Chemokines regulate hippocampal neuronal signaling and gp120 neurotoxicity

Abstract

The HIV-1 envelope protein gp120 induces apoptosis in hippocampal neurons. Because chemokine receptors act as cellular receptors for HIV-1, we examined rat hippocampal neurons for the presence of functional chemokine receptors. Fura-2-based Ca imaging showed that numerous chemokines, including SDF-1alpha, RANTES, and fractalkine, affect neuronal Ca signaling, suggesting that hippocampal neurons possess a wide variety of chemokine receptors. Chemokines also blocked the frequency of spontaneous glutamatergic excitatory postsynaptic currents recorded from these neurons and reduced voltage-dependent Ca currents in the same neurons. Reverse transcription-PCR demonstrated the expression of CCR1, CCR4, CCR5, CCR9/10, CXCR2, CXCR4, and CX3CR1, as well as the chemokine fractalkine in these neurons. Both fractalkine and macrophage-derived chemokine (MDC) produced a time-dependent activation of extracellular response kinases (ERK)-1/2, whereas no activation of c-JUN NH2-terminal protein kinase (JNK)/stress-activated protein kinase, or p38 was evident. Furthermore, these two chemokines, as well as SDF-1alpha, activated the Ca- and cAMP-dependent transcription factor CREB. Several chemokines were able also to block gp120-induced apoptosis of hippocampal neurons, both in the presence and absence of the glial feeder layer. These data suggest that chemokine receptors may directly mediate gp120 neurotoxicity.

Figure 1

[Ca²⁺]_i transients evoked by chemokines and gp120s in hippocampal neurons. Chemokines that activate different CCRs (MDC, TARC, RANTES, MIP-1α, and I-309), CXCRs (IL-8 and SDF-1α), as well as CX₃CR1 (fractalkine) receptors were used for the times indicated by the bars (2–4 min). Representative traces from single neurons are shown in each panel. A total of 85 cells of 167 showed [Ca²⁺]_i increases when stimulated by hMDC (32 of 62 by mMDC). Nine of 34 neurons were responsive to TARC, 14 of 21 to RANTES, 12 of 19 to MIP-1α, 12 of 42 to I-309, 23 of 58 to fractalkine, 18 of 43 to IL-8, 21 of 44 to SDF-1α, 26 of 55 to HIV-1_IIIB gp120, and 9 of 23 to SIV_mac251 gp120.

Figure 2

Inhibition of spontaneous synaptic activity by chemokines. (A) [Ca²⁺]_i oscillations were blocked by hMDC and mMDC, fractalkine, RANTES, and SDF-1α in different neurons. In total, 18 of 37 oscillating neurons were responsive to mMDC (10 nM), 19 of 20 to fractalkine (100 nM), 7 of 7 to RANTES (50 nM), and 17 of 17 to SDF-1α (50 nM). (B) Fractalkine (50 nM) reduced the frequency, but not the amplitude, of excitatory postsynaptic currents recorded from hippocampal neurons (7 of 9 neurons). Similar results were obtained with mMDC (data not shown; 3 of 7 neurons). ∗ = P < 0.001 paired Student’s t test.

Figure 3

Reverse-transcription–PCR reaction products from RNA extracted from pure hippocampal neuronal cultures were run on 1.2% agarose gels. Arrows indicate expected PCR product sizes. Lanes labeled (-) represent water controls.

Figure 4

Effect of chemokines on ERK1/2, JNK1/2, p38, and CREB in hippocampal neuronal cultures. Western blots of total cell lysates (A) and of nuclear extracts (B) were obtained from pure neuronal cultures treated with 100 nM fractalkine after a 4-hr preincubation in balanced salt solution in the absence of glia. Antibodies selectively recognizing the activated form of ERK1/2, JNK1/2, p38, and CREB were used. Similar results were obtained with hMDC (10 nM; not shown). (C) Densitometric analysis of the effect of fractalkine (100 nM) on ERK1/2, JNK1/2, p38, and CREB. Each data point represents the average of two to three similar experiments. (D) Nuclear localization of activated CREB (pCREB) in neurons treated with 10 nM hMDC, 100 nM fractalkine, and 100 nM SDF-1α (10 min in the absence of glial feeder layer).

Figure 5

Effect of chemokines on neuronal survival. (A) Apoptosis induced by removal of the glial cell feeder layer was reduced by hMDC or mMDC (10 nM), RANTES (10 nM), fractalkine (100 nM), SDF-1α (50 nM), and TGF-β1 (5 ng/ml). Data are expressed as mean ± SEM of the percent of dead cells from five different cultures. In each experiment, two to three coverslips per treatment were evaluated and neurons were counted from 10–12 fields of each coverslip. The neurons analyzed for each data point were: 3409 (control with glia), 2828 (control without glia), 591 (TGF-β1), 1948 (MDC), 2376 (RANTES), 2095 (fractalkine), and 1499 (SDF-1α). ^ = P < 0.0001 vs. control with glia; ∗∗ = P < 0.0001 and ∗ = P < 0.001 vs. control without glia. (B) Neurotoxicity induced by HIV-1_IIIB gp120 and SIV_mac251 gp120 in hippocampal cultures. gp120_IIIB-induced neurotoxicity was inhibited by different types of chemokines. Chemokines (at the same concentrations reported above) and/or gp120 (200 pM) were added to the neuronal cultures at 7 days in culture and apoptotic cells counted after 3–4 days. The mean ± SEM of neuronal death from 12 separate experiments is reported and three to five coverslips (10–12 fields per coverslip) were counted from each experiment. The total number of cells counted for each treatment was: 6,145 (control), 6,812 (gp120_IIIB), 1,778 (gp120_mac251), 2,154 (hMDC), 2,615 (hMDC+gp120_IIIB), 2,367 (RANTES), 2,998 (RANTES+gp120_IIIB), 1,872 (fractalkine), 2,449 (fractalkine + gp120_IIIB), 3,125 (SDF-1α), and 3,750 (SDF-1α+gp120_IIIB). ^ = P < 0.0001 vs. control; ∗∗ = P < 0.0001 vs. gp120_IIIB alone. (C) HIV-1_IIIB gp120-induced neurotoxicity in neuronal cultures in the absence of glia. Neurons were treated for 24 hr with gp120 (200 pM) and/or hMDC (100 nM) or fractalkine (100 nM). Total number of cells counted from three different experiments: control = 2932, gp120 = 2740; MDC = 1,086, MDC + gp120 = 1,068, fractalkine = 1,300, fractalkine + gp120 = 1,631; ∗ = P < 0.01 vs. control.