study, differential gene expression is not confounded by the presence of other neural cell lineages and thus reflects the direct effect of fingolimod on neurons. Therefore, fingolimod’s neuroprotective potential may be inferred from its observed pleotropic effects on neuronal gene expression related to activin signaling, glucose metabolism, immune responses, and amyloid-β production.
Fingolimod treatment has a potentially neuroprotective effect through the differential down regulation in neurons of the gene for follistatin, a major antagonist of activin signaling. Specifically, FTY720-P treatment significantly downregulated follistatin gene expression in all culture treatments, albeit with slightly different FC values among the three culture treatments (untreated, Mock-exposed, HIV-exposed). Follistatin is an antagonist of several TGF-beta family members, mainly activin A. Follistatin binds activin A with great affinity and targets it for lysosomal degradation, thus neutralizing its biological function (Hedger et al.
2011). Activin A has multiple neurotrophic effects. In chicken embryos, activin A stimulates and follistatin inhibits neurite outgrowth in primary cultures from dorsal root ganglia (Fang et al.
2012) but does not inhibit nerve growth factor-induced growth. In an experimental mouse model of kainic acid-induced brain injury, the protective effects of basic fibroblast growth factor (bFGF) were mediated through activin A. The abolishment of activin A function by follistatin resulted in excitotoxic cell death (Tretter et al.
2000). Rat spiral ganglion neurons incubated with activin A, brain-derived neurotrophic factor (BDNF), and erythropoietin had significantly improved survival and neurite outgrowth as compared to control cultures not containing these factors (Kaiser et al.
2013). Even though the gene for activin A itself is not differentially upregulated by FTY720-P in any culture conditions, the reduction in the mRNA for its inhibitor, follistatin, can indirectly result in increased neurotrophic activin actions.
Analyses of both functionally enriched canonical pathways and differentially expressed gene sets indicate that differential gene expression related to neuronal glucose utilization is enhanced by fingolimod treatment. In the comparison of Mock-exposed versus untreated cultures, upregulation was significant for 11 glycolysis-related genes in the presence of FTY720-P and for nearly half those genes in the absence of FTY720-P. In the comparison of HIV-exposed versus untreated cultures, however, this upregulation was significant only in the presence of the drug and not in its absence for all genes examined. This implies that HIV exposure alone does not significantly enhance glycolysis-related gene expression compared to that in untreated cultures, while Mock exposure does. The functional link between FTY720-P treatment and more robust glucose metabolism in the HIV-exposed neurons may be through signaling pathways that FTY720 is known to affect. In particular, there may be a role for the Akt-mediated survival pathways in neurons. Fingolimod activates ERK and Akt signaling pathways through phosphorylation of the ERK1/ERK2 or Akt substrates, something we observed with the hNP1 cultures (data not shown). Akt signaling is linked to upregulation of glucose metabolism in proliferating cells and in particular in cancer cells (Simons et al.
2012). Akt has been shown to be activated by FTY720 in rodent brain (Ren et al.
2017; Zhang et al.
2016) and human neuroblastoma cells (Ren et al.
2017) under conditions of stress such as traumatic brain injury, illustrated in the rodent model (Zhang et al.
2016). Akt may increase expression of glycolysis genes through inhibition of transcription factors that otherwise repress the expression of glycolysis genes (Simons et al.
2012). Moreover, energy utilization is of paramount importance in the protection of neurons. In neurodegenerative diseases such as Alzheimer’s, low glucose metabolism has been associated with cognitive decline (Furst et al.
2012). In multiple sclerosis where axonal demyelination is one of the primary hallmarks of the disease, neurons increase energy consumption in order to restore axonal impulse conduction (Trapp and Stys,
2009). A recent study of human CNS-derived oligodendrocytes (Rone et al.
2016) highlights the role of glycolysis in energy production by cells under metabolic stress. Stress-induced reduction in glycolytic ATP production can exacerbate myelin process withdrawal and compromise metabolic support of neurons.
An increasing body of evidence goes against the notion that neurons are immunosilent; on the contrary, neuronal immune responses can play a significant role in defense against pathogens, whether they are neurons directly infected by neurotropic viruses (see review by Chakraborty et al. (
2010) or uninfected neurons in an inflammatory milieu (Boulanger and Shatz
2004). Fingolimod may modulate neuronal immune-related gene expression in an inflammatory milieu, leading to a net neuroprotective effect. In the HIV exposure paradigm of this study, the differentiating hNP1 cells are exposed to HIV proteins such as tat, gp120, nef, and vpr that can be neurotoxic (reviewed by Mocchetti et al. (
2012). In the presence or absence of FTY720-P, exposure to Mock and HIV supernatants significantly upregulated genes associated with immune responses, as compared to untreated cultures (Fig.
7a, b). Even comparing HIV-exposed to Mock-exposed cultures, differential gene expression is significant for a number of MHC class I antigen-presenting genes as well as beta-2-microglobulin (B2M) (Fig.
7c). Consistent with the neuronal expression profile, MHC class II genes were not upregulated. With or without FTY720-P treatment, there are similar fold changes for the pairwise comparison of Mock-exposed versus untreated cultures (Fig.
7b). However, fold change values trend lower for the pairwise comparison of HIV-exposed versus untreated cultures in the presence of FTY720-P (Fig.
7a). Among the differentially expressed immune genes, B2M expression is 33% lower in the HIV-exposed hNP1 cultures treated with FTY720-P. Beta2-microglobulin, a component of MHC class I molecules, is reportedly elevated in the CSF of patients with HIV-associated dementing illnesses and in patients with Alzheimer’s disease (Brew et al.
1996; Carrette et al.
2003). The active yet lower immune gene expression in HIV-exposed hNP1 neurons presents a potential for neuroprotective consequences of FTY720-P treatment. Fingolimod’s net effect on the immune gene expression would be to lower the intensity of inflammation that could otherwise be detrimental to neurons in chronic HIV infection and aging (Canizares et al.
2014).
Further potential for fingolimod-mediated neuroprotection is reflected in the downregulating effect of FTY720-P on the APP gene and APP protein expression in HIV-exposed neurons. This depressive effect, though modest, was significant. Moreover, together with interferon-related and MHC I antigen-presenting genes, APP mRNA was downregulated by FTY720-P treatment significantly but only in the context of HIV exposure. This suggests an interaction between the expression of the APP gene and the inflammatory milieu created by HIV exposure of the neurons. APP is cleaved by β- and γ-secretases resulting in the production of amyloid-β peptides (De Strooper et al.
2010). Aggregation and plaque formation by amyloid-β constitute one of the defining pathological findings in the brain of Alzheimer’s disease patients (Blennow et al.
2006). In HIV-infected individuals, cerebral amyloid-β plaques predict HAND (Soontornniyomkij et al.
2012).
S1P, an essential component of the membrane lipid bilayer, has been found to have profound effects on neurodegeneration by modulating the cleavage of APP into amyloid-β protein among other possible pathways (reviewed by van Echten-Deckert et al. (
2014)). Neurotoxicity mediated by amyloid-β protein was found to be ameliorated in primary mouse cortical neurons by fingolimod through upregulation of BDNF (Doi et al.
2013). An additional study, also using primary mouse cortical neurons, suggests that fingolimod can reduce amyloid-β protein production by a mechanism that is independent of the downstream SP1 receptor signaling pathways (Takasugi et al.
2013). In our study, expression of BDNF was not altered by addition of FTY-720 (data not shown), but protection against neurotoxicity mediated by amyloid-β might be achieved by a reduction of APP production.
This study adds to the body of evidence indicating that fingolimod can act directly on human neurons to modulate cell proliferation and survival. This study has the limitations inherent to the cell culture system to the extent that human neuronal cell lines can reflect the response of primary human neurons. In addition, technical detection of differentially regulated genes is limited by the number of probes included in the gene expression microarray platform and in the knowledge base of the software used for functional genomic analyses. But, given the study limitations, the data indicate that fingolimod has the capacity to alter neuronal gene expression in an inflammatory milieu, potentiating cellular responses that can have a net neuroprotective effect. That capacity is demonstrated here in the viral-specific inflammatory milieu generated by HIV exposure of the neurons. Through modest but significant downregulation of APP gene expression, fingolimod could protect against brain atrophy and cognitive decline associated with amyloid-β accumulation in chronic HIV infection. The drug fingolimod itself, as used in current neurological practice, is very problematic for the treatment of HIV infection or HAND since fingolimod’s actions on lymphocyte trafficking lead to a dramatic reduction in the number of circulating CD4-positive lymphocytes. But this study identifies neuroprotective avenues that can be targets for more specific S1P analogue development, particularly with respect to protection against amyloid-β accumulation. Thus, S1P signaling and related cell interactions can be exploited to treat and/or prevent neurodegenerative disease states such as HAND that are associated with chronic inflammation.