Presenilins regulate synaptic plasticity and mitochondrial calcium homeostasis in the hippocampal mossy fiber pathway.

BACKGROUND: Presenilins play a major role in the pathogenesis of Alzheimer's disease, in which the hippocampus is particularly vulnerable. Previous studies of Presenilin function in the synapse, however, focused exclusively on the hippocampal Schaffer collateral (SC) pathway. Whether Presenilins play similar or distinct roles in other hippocampal synapses is unknown. METHODS: To investigate the role of Presenilins at mossy fiber (MF) synapses we performed field and whole-cell electrophysiological recordings and Ca2+ imaging using acute hippocampal slices of postnatal forebrain-restricted Presenilin conditional double knockout (PS cDKO) and control mice at 2 months of age. We also performed quantitative electron microscopy (EM) analysis to determine whether mitochondrial content is affected at presynaptic MF boutons of PS cDKO mice. We further conducted behavioral analysis to assess spatial learning and memory of PS cDKO and control mice at 2 months in the Morris water maze. RESULTS: We found that long-term potentiation and short-term plasticity, such as paired-pulse and frequency facilitation, are impaired at MF synapses of PS cDKO mice. Moreover, post-tetanic potentiation (PTP), another form of short-term plasticity, is also impaired at MF synapses of PS cDKO mice. Furthermore, blockade of mitochondrial Ca2+ efflux mimics and occludes the PTP deficits at MF synapses of PS cDKO mice, suggesting that mitochondrial Ca2+ homeostasis is impaired in the absence of PS. Quantitative EM analysis showed normal number and area of mitochondria at presynaptic MF boutons of PS cDKO mice, indicating unchanged mitochondrial content. Ca2+ imaging of dentate gyrus granule neurons further revealed that cytosolic Ca2+ increases induced by tetanic stimulation are reduced in PS cDKO granule neurons in acute hippocampal slices, and that inhibition of mitochondrial Ca2+ release during high frequency stimulation mimics and occludes the Ca2+ defects observed in PS cDKO neurons. Consistent with synaptic plasticity impairment observed at MF and SC synapses in acute PS cDKO hippocampal slices, PS cDKO mice exhibit profound spatial learning and memory deficits in the Morris water maze. CONCLUSIONS: Our findings demonstrate the importance of PS in the regulation of synaptic plasticity and mitochondrial Ca2+ homeostasis in the hippocampal MF pathway.

Delivery of a monomeric p53 subdomain with mitochondrial targeting signals from pro-apoptotic Bak or Bax.

PURPOSE: p53 targeted to the mitochondria is the fastest and most direct pathway for executing p53 death signaling. The purpose of this work was to determine if mitochondrial targeting signals (MTSs) from pro-apoptotic Bak and Bax are capable of targeting p53 to the mitochondria and inducing rapid apoptosis. METHODS: p53 and its DNA-binding domain (DBD) were fused to MTSs from Bak (p53-BakMTS, DBD-BakMTS) or Bax (p53-BaxMTS, DBD-BaxMTS). Mitochondrial localization was tested via fluorescence microscopy in 1471.1 cells, and apoptosis was detected via 7-AAD in breast (T47D), non-small cell lung (H1373), ovarian (SKOV-3) and cervical (HeLa) cancer cells. To determine that apoptosis is via the intrinsic apoptotic pathway, TMRE and caspase-9 assays were conducted. Finally, the involvement of p53/Bak specific pathway was tested. RESULTS: MTSs from Bak and Bax are capable of targeting p53 to the mitochondria, and p53-BakMTS and p53-BaxMTS cause apoptosis through the intrinsic apoptotic pathway. Additionally, p53-BakMTS, DBD-BakMTS, p53-BaxMTS and DBD-BaxMTS caused apoptosis in T47D, H1373, SKOV-3 and HeLa cells. The apoptotic mechanism of p53-BakMTS and DBD-BakMTS was Bak dependent. CONCLUSION: Our data demonstrates that p53-BakMTS (or BaxMTS) and DBD-BakMTS (or BaxMTS) cause apoptosis at the mitochondria and can be used as a potential gene therapeutic in cancer.

A chimeric p53 evades mutant p53 transdominant inhibition in cancer cells.

Because of the dominant negative effect of mutant p53, there has been limited success with wild-type (wt) p53 cancer gene therapy. Therefore, an alternative oligomerization domain for p53 was investigated to enhance the utility of p53 for gene therapy. The tetramerization domain of p53 was substituted with the coiled-coil (CC) domain from Bcr (breakpoint cluster region). Our p53 variant (p53-CC) maintains proper nuclear localization in breast cancer cells detected via fluorescence microscopy and shows a similar expression profile of p53 target genes as wt-p53. Additionally, similar tumor suppressor activities of p53-CC and wt-p53 were detected by terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL), annexin-V, 7-aminoactinomycin D (7-AAD), and colony-forming assays. Furthermore, p53-CC was found to cause apoptosis in four different cancer cell lines, regardless of endogenous p53 status. Interestingly, the transcriptional activity of p53-CC was higher than wt-p53 in 3 different reporter gene assays. We hypothesized that the higher transcriptional activity of p53-CC over wt-p53 was due to the sequestration of wt-p53 by endogenous mutant p53 found in cancer cells. Co-immunoprecipitation revealed that wt-p53 does indeed interact with endogenous mutant p53 via its tetramerization domain, while p53-CC escapes this interaction. Therefore, we investigated the impact of the presence of a transdominant mutant p53 on tumor suppressor activities of wt-p53 and p53-CC. Overexpression of a potent mutant p53 along with wt-p53 or p53-CC revealed that, unlike wt-p53, p53-CC retains the same level of tumor suppressor activity. Finally, viral transduction of wt-p53 and p53-CC into a breast cancer cell line that harbors a tumor derived transdominant mutant p53 validated that p53-CC indeed evades sequestration and consequent transdominant inhibition by endogenous mutant p53.

The DNA binding domain of p53 is sufficient to trigger a potent apoptotic response at the mitochondria.

The tumor suppressor p53 is one of the most studied proteins in human cancer.1-3 While nuclear p53 has been utilized for cancer gene therapy, mitochondrial targeting of p53 has not been fully exploited to date.4,5 In response to cellular stress, p53 translocates to the mitochondria and directly interacts with Bcl-2 family proteins including antiapoptotic Bcl-XL and Bcl-2 and proapoptotic Bak and Bax.6 Antiapoptotic Bcl-XL forms inhibitory complexes with proapoptotic Bak and Bax preventing their homo-oligomerization.7 Upon translocation to the mitochondria, p53 binds to Bcl-XL, releases Bak and Bax from the inhibitory complex and enhances their homo-oligomerization.8 Bak and Bax homotetramer formation disrupts the mitochondrial outer membrane, releases antiapoptotic factors such as cytochrome c and triggers a rapid apoptotic response mediated by caspase induction.9 It is still unclear if the MDM2 binding domain (MBD), the proline-rich domain (PRD) and/or DNA binding domain (DBD) of p53 are the domains responsible for interaction with Bcl-XL.10-17 The purpose of this work is to determine if a smaller functional domain of p53 is capable of inducing apoptosis similarly to full length p53. To explore this question, different domains of p53 (MBD, PRD, DBD) were fused to the mitochondrial targeting signal (MTS) from Bcl-XL to ensure Bcl-XL specific targeting.18 The designed constructs were tested for apoptotic activity (TUNEL, Annexin-V, and 7-AAD) in 3 different breast cancer cell lines (T47D, MCF-7, MDA-MB-231), in a cervical cancer cell line (HeLa) and in non-small cell lung adenocarcinoma cells H1373. Our results indicate that DBD-XL (p53 DBD fused to the Bcl-XL MTS) reproduces (in T47D cells) or demonstrates increased apoptotic activity (in MCF-7, MDA-MB-231, and HeLa cells) compared to p53-XL (full length p53 fused to Bcl-XL MTS). Additionally, mitochondrial dependent apoptosis assays (TMRE, caspase-9), co-IP and overexpression of Bcl-XL in T47D cells suggest that DBD fused to XL MTS may bind to and inhibit Bcl-XL. Taken together, our data demonstrates for the first time that the DBD of p53 may be the minimally necessary domain for achieving apoptosis at the mitochondria in multiple cell lines. This work highlights the role of small functional domains of p53 as a novel cancer biologic therapy.

A single mutant, A276S of p53, turns the switch to apoptosis.

The tumor suppressor protein p53 induces apoptosis, cell cycle arrest, and DNA repair along with other functions in a transcription-dependent manner [Vousden, K. H. Cell 2000, 103(5), 691-694]. The selection of these functions depends on sequence-specific recognition of p53 to a target decameric sequence of gene promoters [Kitayner, M.; et al. Mol. Cell 2006, 22(6), 741-753]. Amino acid residues in p53 that directly bind to DNA were analyzed, and the replacement of A276 in p53 with selected amino acids elucidated its importance in promoter transcription. For most apoptotic and cell cycle gene promoters, position 9 of the target decameric sequence is a cytosine, while for DNA repair gene promoters, thymine is found instead. Therefore, selective binding to the cytosine at the ninth position may transcribe apoptotic gene promoters and thus can induce apoptosis and cell cycle arrest. Molecular modeling with PyMOL indicated that substitution of a hydrophilic residue, A276S, would prefer binding to cytosine at the ninth position of the target decameric sequence, whereas substitution of a hydrophobic residue (A276F) would fail to do so. Correspondingly, A276S demonstrated higher transcription of PUMA, PERP, and p21(WAF1/CIP1)gene promoters containing a cytosine at the ninth position and lower transcription of GADD45 gene promoter containing a thymine at the ninth position compared to wild-type p53. Cell cycle analysis showed that A276S maintained similar G1/G0 phase arrest as wild-type p53. Additionally, A276S induced higher apoptosis than wild-type p53 as measured by DNA segmentation and 7-AAD assay. Since the status of endogenous p53 can influence the activity of the exogenous p53, we examined the activity of A276S in HeLa cells (wild-type endogenous p53) in addition to T47D cells (mutated and mislocalized endogenous p53). The same apoptotic trend in both cell lines suggested A276S can induce cell death regardless of endogenous p53 status. Cell proliferation assay depicted that A276S efficiently reduced the viability of T47D cells more than wild-type p53 over time. We conclude that the predicted preferred binding of A276S to cytosine at the ninth position better transactivates a number of apoptotic gene promoters. Higher induction apoptosis than wild-type p53 makes A276S an attractive candidate for therapy to eradicate cancer.

Controlled access of p53 to the nucleus regulates its proteasomal degradation by MDM2.

The tumor suppressor p53 can be sent to the proteasome for degradation by placing its nucleo-cytoplasmic shuttling under ligand control. Endogenous p53 is ubiquitinated by MDM2 in the nucleus, and controlling the access of p53 to the nuclear compartment regulates its ubiquitination and proteasomal degradation. This was accomplished by the use of a protein switch that places nuclear translocation under the control of externally applied dexamethasone. Fluorescence microscopy revealed that sending protein switch p53 (PS-p53) to the nucleus produces a distinct punctate distribution in both the cytoplasm and nucleus. The nuclear role in accessing the proteasome was investigated by inhibiting classical nuclear export with leptomycin B. Trapping PS-p53 in the nucleus only allows this punctate staining in that compartment, suggesting that PS-p53 must translocate first to the nuclear compartment for cytoplasmic punctate staining to occur. The role of MDM2 binding was explored by inhibiting MDM2/p53 binding with nutlin-3. Inhibition of this interaction blocked both nuclear export and cytoplasmic and nuclear punctate staining, providing evidence that any change in localization after nuclear translocation is due to MDM2 binding. Further, blocking the proteolytic activity of the proteasome maintained the nuclear localization of the construct. Truncations of p53 were made to determine smaller constructs still capable of interacting with MDM2, and their subcellular localization and degradation potential was observed. PS-p53 and a smaller construct containing the two MDM2 binding regions of p53 (Box I + V) were indeed degraded by the proteasome as measured by loss of enhanced green fluorescent protein that was also fused to the construct. The influence of these constructs on p53 gene transactivation function was assessed and revealed that PS-p53 decreased gene transactivation, while PS-p53(Box I + V) did not significantly change baseline gene transactivation.

Solid phase synthesis of mitochondrial triphenylphosphonium-vitamin E metabolite using a lysine linker for reversal of oxidative stress.

Mitochondrial targeting of antioxidants has been an area of interest due to the mitochondria's role in producing and metabolizing reactive oxygen species. Antioxidants, especially vitamin E (α-tocopherol), have been conjugated to lipophilic cations to increase their mitochondrial targeting. Synthetic vitamin E analogues have also been produced as an alternative to α-tocopherol. In this paper, we investigated the mitochondrial targeting of a vitamin E metabolite, 2,5,7,8-tetramethyl-2-(2'-carboxyethyl)-6-hydroxychroman (α-CEHC), which is similar in structure to vitamin E analogues. We report a fast and efficient method to conjugate the water-soluble metabolite, α-CEHC, to triphenylphosphonium cation via a lysine linker using solid phase synthesis. The efficacy of the final product (MitoCEHC) to lower oxidative stress was tested in bovine aortic endothelial cells. In addition the ability of MitoCEHC to target the mitochondria was examined in type 2 diabetes db/db mice. The results showed mitochondrial accumulation in vivo and oxidative stress decrease in vitro.

Enhanced and selective killing of chronic myelogenous leukemia cells with an engineered BCR-ABL binding protein and imatinib.

The oncoprotein Bcr-Abl stimulates prosurvival pathways and suppresses apoptosis from its exclusively cytoplasmic locale, but when targeted to the mitochondrial compartment of leukemia cells, Bcr-Abl was potently cytotoxic. Therefore, we designed a protein construct to act as a mitochondrial chaperone to move Bcr-Abl to the mitochondria. The chaperone (i.e., the 43.6 kDa intracellular cryptic escort (iCE)) contains an EGFP tag and two previously characterized motifs: (1) an optimized Bcr-Abl binding motif that interacts with the coiled-coil domain of Bcr (ccmut3; 72 residues), and (2) a cryptic mitochondrial targeting signal (cMTS; 51 residues) that selectively targets the mitochondria in oxidatively stressed cells (i.e., Bcr-Abl positive leukemic cells) via phosphorylation at a key residue (T193) by protein kinase C. While the iCE colocalized with Bcr-Abl, it did not relocalize to the mitochondria. However, the iCE was selectively toxic to Bcr-Abl positive K562 cells as compared to Bcr-Abl negative Cos-7 fibroblasts and 1471.1 murine breast cancer cells. The toxicity of the iCE to leukemic cells was equivalent to 10 µM imatinib at 48 h and the iCE combined with imatinib potentiated cell death beyond imatinib or the iCE alone. Substitution of either the ccmut3 or the cMTS with another Bcr-Abl binding domain (derived from Ras/Rab interaction protein 1 (RIN1; 295 residues)) or MTS (i.e., the canonical IMS derived from Smac/Diablo; 49 residues) did not match the cytotoxicity of the iCE. Additionally, a phosphorylation null mutant of the iCE also abolished the killing effect. The mitochondrial toxicity of Bcr-Abl and the iCE in Bcr-Abl positive K562 leukemia cells was confirmed by flow cytometric analysis of 7-AAD, TUNEL, and annexin-V staining. DNA segmentation and cell viability were assessed by microscopy. Subcellular localization of constructs was determined using confocal microscopy (including statistical colocalization analysis). Overall, the iCE was highly active against K562 leukemia cells and the killing effect was dependent upon both the ccmut3 and functional cMTS domains.

Utilizing the estrogen receptor ligand-binding domain for controlled protein translocation to the insoluble fraction.

PURPOSE: The estrogen receptor forms insoluble aggregates in the insoluble cytoskeletal subcellular fraction when bound to the antagonist fulvestrant. The ligand-binding domain was isolated and fused to signal sequences to target subcellular compartments. Sequestering a pro-apoptotic peptide tested the utility of a protein targeted to the insoluble fraction. METHODS: The ligand-binding domain of the estrogen receptor was isolated and fused with signal sequences, either a nuclear localization signal or nuclear export signal. The subcellular localization when bound to drug fulvestrant was examined, specifically its interaction with cytokeratins 8 and 18. The ability to target a therapeutic peptide to the insoluble fraction was tested by fusing a therapeutic coiled-coil from Bcr-Abl in K562 cells. RESULTS: The estrogen receptor ligand-binding domain responds to fulvestrant by translocating to the insoluble fraction. Adding a signal sequence significantly limited the translocation to either the nucleus or cytoplasm. The cytokeratin 8/18 status of the cell did not alter this response. The therapeutic coiled-coil fused to ERLBD was inactivated upon ligand induction. CONCLUSIONS: Isolating the ligand-binding domain of the estrogen receptor creates a ligand-controllable protein capable of translocation to the insoluble fraction. This can be used to sequester an active peptide to alter its function.

Direct induction of apoptosis using an optimal mitochondrially targeted p53.

Targeting the tumor suppressor p53 to the mitochondria triggers a rapid apoptotic response as efficiently as transcription-dependent p53. (1, 2) p53 forms a complex with the antiapoptotic Bcl-XL, which leads to Bak and Bax oligomerization resulting in apoptosis via mitochondrial outer membrane permeabilization. (3, 4) Although p53 performs its main role in the mitochondrial outer membrane, it also interacts with different proteins in the mitochondrial inner membrane and matrix. (5, 6) To further investigate mitochondrial activity of p53, EGFP-p53 was fused to different mitochondrial targeting signals (MTSs) directing it to the mitochondrial outer membrane ("XL-MTS" from Bcl-XL; "TOM-MTS" from TOM20), the inner membrane ("CCO-MTS" from cytochrome c oxidase), or matrix ("OTC-MTS" from ornithine transcarbamylase). Fluorescence microscopy and a p53 reporter dual luciferase assay demonstrated that fusing MTSs to p53 increased mitochondrial localization and nuclear exclusion depending on which MTS was used. To examine if the MTSs initiate mitochondrial damage, we fused each individual MTS to EGFP (a nontoxic protein) as negative controls. We performed caspase-9, TUNEL, annexin-V, and 7-AAD apoptosis assays on T47D breast cancer cells transfected with mitochondrial constructs. Except for EGFP-XL, apoptotic potential was observed in all MTS-EGFP-p53 and MTS-EGFP constructs. In addition, EGFP-p53-XL showed the greatest significant increase in programmed cell death compared to its nontoxic MTS control (EGFP-XL). The apoptotic mechanism for each construct was further investigated using pifithrin-α (an inhibitor of p53 transcriptional activity), pifithrin-µ (a small molecule that reduces binding of p53 to Bcl-2 and Bcl-XL), and overexpressing the antiapoptotic Bcl-XL. Unlike the MTSs from TOM, CCO, and OTC, which showed different apoptotic mechanisms, we conclude that p53 fused to the MTS from Bcl-XL performs its apoptotic potential exclusively through the p53/Bcl-XL specific pathway.

Controlling subcellular delivery to optimize therapeutic effect.

This article focuses on drug targeting to specific cellular organelles for therapeutic purposes. Drugs can be delivered to all major organelles of the cell (cytosol, endosome/lysosome, nucleus, nucleolus, mitochondria, endoplasmic reticulum, Golgi apparatus, peroxisomes and proteasomes) where they exert specific effects in those particular subcellular compartments. Delivery can be achieved by chemical (e.g., polymeric) or biological (e.g., signal sequences) means. Unidirectional targeting to individual organelles has proven to be immensely successful for drug therapy. Newer technologies that accommodate multiple signals (e.g., protein switch and virus-like delivery systems) mimic nature and allow for a more sophisticated approach to drug delivery. Harnessing different methods of targeting multiple organelles in a cell will lead to better drug delivery and improvements in disease therapy.

Reversal of oxidative stress in endothelial cells by controlled release of adiponectin.

Hyperglycemia causes endothelial dysfunction due to its effect on increasing reactive oxygen species (ROS). Adiponectin (Adp) has been reported to suppress hyperglycemia-associated ROS generation. It was hypothesized that administering globular adiponectin (gAdp) via injectable biodegradable thermosensitive triblock copolymer might effectively reduce ROS generation in endothelial cells. In this study, gAdp was incorporated into and released from the polymer gel. The released gAdp was further investigated by comparing it with the intact gAdp with regard to the efficiency in reducing ROS and activating cAMP. The released gAdp effectively suppressed excess ROS production in the in vitro endothelial cell culture model under high-glucose condition via cAMP/PKA pathway. These data provide a rationale for developing controlled release dosage form of gAdp as a therapeutic tool for oxidative stress-related pathology in patients with diabetes.