Converging perturbed microvasculature and microglial clusters characterize Alzheimer disease brain.

We have investigated physical properties of microvasculature and vessel association with microglial clusters in cortical tissue from Alzheimer disease individuals, classified as severe (ADsev) or mild (ADmild), and nondemented controls (ND). Immunostaining with laminin or von Willerbrand factor demonstrated numbers of microvessels and microvascular density were significantly higher in ADsev cases compared with levels in ADmild or ND cases suggesting proangiogenic activity in ADsev brain. Evidence for extravascular laminin immunoreactivity was found in ADsev tissue and was largely absent in ADmild and ND cases suggesting vascular remodeling in ADsev brain included abnormalities in blood vessels. Microgliosis was progressively increased from ND to ADmild to ADsev with the latter demonstrating areas of clustered microglia (groupings of three or more cells) rarely observed in ADmild or ND cases. Microglial clusters in ADsev brain were in close proximity with extravascular laminin and also plasma protein, fibrinogen, implicating vascular perturbation as a component of inflammatory reactivity. ADsev brain also exhibited elevated levels of the pro-inflammatory/angiogenic factors tumor necrosis factor-α (TNF-α) and vascular endothelial growth factor (VEGF) in association, relative to non-association, with microglial clusters. The presence of extravascular laminin and fibrinogen and the vascular modifying factors, TNF-α and VEGF in localization with clusters of activated microglia, is consistent with microglial-induced vascular remodeling in ADsev brain. Microglial-vascular reciprocal interactions could serve a critical role in the amplification and perpetuation of inflammatory reactivity in AD brain.

Thalidomide inhibition of vascular remodeling and inflammatory reactivity in the quinolinic acid-injected rat striatum.

Effects of thalidomide administration on vascular remodeling, gliosis and neuronal viability have been studied in excitotoxin-injected rat striatum. Intrastriatal injection of quinolinic acid (QUIN) caused time-dependent changes (durations of 6 h, 1 and 7 d post-injection) in vascular remodeling. QUIN excitotoxic insult was associated with increased numbers of vessels (laminin or collagen IV markers) demonstrating considerable abnormalities in morphology, including short fragments and vascular loops. Non-lesioned striatum, with injection of phosphate buffer solution (PBS) as a vehicle, showed no evidence for vascular remodeling. A maximal extent of vascular remodeling was measured at 1 d post-QUIN and was correlated with marked increases in microgliosis (ED1 marker) and astrogliosis (glial fibrillary acidic protein [GFAP] marker) relative to control PBS injection. Double staining of laminin with ED1 and GFAP demonstrated areas of close association of glial cells with blood vessels. Treatment of QUIN-injected animals with the anti-inflammatory compound, thalidomide significantly inhibited vascular remodeling (by 43%) and reduced microgliosis (by 33%) but was ineffective in modifying extents of astrogliosis. Intrastriatal QUIN injection was associated with a marked loss of striatal neurons relative to non-lesioned control with thalidomide treatment exhibiting a significant degree of neuroprotection (24% recovery) against QUIN-induced neurotoxicity. These results suggest close links between microglial-mediated inflammatory responses and vascular remodeling, with inflammatory reactivity associated with, and contributing to, neuronal damage in excitotoxically-lesioned striatum.

Generation and characterization of immortalized human microglial cell lines: expression of cytokines and chemokines.

Microglia are a major glial component of the central nervous system (CNS), play a critical role as resident immunocompetent and phagocytic cells in the CNS, and serve as scavenger cells in the event of infection, inflammation, trauma, ischemia, and neurodegeneration in the CNS. Studies of human microglia have been hampered by the difficulty of obtaining sufficient numbers of human microglia. One way to circumvent this difficulty is to establish permanent cell lines of human microglia. In the present study we report the generation of immortalized human microglial cell line, HMO6, from human embryonic telencephalon tissue using a retroviral vector encoding myc oncogene. The HMO6 cells exhibited cell type-specific antigens for microglia-macrophage lineage cells including CD11b (Mac-1), CD68, CD86 (B7-2), HLA-ABC, HLA-DR, and ricinus communis agglutinin lectin-1 (RCA), and actively phagocytosed latex beads. In addition, HMO6 cells showed ATP-induced responses similar to human primary microglia in Ca2+ influx spectroscopy. Both human primary microglia and HMO6 cells showed the similar cytokine gene expression in IL-1beta, IL-6, IL-8, IL-10, IL-12, IL-15, and TNF-alpha. Using HMO6 cells, we investigated whether activation was induced by Amyloid-beta fragments or lipopolysaccharide (LPS). Treatment of HMO6 cells with Amyloid-beta 25-35 fragment (Abeta(25-35)) or Amyloid-beta 1-42 fragment (Abeta(1-42)) led to increased expression of mRNA levels of cytokine/chemokine IL-8, IL-10, IL-12, MIP-1beta MIP-1, and MCP-1, and treatment with LPS produced same results. Expression of TNF-alpha and MIP1-alpha was not detected in unstimulated HMO6 cells, but their expression was later induced by long-term exposure to Abeta(25-35) or Abeta(1-42.) ELISA assays of spent culture media showed increased protein levels of TNF-alpha and IL-8 in HMO6 cells following treatment with Abeta(25-35) or LPS. Taken together, our results demonstrate that treatment of human primary microglia and HMO6 immortalized human microglia cell line with Abeta(25-35), Abeta(1-42) and LPS upregulate gene expression and protein production of proinflammatory cytokines and chemokines in these cells. The human microglial cell line HMO6 exhibits similar properties to those documented in human microglia and should have considerable utility as an in vitro model for the studies of human microglia in health and disease.

Inhibition of store-operated Ca(2+) influx by acidic extracellular pH in cultured human microglia.

The effects of extracellular acidification on Ca(2+)-dependent signaling pathways in human microglia were investigated using Ca(2+)-sensitive fluorescence microscopy. Adenosine triphosphate (ATP) was used to elicit Ca(2+) responses primarily dependent on the depletion of intracellular endoplasmic reticulum (ER) stores, while platelet-activating factor (PAF) was used to elicit responses primarily dependent on store-operated channel (SOC) influx of Ca(2+). The duration of transient responses induced by ATP was not significantly different in standard physiological pH 7.4 (mean duration 30.2 +/- 2.5 s) or acidified pH 6.2 (mean duration 31.7 +/- 2.8 s) extracellular solutions. However, the time course of the PAF response at pH 7.4 was significantly reduced by 87% with external pH at 6.2. These results suggest that acidification of extracellular solutions inhibits SOC entry of Ca(2+) with little or no effect on depletion of ER stores. Changes of extracellular pH over the range from 8.6 to 6.2 during the development of a sustained SOC influx induced by PAF resulted in instantaneous modulation of SOC amplitude indicating a rapidly reversible effect of pH on this Ca(2+) pathway. Whole-cell patch clamp recordings showed external acidification blocked depolarization-activated outward K(+) current indicating cellular depolarization may be involved in the acid pH inhibition. Since SOC mediated influx of Ca(2+) is strongly modulated by membrane potential, the electrophysiological data suggest that acidification may act to inhibit SOC by cellular depolarization. These results suggest that acidification observed during cerebral ischemia may alter microglial responses and functions.

Acute actions of tumor necrosis factor-alpha on intracellular Ca(2+) and K(+) currents in human microglia.

The effects of acute application of the pro-inflammatory cytokine tumor necrosis factor-alpha (TNFalpha) on levels of intracellular Ca(2+) ([Ca(2+)]i) and on whole-cell outward and inward K(+) currents were studied in cultured human microglia. TNFalpha elicited a linear increase in [Ca(2+)]i to a plateau level in microglia bathed in either standard physiological saline solution or Ca(2+)-free physiological saline solution. The rate of increase of [Ca(2+)]i or the level of [Ca(2+)]i attained was not significantly altered in the absence of external Ca(2+) indicating that Ca(2+) influx did not contribute appreciably to the cytokine-induced rise in [Ca(2+)]i. This point was directly confirmed using Mn(2+) quenching where no change in signal fluorescence was observed with TNFalpha treatment of microglia in Ca(2+)-free physiological saline solution. The rate of increase of [Ca(2+)]i induced by TNFalpha in Ca(2+)-free physiological saline solution was not altered by prior application of ATP to deplete inositol triphosphate stores indicating that these stores did not contribute to the cytokine response. In whole-cell patch clamp recordings, the acute treatment of human microglia with TNFalpha led to the expression of an outward K(+) current in one-third (14 of 41) of cells. This current was activated at potentials positive to -30 mV, showed rapid kinetics of activation with no evident inactivation and had an I-V relation exhibiting outward rectification. Analysis of tail currents showed reversal of the outward K(+) current near -70 mV and tetraethylammonium (10 mM) inhibited the outward K(+) current to 24% of control level. Acute application of TNFalpha had no effect to alter inward rectifier currents generated from voltage ramps. The signaling pathways involving TNFalpha modulation of [Ca(2+)]i and K(+) channels in human microglia may contribute to functional and pathological actions of the cytokine in the brain.

Anion channels modulate store-operated calcium influx in human microglia.

Recent work from this laboratory has demonstrated that purinergic-mediated depolarization of human microglia inhibited a store-operated pathway for entry of Ca2+. We have used Fura-2 spectrofluorometry to investigate the effects on store-operated Ca2+ influx induced by replacement of NaCl with Na-gluconate in extracellular solutions. Three separate procedures were used to activate store-operated channels. Platelet activating factor (PAF) was used to generate a sustained influx of Ca2+ in standard physiological saline solution (PSS). The magnitude of this response was depressed by 70% after replacement of PSS with low Cl- PSS. A second procedure used ATP, initially applied in Ca2+-free PSS solution to deplete intracellular stores. The subsequent perfusion of PSS solution containing Ca2+ resulted in a large and sustained entry of Ca2+, which was inhibited by 75% with low Cl- PSS. The SERCA inhibitor cyclopiazonic acid (CPA) was used to directly deplete stores in zero-Ca2+ PSS. Following the introduction of PSS containing Ca2+, a maintained stores-operated influx of Ca2+ was evident which was inhibited by 77% in the presence of the low Cl- PSS. Ca2+ influx was linearly reduced with cell depolarization in elevated K+ (7.5 to 35 mM) suggesting that changes in external Cl- were manifest as altered electrical driving force for Ca2+ entry. However, 50 mM external KCl effectively eliminated divalent entry which may indicate inactivation of this pathway with high magnitudes of depolarization. Patch clamp studies showed low Cl-PSS to cause depolarizing shifts in both holding currents and reversal potentials of currents activated with voltage ramps. The results demonstrate that Cl- channels play an important role in regulating store-operated entry of Ca2+ in human microglia.

Activation of purinergic P2X receptors inhibits P2Y-mediated Ca2+ influx in human microglia.

Purinoceptor (P2X and P2Y) mediated Ca2+ signaling in cultured human microglia was studied using Ca2+ sensitive fluorescence microscopy. ATP (at 100 microM) induced a transient increase in [Ca2+]i in both normal and Ca(2+)-free solution suggesting a primary contribution by release from intracellular stores. This conclusion was further supported by the failure of ATP to cause a divalent cationic influx in Mn2+ quenching experiments. However, when fluorescence quenching was repeated after removal of extracellular Na+, ATP induced a large influx of Mn2+, indicating that inward Na+ current through a non-selective P2X-coupled channel may normally suppress divalent cation influx. Inhibition of Mn2+ entry was also found when microglia were depolarized using elevated external K+ in Na(+)-free solutions. The possibility of P2X inhibition of Ca2+ influx was then investigated by minimizing P2X contributions of purinergic responses using either the specific P2Y agonist, ADP-beta-S in the absence of ATP or using ATP combined with PPADS, a specific inhibitor of P2X receptors. In quenching studies both procedures resulted in large increases in Mn2+ influx in contrast to the lack of effect observed with ATP. In addition, perfusion of either ATP plus PPADS or ADP-beta-S alone caused a significantly enhanced duration (about 200%) of the [Ca2+]i response relative to that induced by ATP. These results show that depolarization induced by P2X-mediated Na+ influx inhibits store-operated Ca2+ entry resulting from P2Y activation, thereby modulating purinergic signaling in human microglia.

Acute application of interleukin-1beta induces Ca(2+) responses in human microglia.

The effects of the pro-inflammatory cytokine interleukin-1-beta (IL-1beta) on levels of intracellular calcium [Ca(2+)](i) in cultured human microglia have been studied using the fluorescent Ca(2+) indicator fura-2. IL-1beta (2 ng/ml) caused a slow, progressive increase in [Ca(2+)](i) in standard Ca(2+)-containing physiological solution (PSS). A similar effect was observed in separate studies using Ca(2+)-free PSS, however, the mean rate of increase was significantly lower than that measured with PSS. Similar results were obtained in a separate protocol, where cells were exposed to both IL-1beta in Ca(2+)-free PSS and PSS. The slope of the IL-1beta induced increase of [Ca(2+)](i) in Ca(2+)-free PSS was not altered when adenosine triphosphate was added prior to application of the cytokine. These results suggest that IL-1beta-induced responses in human microglia involve both a Ca(2+) entry pathway and a mechanism of intracellular increase other than from IP(3)-sensitive stores.

Effects of ATP and elevated K+ on K+ currents and intracellular Ca2+ in human microglia.

We have used whole-cell patch-clamp recordings and calcium microfluorescence measurements to study the effects of ATP and elevated external K+ on properties of human microglia. The application of ATP (at 0.1 mM) led to the activation of a transient inward non-selective cationic current at a cell holding potential of -60 mV and a delayed, transient expression of an outward K+ current activated with depolarizing steps applied from holding level. The ATP response included an increase in inward K+ conductance and a depolarizing shift in reversal potential as determined using a voltage ramp waveform applied from -120 to -50 mV. Fura-2 microspectrofluorescence measurements showed intracellular calcium to be increased following the application of ATP. This response was characterized by an initial transient phase, which persisted in Ca2+-free media and was due to release of Ca2+ from intracellular storage sites. The response had a later plateau phase, consistent with Ca2+ influx. In addition, ATP-induced changes in intracellular Ca2+ exhibited prominent desensitization. Elevated external K+ (at 40 mM) increased inward K+ conductance and shifted the reversal potential in the depolarizing direction, with no effect on outward K+ current or the level of internal Ca2+. The results of these experiments show the differential responses of human microglia to ATP and elevated K+, two putative factors associated with neuronal damage in the central nervous system.

Actions of the benzopyran compound terikalant on macroscopic currents in rat ventricular myocytes.

The putative Class III antiarrhythmic benzopyran compound terikalant (RP62719) has been applied to isolated rat ventricular myocytes. The drug, at extracellular concentrations from 1 to 60 microM, reduced the inactivation time constant of transient outward potassium current (I(to)) with the time constant decreased to 50% of the control value with terikalant at 11 microM. The peak value of I(to) was also diminished with terikalant in excess of 2 microM and analysis of the integral of charge movement showed this quantity to be halved with a drug concentration near 5 microM. The voltage dependence for both activation and inactivation of I(to) were not changed by terikalant and the drug had no effect on the time course of recovery from inactivation. The inhibition of I(to) currents was increased with time during depolarizing pulses suggesting drug interactions with the open channel and analysis of the time dependence of drug block gave estimates of 3.7 x 10(6) M-1 s-1 and 64 s-1 for the respective blocking and unblocking rate constants. At concentrations greater than 5 microM, terikalant also altered the peak amplitudes of inward rectifier K+ currents (IK1) elicited with hyperpolarizing or depolarizing steps from holding potential and diminished IK1 resulting from voltage ramps. The results of this study represent the initial characterization of terikalant actions in a species possessing abundant I(to) in ventricular myocytes.

Temperature dependence of unitary properties of an ATP-dependent potassium channel in cardiac myocytes.

The temperature dependence of the properties of unitary currents in cultured rat ventricular myocytes has been studied. Currents flowing through an ATP-dependent K+ channel were recorded from inside-out patches with the bath temperature varied from 10 degrees to 30 degrees C. The channel conductance was 56 pS at room temperature (22 degrees C), and the amplitudes of unitary currents and the channel conductance exhibited a relatively weak (Q10 from 1.4 to 1.6) dependence on temperature. The temperature dependence of channel mean open times was biphasic with the low temperature (10-20 degrees C) range showing a relatively stronger temperature dependence (Q10 of 2.3) than the high temperature (20-30 degrees C) range (Q10 of 1.6). The activation energies for the two regions were determined from an Arrhenius plot with the activation energy, corresponding to the lower temperature range, near 16 kcal/mol. Thermodynamic analysis, using transition rate theory, indicated that the formation of a transition state prior to channel closure to be associated with a positive entropy component for the high Q10 region.

Effects of tumor necrosis factor on inward potassium current and cell morphology in cultured human oligodendrocytes.

The effects of recombinant human tumor necrosis factor-alpha (rhTNF-alpha) on inward rectifier potassium [K(IR)] currents and on cell morphology have been studied in cultured human oligodendrocytes. Cell-attached patches were used to isolate and record unitary currents through an inward rectifier K+ channel with a conductance of 23 pS. In control oligodendrocytes the mean open times showed an exponential dependence on patch potential with an e-fold decrease over a patch hyperpolarization of 28 mV. Treatment of oligodendrocytes with rhTNF (at 250 ng/ml for 24-48 h) had significant actions to diminish the mean open times of K(IR) relative to control values. At cell resting potential the mean open times were reduced by 60% after rhTNF application; the amplitudes of unitary currents or the extrapolated zero-current potentials were not significantly changed by the cytokine. The rhTNF treatments were not cytotoxic to cultured human oligodendrocytes; however, in some experiments rhTNF caused evident retraction of cell processes.