• cell culture;
  • daidzein;
  • F-actin;
  • HAND ;
  • liquiritigenin;
  • rat


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
Thumbnail image of graphical abstract

HIV-1 infects the brain and, despite antiretroviral therapy, many infected individuals suffer from HIV-1-associated neurocognitive disorders (HAND). HAND is associated with dendritic simplification and synaptic loss. Prevention of synaptodendritic damage may ameliorate or forestall neurocognitive decline in latent HIV-1 infections. The HIV-1 transactivating protein (Tat) is produced during viral latency in the brain and may cause synaptodendritic damage. This study examined the integrity of the dendritic network after exposure to HIV-1 Tat by labeling filamentous actin (F-actin)-rich structures (puncta) in primary neuronal cultures. After 24 h of treatment, HIV-1 Tat was associated with the dendritic arbor and produced a significant reduction of F-actin-labeled dendritic puncta as well as loss of dendrites. Pre-treatment with either of two plant-derived phytoestrogen compounds (daidzein and liquiritigenin), significantly reduced synaptodendritic damage following HIV-1 Tat treatment. In addition, 6 days after HIV-1 Tat treatment, treatment with either daidzein, or liquiritigenin enhanced recovery, via the estrogen receptor, from HIV-1 Tat-induced synaptodendritic damage. These results suggest that either liquiritigenin or daidzein may not only attenuate acute synaptodendritic injury in HIV-1 but may also promote recovery from synaptodendritic damage.

The HIV-1 transactivating protein (Tat) is produced during viral latency in the brain. Treatment with either daidzein or liquiritigenin restored the loss of synaptic connectivity produced by HIV-1 Tat. This neurorestoration was mediated by estrogen receptors (ER). These results suggest that plant-derived phytoestrogens may promote recovery from HIV-1-induced synaptodendritic damage.

Abbreviations used

Combination anti-retroviral therapy




HIV-1 associated neurocognitive disorders





HIV-1 enters the central nervous system early in the infection process and may ultimately result in an array of deficits, collectively known as HIV-1-associated neurocognitive disorders (HAND) (Antinori et al. 2007). Although the incidence of the most severe dementia has decreased as the development of combination anti-retroviral therapy (cART), neurocognitive deficits continue to persist in more than 40% of HIV-1-infected individuals (Lindl et al. 2010; Letendre 2011). The continued prevalence of HAND despite the success of cART (Ances and Ellis 2007; Lindl et al. 2010; Letendre 2011) indicates a need to understand the underlying mechanism(s) of HAND and identify effective treatments.

Neuronal cell death has been observed in post-mortem brain tissue of patients with HAND; however, cell death does not correlate well with neurocognitive status (Adle-Biassette et al. 1999; Kaul et al. 2001; Ellis et al. 2007). Dendritic pruning, decreases in spine density, and degradation of synaptic proteins all correlate more readily with HIV-1-induced neurocognitive decline than does cell death (Masliah et al. 1997; Adle-Biassette et al. 1999; Kaul et al. 2001). Moreover, loss of dendrites and decreased synaptic density in the frontal cortex from HIV+ patients has been reported with latent viral infection (Desplats et al. 2013).

During latent HIV-1 infection, the provirus is incorporated into cellular DNA, without the expression of viral RNA (Wu 2004). Current anti-retroviral drugs effectively suppress peripheral viral load, but do not prevent the continued latent production of HIV-1 neurotoxic proteins (i.e. Tat) once the proviral DNA is incorporated into the brain (Li et al. 2009). HIV-1 provirus containing cells can thereby produce and release Tat into the extracellular space (Pugliese et al. 2005; Bachani et al. 2013) and Tat can interact with the cell surface of other, non-infected cells, such as neurons. HIV-1 Tat protein has been found to decrease dendritic spine and synaptic density in vitro and in vivo (Everall et al. 1999; Kim et al. 2008; Fitting et al. 2010). Therefore, a more detailed understanding of how HIV-1 Tat interacts with the dendritic network may aid in preventing neuropathological dendritic pruning and synaptic loss during latent HIV-1 infection.

Filamentous actin (F-actin) is one of the major cytoskeletal proteins that make up pre- and post-synaptic structures. Polymerization of globular actin (G-actin) into F-actin, a more stable form of actin, occurs prior to spinogenesis (Johnson and Ouimet 2006). Although dendritic spines are rich in F-actin, where F-actin is found in the spine head and shaft (Sekino et al. 2007; Dent et al. 2011), dendritic spines are not the only F-actin rich structures found on the neurite (Halpain et al. 1998; Lau et al. 1999; Johnson and Ouimet 2006; Hotulainen et al. 2009). F-actin is associated with both pre- and post-synaptic structures (Johnson and Ouimet 2006), including non-spiny synapses; therefore, changes in the F-actin rich structures, i.e. dendritic F-actin puncta, suggest overall alterations in synaptic connectivity not limited to spines. Phalloidin, a form of phallotoxin isolated from the death cap mushroom (Amanita phalloides), selectively binds to F-actin, but not monomeric G-actin, and has been used to examine the dynamic activity of F-actin in modulating synaptic plasticity (Kaech et al. 1997; Halpain et al. 1998; Hotulainen et al. 2009; Korobova and Svitkina 2010).

Interestingly, estrogens modulate spine concentrations of F-actin (Kramar et al. 2009; Sanchez et al. 2009), suggesting estrogenic therapeutic approaches to correcting F-actin loss and restoration of neuroplasticity. Phytoestrogens are plant-derived diphenolic compounds found to partially mimic mammalian estrogen in structure and function (Glazier and Bowman 2001; Lephart et al. 2005). Daidzein (DAI) is a phytoestrogen isoflavone found in soybeans (Mortensen et al. 2009). DAI has been shown to be neuroprotective against glutamate excitotoxicity and Aβ25-35-induced apoptosis (Zhao et al. 2002). We have reported that DAI is protective against acute HIV-1 Tat-induced apoptosis (Adams et al. 2012).

The novel flavonoid, liquiritigenin (LQ), is one of the active compounds of MF101, a herbal remedy used to treat menopausal symptoms (Mersereau et al. 2008). LQ has been isolated from Chinese licorice root, glycyrrhiza uralensis, and identified as a selective estrogen receptor beta (ERβ) agonist (Paruthiyil et al. 2009). Previous studies have shown that LQ has anti-inflammatory effects, prevents neurocognitive deficits and neurotoxicity induced by the Aβ protein in animal models (Liu et al. 2009, 2010, 2011). LQ has been used in Traditional Chinese Medicine to treat senile dementia (Lin et al. 2012). Together, the ERβ specificity of LQ as well as the ability of LQ to prevent neurocognitive decline in animal models, suggest that LQ could be neuroprotective against HIV-1 Tat-mediated cell death and synaptodendritic damage.

Neurocognitive deficits associated with latent HIV-1 infection continue to persist despite successful suppression of peripheral viral load by anti-retroviral therapy (Heaton et al. 2010; Desplats et al. 2013). Synaptodendritic injury is closely tied to cognitive decline in HAND (Ellis et al. 2007). However, there are currently no treatments to prevent or attenuate synaptodendritic injury that occurs during HIV-1 infection. In this study, the neurorestorative potential of DAI and LQ was evaluated by monitoring their ability to prevent, and enhance recovery from, synaptodendritic injury produced in vitro by the HIV-1 protein, Tat 1-86B. The effects of DAI and LQ on restoration of the dendritic network were determined following withdrawal of Tat, and the role of estrogen receptors in the phytoestrogen-induced recovery process was determined through the use of tamoxifen. Collectively, these studies determined the neuroprotective and neurorestorative potential of DAI and LQ treatment in HIV-1 Tat-induced synaptodendritic injury.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Primary cell cultures

Primary cell cultures were prepared from embryonic day 18 Sprague-Dawley rat fetuses as previously described (Aksenova et al. 2009; Bertrand et al. 2011). Timed-pregnant females were obtained from Harlan Laboratories Inc., Indianapolis, IN, USA. In brief, hippocampal regions were dissected and incubated for 10 min in a solution of 2 mg/mL trypsin in Hank's balanced salt solution (HBSS) buffered with 10 mM HEPES (GIBCO Life Technologies, Grand Island, NY, USA). The tissue was then exposed for 2 min to soybean trypsin inhibitor (1 mg/mL in HBSS) and rinsed three times in HBSS. Cells were dissociated by trituration and distributed to poly-l-lysine-coated glass bottom 35 mm dishes (MatTek Corporation, Ashland, MA, USA) for cytomorphological studies and 96-well plastic plates (Costar, Cambridge, MA, USA) for cell viability measurements. Initial plating density in 96-well plates was approximately 160–180 cells/mm2; in the 35 mm dishes the initial density was 120–140 cells/mm2. A lower plating density was used in 35 mm dishes for clear identification of second order branching patterns. At the time of plating, culture dishes contained Dulbecco's modified Eagle's medium/Ham's nutrient mixture F-12 (GIBCO, Life Technologies) supplemented with 10% fetal bovine serum (Sigma Chemicals, St. Louis, MO, USA). After a 24-h period, Dulbecco's modified Eagle's medium/Ham's nutrient mixture F-12 was replaced with an equal amount of Neurobasal (serum-free) medium, without phenol red, supplemented with 2% v/v B27, 2 mM GlutaMAX supplement, and 0.5% w/v d-1 glucose (all ingredients from GIBCO). Cultures were maintained at 37°C in a 5% CO2/95% room air-humidified incubator at all times. Serum-free culture medium was supplemented at weekly intervals. Cultures were used for experiments at the age of 21 days in vitro (DIV) and were > 85–90% neuronal as determined by anti-MAP-2/anti-GFAP/Hoechst fluorescent staining (Bertrand et al. 2011).

All animal protocols were reviewed and approved by the Animal Care and Use Committee at the University of South Carolina (animal assurance number: A3049-01).

Experimental treatment of cell cultures

The treatment of hippocampal cell cultures with HIV-1 Tat was carried out by the addition of 10 μL freshly prepared stock solutions of recombinant Tat 1-86B (LAI/Bru strain of HIV-1 clade B, GenBank accession no. K02013) (Diatheva, Fano, Italy) to the serum-free cell culture growth medium (50 nM final concentration). An equal volume of vehicle was added to control cell cultures. Prior work has shown that Tat 1-86B effects on the synaptodendritic arbor to be highly selective and dependent upon the presence of an intact cysteine domain, which is present in HIV-1 Tat 1-86B (Bertrand et al. 2013).

Hippocampal cell cultures were treated with DAI (≥ 98.5% purity; Indofine Chemical Hillsborough, NJ, USA), LQ (≥ 98.5% purity; Indofine Chemical Hillsborough) and tamoxifen (TMX) (Tamoxifen citrate; Tocris Bioscience, Ellisville, MD, USA). DAI and TMX were initially dissolved in dimethylsulfoxide and then diluted in phosphate-buffered saline (PBS). LQ was initially dissolved in methanol and then diluted in PBS. For the studies of acute injury and cell death, the cells were treated with DAI and LQ for 24 h prior to Tat treatment and assessed at either 24 h (F-actin puncta) or 48 h (cell death). For recovery studies, DAI or LQ, (with or without TMX) were added to cultures 6 days after Tat treatment with assessments of F-actin puncta conducted 3 days later (i.e. 9 days after initial Tat treatment). TMX was used as initially described (Murphy and Segal 1996) in studies of primary hippocampal cell cultures/spine morphology to modulate spine density. TMX was found to block the 17-β-estradiol-induced increase in spine density in the cultured hippocampal cells (Segal and Murphy 2001), and not to increase neural process outgrowth or morphological complexity (O'Neill et al. 2004). For our hippocampal cell culture/morphological endpoints, the long-standing evidence in literature supported the use of TMX.

Cell viability

Hippocampal cell viability was assessed in primary hippocampal cell cultures prepared in 96-well culture plates (Costar) using the microplate reader-formatted variant of the fluorescent calcein AM/ethidium bromide cell labeling assay (Live/Dead kit; Invitrogen Life Technologies) as described previously (Aksenov et al. 2009; Aksenova et al. 2009; Adams et al. 2010, 2012). Initial cell viability measurements were carried out after 1, 24, 72, 96, and 144 h incubation periods with Tat 1-86B (50 nM). Subsequent cell viability measurements were carried out after a 48-h incubation period with Tat 1-86 B (50 nM) which was preceded by a 24-h pre-treatment with DAI or LQ. Fluorescence was measured using a Bio-Tek Synergy HT microplate reader (Bio-Tek Instruments Inc., Winooski, VT, USA). Data from treated cells were normalized to data from untreated control cells in adjacent wells from the same plate and are presented as a percent of control. For cell viability tests, groups of 7–8 culture wells were analyzed. Two replicates for each experiment were performed using cell culture preparations from different animals.

Fluorescent labeling and immunocytochemistry

The immunofluorescent labeling of primary hippocampal cell cultures was carried out in glass bottom cell culture dishes (MatTek Corporation). F-actin was visualized using a modified protocol for filamentous actin (F-actin) staining (Invitrogen Life Technologies). Briefly, treated and control cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100 and then treated for 20 min at 21°C with an F-actin specific probe, AlexaFluor 488 Phalloidin (1 : 40) (Invitrogen Life Technologies). Following F-actin labeling, cells were incubated with 10% normal horse serum (NHS) in PBS to block non-specific binding and co-labeled with rabbit polyclonal anti-MAP-2 (1 : 1000) (Santa Cruz Biotechnology, Inc. Santa Cruz, CA, USA) or anti-Tat rabbit polyclonal antibody (1 : 1000) (Diatheva) overnight. The secondary antibody for MAP-2 and Tat labeling was Alexa Red 594-conjugated goat anti-rabbit IgG (1 : 500, Invitrogen Life Technologies). Identification of cell nuclei was accomplished using direct Hoescht dye techniques (ImmunoChemistry Technologies LLC, Bloomington, MN, USA).

F-actin puncta

Images of co-labeled F-actin/MAP-2 neurons were acquired using a high-resolution CCD camera attached to the Nikon Eclipse TE2000-E inverted fluorescent computer-controlled microscope (20X objective, 1600 × 1200 pixel image size, 0.17 μm/px image resolution at 1X zoom; Nikon Instruments, Melville, NY, USA). Between 3 and 5 Green (F-actin)/Red (MAP-2) immunolabeled/Blue (Hoescht) fluorescent images of individual neurons with clearly defined dendritic arbors and normal nuclear morphology per culture well were randomly selected and analyzed using the NIS-Elements software package (Nikon Instruments). For each neuron selected, F-actin rich structures were identified in several (3–4) second order dendritic fragments (length range 25–75 μm) with continuous MAP-2 immunofluorescence. Fine filapodia, spine protrusions, and F-actin patches were considered F-actin rich structures; in contrast, growth cones (F-actin rich structures located at the most distal dendritic terminus) were excluded (Bertrand et al. 2013).

Trained independent observers manually counted F-actin puncta from identically treated cultures at different times, with a very high correlation (r2 = 0.97), indicating that the green F-actin puncta were readily identified.

For computer-assisted methods, F-actin labeled puncta were identified by subtracting 20–50 au from the green channel to set the threshold for baseline staining of the dendritic shaft. Density was calculated by dividing total F-actin labeled puncta (N) by the length (L) of the MAP-2 labeled dendrites. For this study, puncta (size ≤ 1.5 μm) of F-actin fluorescence with a peak intensity of at least 50% above the average intensity of staining in the dendritic shaft were included in each selected dendritic fragment.

Statistical analysis

Statistical comparisons were made using anova techniques with specific a priori contrasts and regressions, and were used to determine specific treatment effects. Pearson's product-moment correlation coefficient was calculated to verify inter-rater reliability and correlations between computer-assisted profiling and manual counting. Comparisons and correlations were calculated using BMDP version 2009 (Statistical Solutions, Saugus, MA, USA). Significant differences were set at < 0.05. GraphPad/Prism (V5.02; GraphPad Software, San Diego, CA, USA), was used for time course analyses.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Tat 1-86B synaptodendritic injury and cell death time course

After treatment with 50 nM Tat 1-86B cell viability was monitored at 1, 24, 48, 72, 96, and 144 h (Fig. 1a). Tat 1-86B failed to induce significant cell death through 24 h. Tat 1-86B induced significant cell death at 48 h (< 0.001). Cell death reached a plateau at 72 h, with no additional changes at 96 or 144 h.


Figure 1. Time-course of Tat-induced cell death and synaptodendritic alterations. (a) The viability of primary cultures was assessed at 1, 24, 48, 72, 96, and 144 h after treatment with Tat. There was no significant cell death through 24 h; however, by 48 h there was a significant decrease in cell viability (< 0.001). By 72 h, cell death reached a plateau and remained stable through 144 h. Mean ± 95% confidence interval (CI). (b) F-actin/MAP-2 staining of cell cultures at 1, 24, 72, and 144 h after Tat 1-86B exposure. Tat 1-86B treatment produced a simplification of the fine network, as well as bundling and fragmentation of neurites at 24 h, which preceded Tat-mediated cell death. Tat-induced alterations in the neuronal network persisted through 144 h. (c) Representative images (20X) of second branch order neuritic fragments co-stained with Alexa 488 phalloidin and rabbit polyclonal anti-Tat primary/Alexa Red 594. Tat was associated with the F-actin labeled network at 1 h and 24 h post-treatment.

Download figure to PowerPoint

Cell cultures were co-stained for F-actin and MAP-2 at 1, 24, 72, and 144 h after addition of Tat (50 nM). Tat exposure resulted in degradation of pyramidal cell dendrites, indicated by loss of F-actin staining (Fig. 1b). In addition to simplification of the cellular network, the presence of fibril bundling and fragmentation suggested microtubule dysfunction. Tat 1-86B-mediated F-actin network degradation occurred earlier than Tat-mediated cell death, with degradation observed at 24 h.

Figure 1c illustrates the association of Tat 1-86B (red) with the neuronal network (green) at 1 and 24 h in primary hippocampal cell cultures. Subsequent acute experiments were performed at 24 h because of the visible change in F-actin rich structures, lack of overt cell death, and persistence of Tat 1-86B within the neuronal network.

Computer-assisted identification of F-actin puncta

Computer-assisted intensity profiling was used to determine the number of F-actin rich structures within selected dendritic segments. Figure 2a shows representative selections of Tat 1-86B-treated and vehicle-treated controls. The computer-assisted profiles of MAP-2 immunofluorescence (25–75 μm length; intensity range 37–110 arbitrary units, au) in cultures were continuous, indicating the absence of fragmentation of MAP-2 labeled dendrites, regardless of treatment group. The peak F-actin fluorescence (puncta) varied from 60–200 au within the dendritic segments.


Figure 2. Computer-assisted identification of F-actin puncta after Tat treatment. (a) Segments of second order dendritic bran-ches (shown as boxes) were selected from 20X images of F-actin/MAP2/Hoechst-stained Tat-treated and non-treated control neurons. The computer-assisted intensity profile of F-actin fluorescence (green) showed numerous peaks corresponding to F-actin rich structures and valleys corresponding to low baseline staining of dendritic shafts. The computer-based inten-sity profile of MAP2-specific immunoflu-orescence (red) showed uniform labeling of all dendritic segments selected for counting of the F-actin puncta density. (b) The total number of F-actin puncta (N; green +) was divided over the length of the neurite (L; green line) to provide the mean number of F-actin labeled puncta per 10 μm of neuronal dendrite ± SEM. F-actin puncta were significantly reduced in neurons exposed to Tat for 24 h using both computer-assisted intensity profiling and manual counting methods. F-actin puncta are readily distinguished and there was a very high correlation between the two methods (r2 = 0.97).

Download figure to PowerPoint

As shown in Fig. 2b, F-actin rich puncta per 10 μm of F-actin/MAP-2 labeled dendrite, was significantly (non-overlapping distributions) decreased in neurons exposed to 50 nM Tat for 24 h using both computer-assisted and manual counting methods. There was a very high correlation between the two methods (r2 0.97).

Phytoestrogens are not neurotoxic at 1.0 μM doses

An overall significant effect of phytoestrogens in protecting against Tat-induced cell death was suggested (F(4,71) = 54.4, p < 0.001), with a most prominent linear effect of dose (F(1,71) = 166.4, p < 0.001) and a less prominent quadratic effect of dose (F(1,71) = 7.3, p < 0.001).

To determine the ability of DAI to provide protection against neuronal death, cells were pre-treated with DAI for 24 h and then incubated with Tat 1-86B for 48 h. There was an overall significant main effect of DAI treatment on viability F(4,36) = 40.2, < 0.001, as pre-treatment with DAI protected against Tat-induced toxicity (Fig. 3a). Planned comparisons indicated that cell viability following treatment with 1.0 μM DAI alone was not significantly different from untreated control cultures (F < 1.0), suggesting that DAI alone did not produce neurotoxicity at the highest dose. A quadratic dose-dependent effect of DAI was found (F(1,36) = 32.6, p < 0.001), with complete neuroprotection achieved, as pre-treatment with DAI prevented Tat 1-86B-induced cell death in both the 0.2 and 1.0 μM DAI treatment groups (Fs(1,36) > 78.0, ps < 0.001).


Figure 3. Pre-treatment with phytoestrogens provides dose-dependent protection against neuronal cell death. (a) Cell viability after daidzein (DAI) 24 h pre-treatment and a 48 h incubation with 50 nM Tat 1-86B. Treatment with 1 μM DAI alone had no significant effect on cell viability relative to control cultures (F < 1.0). DAI pre-treatment provided complete neuroprotection in both the 0.2 μM and 1 μM treatment groups (ps < 0.001). Results are presented as mean % of control ± SEM. *Indicates significant protection of Tat-induced neurotoxicity. (b) Cell viability after 24 h liquiritigenin (LQ) pre-treatment and a 48 h incubation with 50 nM Tat 1-86B. There was no significant difference between LQ-treated cultures and controls, indicating 1 μM LQ was not neurotoxic. A linear dose-dependent effect of LQ was found (p < 0.001) with significant, although not complete, neuroprotection. Results are presented as mean % of control ± SEM. *Indicates significant protection of Tat-induced neurotoxicity.

Download figure to PowerPoint

To determine the ability of LQ to provide protection against cell death, cells were pre-treated with LQ and then incubated with Tat 1-86B for 48 h. There was an overall significant main effect of LQ treatment on viability (F(4,35) = 25.7, < 0.001), as pre-treatment with LQ protected against Tat-induced toxicity (Fig. 3b). There was no significant difference between LQ-treated cultures and controls, indicating that LQ was not neurotoxic (F < 1.0). A linear dose-dependent effect of LQ was found (F(1,35) = 77.6, p < 0.001), with significant, although not complete, neuroprotection. Specifically, planned comparisons indicated that pre-treatment with LQ provided significant prevention of Tat-mediated cell death in both the 0.2 and 1.0 μM LQ dose groups (Fs(1,35) > 7.2, ps < 0.01), but that both LQ dose groups displayed significantly decreased cell viability relative to controls (Fs(1,35) > 7.6, ps < 0.009).

Based on the lack of neurotoxicity of 1.0 DAI and 1.0 LQ (i.e. phytoestrogen treatments not different from untreated controls), the 1.0 μM doses were used in subsequent studies of dendritic recovery.

Phytoestrogen pre-treatment provides protection against acute HIV-1 Tat-mediated synaptodendritic injury

An overall significant effect of phytoestrogens in protecting against acute Tat-induced synaptodendritic injury was suggested by the significant phytoestrogen treatment effect (F(1,58) = 8.9, p < 0.004) and the significant interaction between Tat and phytoestrogen treatments (F(1,58) = 4.9, p < 0.03), without any significant difference between the two phytoestrogens (DAI vs. LQ) in their neuroprotection (Fs < 1.0).


Fine network integrity was assessed by counting F-actin puncta after pre-treatment with DAI (Fig. 4a) and then 24 h of incubation with Tat 1-86B (Fig. 4b). Planned contrasts indicated that DAI was not significantly different from controls (F < 1.0), suggesting that DAI was neither toxic nor stimulatory. However, Tat 1-86B produced a significant loss of F-actin puncta (F(1,31) = 10.2, p < 0.003), and DAI provided significant protection against the synaptodendritic damage/puncta loss caused by Tat (F(1,35) = 4.7, p < 0.039).


Figure 4. Pre-treatment with DAI protects against Tat-induced loss of F-actin puncta. (a) Chemical structure of daidzein (7,4′-Dihydroxyisoflavone). (b) F-actin puncta following pre-treatment (24 h) with 1 μM daidzein (DAI) and incubation (24 h) with 50 nM Tat 1-86B. DAI was neither toxic nor stimulatory to production of F-actin puncta; however, Tat treatment produced a significant loss of F-actin puncta (p < 0.003), and DAI provided significant protection against the puncta loss caused by Tat. Results are presented as mean number of F-actin labeled puncta per 10 μm of neuronal dendrite ± SEM. *Indicates a significant loss of F-actin puncta after Tat 1-86B treatment when compared with vehicle-treated controls (< 0.01). *Indicates 1 μM DAI pre-treatment provided significant protection from damage induced by Tat 1-86B (50 nM; 24 h) when compared with cultures treated with Tat 1-86B alone. (c–e) Neurons from (c) vehicle-treated control cultures, (d) Tat 1-86B-treated (50 nM; 24 h) cultures, and (e) pre-treated DAI (1 μM; 24 h) + incubated with Tat 1-86B (50 nM; 24 h) cultures displaying typical F-actin (green), MAP-2 (red) and Hoescht (blue) staining for each treatment group. The control image shows robust F-actin presence, complex branching patterns, and an extensive fine network. Tat 1-86B treatment induced a simplification of the network. In contrast, in the DAI pre-treated culture, Tat 1-86B failed to cause network simplification, suggesting DAI pre-treatment protected against Tat-induced synaptodendritic alterations.

Download figure to PowerPoint

Representative images (Fig. 4c–e) show that relative to control cultures (Fig. 4c), Tat treatment induced a simplification of the network (Fig. 4d). Moreover, there was no significant difference between DAI+Tat-treated cultures and controls (Fig. 4e), indicating that DAI pre-treatment protected against F-actin puncta loss induced by Tat 1-86B.


To determine the ability of LQ to provide synaptodendritic protection, cultures were pre-treated with LQ (Fig. 5a) and then incubated with Tat 1-86B (Fig. 5b). Planned comparisons indicated that pre-treatment with LQ did not alter puncta density from controls (F < 1.0), suggesting that LQ was neither toxic nor stimulatory. However, Tat 1-86B again produced a significant loss of F actin puncta (F(1,27) = 13.8, p < 0.001), and LQ provided significant protection against the synaptodendritic damage/puncta loss caused by Tat (F(1,27) = 6.7, p < 0.016).


Figure 5. Pre-treatment with liquiritigenin (LQ) provides protection against Tat-induced loss of F-actin puncta. (a) Chemical structure of liquiritigenin (4′,7-Dihydroxyflavanone). (b) Quantification of F-actin puncta with a pre-treatment (24 h) of 1 μM LQ and 50 nM Tat 1-86B (24 h). LQ did not alter F-actin puncta density from control (F < 1.0), suggesting that LQ was neither toxic nor stimulatory. Tat treatment produced a significant loss of F-actin puncta (p < 0.001) and LQ provided significant protection against the puncta loss caused by Tat (p < 0.016). Results are presented as mean number of F-actin labeled puncta per 10 μm of neuronal dendrite ± SEM. *Indicates a significant loss of F-actin puncta after Tat 1-86B treatment when compared with vehicle-treated controls (< 0.01). *Indicates 1 μM LQ pre-treatment provided significant protection from F-actin puncta loss induced by Tat 1-86B (50 nM; 24 h) when compared with cultures treated with Tat 1-86B alone. (c–e) Neurons from (c) vehicle-treated control cultures, (d) Tat 1-86B-treated (50 nM; 24 h) cultures, and (e) pre-treated LQ (1 μM; 24 h) + incubated with Tat 1-86B (50 nM; 24 h) cultures, displaying typical F-actin (green), MAP-2 (red) and Hoescht (blue) staining for each treatment group. The control image shows robust F-actin presence, complex branching patterns, and an extensive fine network. Tat 1-86B treatment induced a simplification of the dendritic network. In contrast, in the LQ pre-treated culture, Tat 1-86B failed to cause network simplification, indicating that LQ pre-treatment protected against Tat-induced synaptodendritic alterations.

Download figure to PowerPoint

Representative images are shown in Fig. 5c–e. Relative to controls (Fig. 5c), Tat treatment produced significant damage to the network (Fig. 5d). The difference between LQ+Tat-treated cultures and controls was not significant (Fig. 5e), indicating that pre-treatment with LQ protects against Tat 1-86B-induced loss of F-actin rich structures.

Neurorestoration of prior HIV-1 Tat-mediated synaptodendritic injury via by Daidzein and Liquiritigenin

After the cytotoxic effect of Tat was fully developed (6 days), the Tat-containing medium was replaced with fresh medium (i.e. without Tat). Three days after medium replacement, a significant decrease (F(1,64) = 32.2, p < 0.001) in the F-actin puncta remained in cultures with prior Tat exposure, relative to controls (Fig. 6). The cell cultures treated with DAI, LQ, or TMX were not significantly different from control (Fs < 1.0), suggesting a lack of toxicity from these compounds. However, when either 1.0 μm LQ or DAI were included in the replaced medium, a significant increase in F-actin puncta was found (F(1,64) = 10.6, p < 0.002). To determine if the DAI or LQ neurorestoration was dependent on ER-signaling, a similar experiment was performed in the presence of tamoxifen (TMX). Inclusion of TMX with either DAI or LQ in the cultures with prior Tat exposure blocked the restoration of F-actin puncta by the phytoestrogens (F(1,64) = 8.0, p < 0.006).


Figure 6. Phytoestrogens enhance the recovery of F-actin puncta from HIV-1 Tat-induced synaptodendritic injury in an estrogen receptor-dependent mechanism. The neurorecovery experiment was initiated by medium replacement after 6 days (i.e. after the initial cytotoxic effects of Tat). Three days after medium replacement (9 days after initial Tat treatment) a significant loss of F-actin puncta remained (p < 0.001) in Tat-treated cultures. The cell cultures treated on day 6 with either daidzein (DAI), liquiritigenin (LQ) or tamoxifen (TMX) were not significantly different from controls. When either DAI or LQ were added on day 6 to the cultures initially treated with Tat, a significant increase in F-actin puncta was detected (p < 0.002). This enhancement by DAI or LQ was blocked by the estrogen receptor antagonist TMX (100 nM), suggesting involvement of estrogen receptors in mediating the recovery of F-actin puncta. Results are presented as mean number of F-actin labeled puncta per 10 μm of neuronal dendrite ± SEM. *Indicates significant differences (< 0.05) between indicated groups.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The major findings of this study are (i) acute pre-treatment with either LQ or DAI, two plant-derived phytoestrogenic compounds, prevents the loss of F-actin puncta induced by HIV-1 Tat. Synaptodendritic damage was quantified by the significant loss of F-actin puncta after a short exposure to Tat, and Tat was found to be acutely associated with the dendtric arbor. Moreover, (ii) these phytoestrogens were able to promote restoration of F-actin dendritic puncta and dendrites following damage from HIV-1 Tat in hippocampal cells. This recovery occurred via an estrogen receptor-mediated mechanism, as restoration of F-actin puncta was blocked by tamoxifen. These results suggest that DAI, LQ, and compounds of similar chemical structures, could be beneficial in HIV-1 therapeutics because of their ability to either prevent, and/or enhance recovery, from synaptodendritic injury.

The brain serves as a reservoir for HIV-1, as cART effectively suppresses viral replication in the periphery, but does not eradicate the virus from the brain (Heaton et al. 2010). Synaptodendritic injury has been found in latent HIV-1 infection, in which active viral replication is not present in the brain, despite the presence of proviral DNA. Latent HIV-1 infection in humans produces dendritic loss (decreased MAP2 staining) and synaptic loss (decreased synaptophysin staining) in the frontal cortex without viral replication (Desplats et al. 2013). HIV-1 proviral DNA may produce early proteins, such as Tat, without productive viral infection (Wu 2004). In this study, the co-localization of Tat within the F-actin rich dendritic network and the concomitant loss of puncta suggests a mechanistic interaction between Tat and synaptodendritic injury; however, it is currently unknown whether the synaptodendritic loss in human latent HIV-1 infections is caused by Tat expression per se. Nevertheless, our in vitro model of HIV-1 Tat exposure demonstrates dendritic loss (decreased MAP2 staining) and synaptic loss (decreased F-actin puncta), similar to that observed in post-mortem tissue from humans brains with latent HIV-1 infection.

F-actin puncta encompass a variety of neuronal structures, the most well understood being dendritic spines. Dendritic spines are F-actin rich protrusions from the dendritic arbor, are the sites of post-synaptic excitatory synapses, and change shape rapidly in response to extracellular signaling (Kaech et al. 1997; Calabrese et al. 2006; Hotulainen et al. 2009; Dent et al. 2011). Non-spiny synapses, typically GABA-ergic (Craig et al. 1994), have also been found to contain a significant concentration of F-actin and appear as patches or ‘hot spots’ on the dendritic arbor (Halpain et al. 1998; Lau et al. 1999). We found that intensity profiling, a semi-automatic computer-based method, and manual counting produced similar puncta densities in vitro. As a result of the varying morphology of spines, the ability of F-actin staining to detect non-spiny synapses, and the early filamentous stages of spinogenesis, the quantification of all F-actin rich structures, or puncta, provides a measurement of overall synaptic integrity, health of the dendritic network, and potential to recover from Tat-induced injury.

Although F-actin is known to play a role in spinogenesis and synaptic plasticity, the staining of F-actin is a relatively new technique used to image dendritic spines and monitor synaptic integrity (Matus et al. 1982; Allison et al. 1998; Sekino et al. 2007). In comparison, the Golgi method has been used for well over a century to identify changes in spine morphology and density, and is still in use. The Golgi method has been favored over the years because of the particular ability of Golgi to produce images with little to no background and complete neuronal filling. However, Golgi randomly and unpredictably stains neurons and has been found to significantly underestimate spine number and, moreover, fails to identify non-spiny thin synaptic structures and patch morphology (Mancuso et al. 2012). Non-spiny inhibitory synapses, which can be detected as patches by F-actin staining, but are not detected by the Golgi method, are important for normal communication between neurons (Heller et al. 2012).

In this study, we used F-actin to assess the synaptodendritic injury induced by HIV-1 Tat, whereas other studies have used Golgi (Sa et al. 2004; Fitting et al. 2010), green fluorescent protein coupled to post-synaptic density 95 (GFP-PSD95) labeling (Kim et al. 2008; Shin et al. 2012), or MAP-2 staining (Maragos et al. 2003) to evaluate the damage to the dendritic arbor. Golgi detects changes in spine morphology, but not changes in non-spiny synapses or thin filopodia. GFP-PSD95 staining can detect only post-synaptic excitatory structures, as PSD-95 has not been found in non-spiny inhibitory synapses (Heller et al. 2012) and is not found in pre-synaptic structures. Although pre- and post-synaptic terminals generally correspond to one another (Craig et al. 1994), PSD-95 can still be located after F-actin has disappeared from the spine (Halpain et al. 1998), suggesting that F-actin may be a more sensitive marker. MAP-2 staining can detect changes in overall structure of the neuronal arbor, but cannot reliably detect subtle changes in spine or patch morphology as microtubules have been observed to enter dendritic spines selectively and microtubule entry is activity dependent (Dent et al. 2011). Although the aforementioned techniques vary, and the studies determined the effects of HIV-1 Tat both in vivo (Maragos et al. 2003; Fitting et al. 2010) and in vitro (Kim et al. 2008; Shin et al. 2012), collectively these studies indicate that HIV-1 Tat produces dendritic damage. Therefore, identifying mechanisms or compounds that either prevent or reverse synaptodendritic damage inflicted by HIV-1 Tat may be useful approaches for providing neuroprotection during HIV-1 infection.

As a potential therapeutic approach for recovery of synaptodendritic damage, estrogen has long been known to rapidly increase dendritic spine density in vivo (Gould et al. 1990; Woolley and McEwen 1992), specifically through regulation of the actin cytoskeleton in spines (Sanchez et al. 2009). Phytoestrogens are non-steroidal, diphenolic structures found in plants that have similar chemical and structural properties to 17-β-estradiol (Glazier and Bowman 2001; Lephart et al. 2005). We found that phytoestrogens promote recovery from the synaptodendritic injuries produced by HIV-1 Tat. Although synaptodendritic injury is correlated with the symptoms of HAND (Ellis et al. 2007; Desplats et al. 2013), the extent to which synaptic restoration is possible remains unknown. Also unknown is whether there might be a critical therapeutic window for promoting effective recovery wherein neurorestoration remains possible. Unfortunately, studies of therapeutic pathways for enhancing dendritic recovery are few (Kim et al. 2008; Shin et al. 2012); however, estrogen has been shown to promote spine formation via modulation of F-actin in spines (Kramar et al. 2009; Sanchez et al. 2009). The current data suggest that phytoestrogens, possibly acting via F-actin, may provide a novel intervention for promoting neurorestoration. The activation of estrogen receptors has been shown to play a key role in modulation of dendritic spine dynamics (Liu et al. 2008; Kramar et al. 2009; Sanchez et al. 2009; Srivastava et al. 2010; Phan et al. 2011), suggesting the estrogen receptor may be a useful target in ameliorating synaptodendritic injury, such as that seen in HAND.

Previous studies have shown that LQ is a highly specific ERβ agonist, with a 75-fold binding preference to ERβ over ERα (Mersereau et al. 2008) and only activates ERβ (Kupfer et al. 2008; Paruthiyil et al. 2009). DAI preferentially binds to ERβ, with a 14-fold selectivity for ERβ (Zhao et al. 2009). The selectivity of LQ for ERβ suggests that LQ acts through an ERβ dependent mechanism to prevent synaptodendritic damage induced by HIV-1 Tat; however, more experimentation is needed to determine the receptor-dependent mechanism of LQ.

Our results illustrate that HIV-1 Tat causes an early reduction in F-actin positive puncta. Interventions aimed at promoting synaptodendritic integrity following HIV-1 infection of the nervous system would therefore appear to be an effective approach for preventing HAND. Moreover, our results indicate that damage by HIV-1 Tat may be repaired by the phytoestrogens, liquiritigenin and daidzein, and estrogen receptor actions mediate this neurorestoration. Although it is presently unknown if such reversals can improve neurocognitive outcomes, phytoestrogenic compounds, like DAI and LQ, may prevent cumulative injury to the dendritic network, and ultimately, aid recovery from HIV-associated neurocognitive disorders.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

This work was funded by NIH grants DA013137, DA031604, DA035714 and HD043680. Partial support was provided by a NIH T32 training grant in Biomedical-Behavioral science. Tori D. Espensen-Sturges is now pursuing her graduate studies at the University of Minnesota. The authors regretfully acknowledge the untimely passing of Dr. Michael Aksenov, who contributed much to this work. The authors have no conflicts of interest to declare.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • Adams S. M., Aksenova M. V., Aksenov M. Y., Mactutus C. F. and Booze R. M. (2010) ER-beta mediates 17beta-estradiol attenuation of HIV-1 Tat-induced apoptotic signaling. Synapse 64, 829838.
  • Adams S. M., Aksenova M. V., Aksenov M. Y., Mactutus C. F. and Booze R. M. (2012) Soy isoflavones genistein and daidzein exert anti-apoptotic actions via a selective ER-mediated mechanism in neurons following HIV-1 Tat(1-86) exposure. PLoS ONE 7, e37540.
  • Adle-Biassette H., Chretien F., Wingertsmann L., Hery C., Ereau T., Scaravilli F., Tardieu M. and Gray F. (1999) Neuronal apoptosis does not correlate with dementia in HIV infection but is related to microglial activation and axonal damage. Neuropathol. Appl. Neurobiol. 25, 123133.
  • Aksenov M. Y., Aksenova M. V., Mactutus C. F. and Booze R. M. (2009) Attenuated neurotoxicity of the transactivation-defective HIV-1 Tat protein in hippocampal cell cultures. Exp. Neurol. 219, 586590.
  • Aksenova M. V., Aksenov M. Y., Adams S. M., Mactutus C. F. and Booze R. M. (2009) Neuronal survival and resistance to HIV-1 Tat toxicity in the primary culture of rat fetal neurons. Exp. Neurol. 215, 253263.
  • Allison D. W., Gelfand V. I., Spector I. and Craig A. M. (1998) Role of actin in anchoring postsynaptic receptors in cultured hippocampal neurons: differential attachment of NMDA versus AMPA receptors. J. Neurosci. 18, 24232436.
  • Ances B. M. and Ellis R. J. (2007) Dementia and neurocognitive disorders due to HIV-1 infection. Semin. Neurol. 27, 8692.
  • Antinori A., Arendt G., Becker J. T. et al. (2007) Updated research nosology for HIV-associated neurocognitive disorders. Neurology 69, 17891799.
  • Bachani M., Sacktor N., McArthur J. C., Nath A. and Rumbaugh J. (2013) Detection of anti-tat antibodies in CSF of individuals with HIV-associated neurocognitive disorders. J. Neurovirol. 19, 8288.
  • Bertrand S. J., Aksenova M. V., Aksenov M. Y., Mactutus C. F. and Booze R. M. (2011) Endogenous amyloidogenesis in long-term rat hippocampal cell cultures. BMC Neurosci. 12, 38.
  • Bertrand S. J., Aksenova M. V., Mactutus C. F. and Booze R. M. (2013) HIV-1 Tat protein variants: critical role for the cysteine region in synaptodendritic injury. Exp. Neurol. 248, 228235.
  • Calabrese B., Wilson M. S. and Halpain S. (2006) Development and regulation of dendritic spine synapses. Physiology (Bethesda) 21, 3847.
  • Craig A. M., Blackstone C. D., Huganir R. L. and Banker G. (1994) Selective clustering of glutamate and gamma-aminobutyric acid receptors opposite terminals releasing the corresponding neurotransmitters. Proc. Natl Acad. Sci. USA 91, 1237312377.
  • Dent E. W., Merriam E. B. and Hu X. (2011) The dynamic cytoskeleton: backbone of dendritic spine plasticity. Curr. Opin. Neurobiol. 21, 175181.
  • Desplats P., Dumaop W., Smith D., Adame A., Everall I., Letrendre S., Ellis R., Cherner M., Grant I. and Masliah E. (2013) Molecular and pathologic insights from latent HIV-1 infection in the human brain. Neurology 80, 14151423.
  • Ellis R., Langford D. and Masliah E. (2007) HIV and antiretroviral therapy in the brain: neuronal injury and repair. Nat. Rev. Neurosci. 8, 3344.
  • Everall I. P., Heaton R. K., Marcotte T. D., Ellis R. J., McCutchan J. A., Atkinson J. H., Grant I., Mallory M. and Masliah E. (1999) Cortical synaptic density is reduced in mild to moderate human immunodeficiency virus neurocognitive disorder. HNRC Group. HIV Neurobehavioral Research Center. Brain Pathol. 9, 209217.
  • Fitting S., Xu R., Bull C., Buch S. K., El-Hage N., Nath A., Knapp P. E. and Hauser K. F. (2010) Interactive comorbidity between opioid drug abuse and HIV-1 Tat: chronic exposure augments spine loss and sublethal dendritic pathology in striatal neurons. Am. J. Pathol. 177, 13971410.
  • Glazier M. G. and Bowman M. A. (2001) A review of the evidence for the use of phytoestrogens as a replacement for traditional estrogen replacement therapy. Arch. Intern. Med. 161, 11611172.
  • Gould E., Woolley C. S., Frankfurt M. and McEwen B. S. (1990) Gonadal steroids regulate dendritic spine density in hippocampal pyramidal cells in adulthood. J. Neurosci. 10, 12861291.
  • Halpain S., Hipolito A. and Saffer L. (1998) Regulation of F-actin stability in dendritic spines by glutamate receptors and calcineurin. J. Neurosci. 18, 98359844.
  • Heaton R. K., Clifford D. B., Franklin D. R. Jr et al. (2010) HIV-associated neurocognitive disorders persist in the era of potent antiretroviral therapy: CHARTER Study. Neurology 75, 20872096.
  • Heller E. A., Zhang W., Selimi F., Earnheart J. C., Slimak M. A., Santos-Torres J., Ibanez-Tallon I., Aoki C., Chait B. T. and Heintz N. (2012) The biochemical anatomy of cortical inhibitory synapses. PLoS ONE 7, e39572.
  • Hotulainen P., Llano O., Smirnov S., Tanhuanpaa K., Faix J., Rivera C. and Lappalainen P. (2009) Defining mechanisms of actin polymerization and depolymerization during dendritic spine morphogenesis. J. Cell Biol. 185, 323339.
  • Johnson O. L. and Ouimet C. C. (2006) A regulatory role for actin in dendritic spine proliferation. Brain Res. 1113, 19.
  • Kaech S., Fischer M., Doll T. and Matus A. (1997) Isoform specificity in the relationship of actin to dendritic spines. J. Neurosci. 17, 95659572.
  • Kaul M., Garden G. A. and Lipton S. A. (2001) Pathways to neuronal injury and apoptosis in HIV-associated dementia. Nature 410, 988994.
  • Kim H. J., Martemyanov K. A. and Thayer S. A. (2008) Human immunodeficiency virus protein Tat induces synapse loss via a reversible process that is distinct from cell death. J. Neurosci. 28, 1260412613.
  • Korobova F. and Svitkina T. (2010) Molecular architecture of synaptic actin cytoskeleton in hippocampal neurons reveals a mechanism of dendritic spine morphogenesis. Mol. Biol. Cell 21, 165176.
  • Kramar E. A., Chen L. Y., Brandon N. J., Rex C. S., Liu F., Gall C. M. and Lynch G. (2009) Cytoskeletal changes underlie estrogen's acute effects on synaptic transmission and plasticity. J. Neurosci. 29, 1298212993.
  • Kupfer R., Swanson L., Chow S., Staub R. E., Zhang Y. L., Cohen I. and Christians U. (2008) Oxidative in vitro metabolism of liquiritigenin, a bioactive compound isolated from the Chinese herbal selective estrogen beta-receptor agonist MF101. Drug Metab. Dispos. 36, 22612269.
  • Lau P. M., Zucker R. S. and Bentley D. (1999) Induction of filopodia by direct local elevation of intracellular calcium ion concentration. J. Cell Biol. 145, 12651275.
  • Lephart E. D., Setchell K. D. and Lund T. D. (2005) Phytoestrogens: hormonal action and brain plasticity. Brain Res. Bull. 65, 193198.
  • Letendre S. (2011) Central nervous system complications in HIV disease: HIV-associated neurocognitive disorder. Top. Antivir. Med. 19, 137142.
  • Li W., Li G., Steiner J. and Nath A. (2009) Role of Tat protein in HIV neuropathogenesis. Neurotox. Res. 16, 205220.
  • Lin Z., Gu J., Xiu J., Mi T., Dong J. and Tiwari J. K. (2012) Traditional chinese medicine for senile dementia. Evid. Based Complement. Alternat. Med. 2012, 692621.
  • Lindl K. A., Marks D. R., Kolson D. L. and Jordan-Sciutto K. L. (2010) HIV-associated neurocognitive disorder: pathogenesis and therapeutic opportunities. J. Neuroimmune Pharmacol. 5, 294309.
  • Liu F., Day M., Muniz L. C. et al. (2008) Activation of estrogen receptor-beta regulates hippocampal synaptic plasticity and improves memory. Nat. Neurosci. 11, 334343.
  • Liu R. T., Zou L. B. and Lu Q. J. (2009) Liquiritigenin inhibits Abeta(25-35)-induced neurotoxicity and secretion of Abeta(1-40) in rat hippocampal neurons. Acta Pharmacol. Sin. 30, 899906.
  • Liu R. T., Zou L. B., Fu J. Y. and Lu Q. J. (2010) Effects of liquiritigenin treatment on the learning and memory deficits induced by amyloid beta-peptide (25-35) in rats. Behav. Brain Res. 210, 2431.
  • Liu R. T., Tang J. T., Zou L. B., Fu J. Y. and Lu Q. J. (2011) Liquiritigenin attenuates the learning and memory deficits in an amyloid protein precursor transgenic mouse model and the underlying mechanisms. Eur. J. Pharmacol. 669, 7683.
  • Mancuso J. J., Chen Y., Li X., Xue Z. and Wong S. T. (2012) Methods of dendritic spine detection: from Golgi to high-resolution optical imaging. Neuroscience. doi:10.1016/j.neuroscience.2012.04.010. [Epub ahead of print].
  • Maragos W. F., Tillman P., Jones M., Bruce-Keller A. J., Roth S., Bell J. E. and Nath A. (2003) Neuronal injury in hippocampus with human immunodeficiency virus transactivating protein, Tat. Neuroscience 117, 4353.
  • Masliah E., Heaton R. K., Marcotte T. D. et al. (1997) Dendritic injury is a pathological substrate for human immunodeficiency virus-related cognitive disorders. HNRC Group. The HIV Neurobehavioral Research Center. Ann. Neurol. 42, 963972.
  • Matus A., Ackermann M., Pehling G., Byers H. R. and Fujiwara K. (1982) High actin concentrations in brain dendritic spines and postsynaptic densities. Proc. Natl Acad. Sci. USA 79, 75907594.
  • Mersereau J. E., Levy N., Staub R. E. et al. (2008) Liquiritigenin is a plant-derived highly selective estrogen receptor beta agonist. Mol. Cell. Endocrinol. 283, 4957.
  • Mortensen A., Kulling S. E., Schwartz H. et al. (2009) Analytical and compositional aspects of isoflavones in food and their biological effects. Mol. Nutr. Food Res. 53(Suppl 2), S266S309.
  • Murphy D. D. and Segal M. (1996) Regulation of dendritic spine density in cultured rat hippocampal neurons by steroid hormones. J. Neurosci. 16, 40594068.
  • O'Neill K. O., Chen S. and Brinton R. D. (2004) Impact of the selective estrogen receptor modulator, tamoxifen, on neuronal outgrowth and survival following toxic insults associated with aging and Alzheimer's disease. Exp. Neurol. 188, 268278.
  • Paruthiyil S., Cvoro A., Zhao X. et al. (2009) Drug and cell type-specific regulation of genes with different classes of estrogen receptor beta-selective agonists. PLoS ONE 4, e6271.
  • Phan A., Lancaster K. E., Armstrong J. N., MacLusky N. J. and Choleris E. (2011) Rapid effects of estrogen receptor alpha and beta selective agonists on learning and dendritic spines in female mice. Endocrinology 152, 14921502.
  • Pugliese A., Vidotto V., Beltramo T., Petrini S. and Torre D. (2005) A review of HIV-1 Tat protein biological effects. Cell Biochem. Funct. 23, 223227.
  • Sa M. J., Madeira M. D., Ruela C., Volk B., Mota-Miranda A. and Paula-Barbosa M. M. (2004) Dendritic changes in the hippocampal formation of AIDS patients: a quantitative Golgi study. Acta Neuropathol. 107, 97110.
  • Sanchez A. M., Flamini M. I., Fu X. D., Mannella P., Giretti M. S., Goglia L., Genazzani A. R. and Simoncini T. (2009) Rapid signaling of estrogen to WAVE1 and moesin controls neuronal spine formation via the actin cytoskeleton. Mol. Endocrinol. 23, 11931202.
  • Segal M. and Murphy D. (2001) Estradiol induces formation of dendritic spines in hippocampal neurons: functional correlates. Horm. Behav. 40, 156159.
  • Sekino Y., Kojima N. and Shirao T. (2007) Role of actin cytoskeleton in dendritic spine morphogenesis. Neurochem. Int. 51, 92104.
  • Shin A. H., Kim H. J. and Thayer S. A. (2012) Subtype selective NMDA receptor antagonists induce recovery of synapses lost following exposure to HIV-1 Tat. Br. J. Pharmacol. 166, 10021017.
  • Srivastava D. P., Woolfrey K. M., Liu F., Brandon N. J. and Penzes P. (2010) Estrogen receptor ss activity modulates synaptic signaling and structure. J. Neurosci. 30, 1345413460.
  • Woolley C. S. and McEwen B. S. (1992) Estradiol mediates fluctuation in hippocampal synapse density during the estrous cycle in the adult rat. J. Neurosci. 12, 25492554.
  • Wu Y. (2004) HIV-1 gene expression: lessons from provirus and non-integrated DNA. Retrovirology 1, 13. doi:10.1186/1742-4690-1-13.
  • Zhao L., Chen Q. and Diaz B. R. (2002) Neuroprotective and neurotrophic efficacy of phytoestrogens in cultured hippocampal neurons. Exp. Biol. Med. (Maywood) 227, 509519.
  • Zhao L., Mao Z. and Brinton R. D. (2009) A select combination of clinically relevant phytoestrogens enhances estrogen receptor beta-binding selectivity and neuroprotective activities in vitro and in vivo. Endocrinology 150, 770783.