Presenilin-2 modulation of ER-mitochondria interactions

Abstract

Presenilin (PS) mutations are the main cause of Familial Alzheimer’s Disease (FAD), and have been demonstrated to cause an imbalance of intracellular Ca²⁺ homeostasis. Though PS1 and 2 are generally considered to behave similarly in terms of their effects on Ca²⁺ handling, we have recently described a novel function, which is unique to PS2, i.e. the modulation of ER-mitochondria juxtaposition. Accordingly, PS2, but not PS1, affects the Ca²⁺ cross-talk between these organelles, a key feature in determining cell fate. In particular, PS2 over-expression, and more drastically that of FAD-linked PS2 mutants, strongly increases the interaction between ER and mitochondria, thus facilitating mitochondrial Ca²⁺ uptake. The likely mechanisms behind this phenomenon and its potential effects in cell physiology and pathology are discussed.

This article refers to:

Presenilins 1 and 2 (PS1, PS2) are the catalytic core of γ-secretase, the enzymatic complex responsible for the final step in amyloid-beta (Aβ) peptides' production and mutations in their genes are the main cause of Familial Alzheimer's Disease (FAD).1,2

Presenilin is a nine trans-membrane domain protein that, once incorporated into the γ-secretase complex, rapidly undergoes endoproteolysis at its large cytosolic loop, generating two fragments (N- and C-terminal) which are considered the mature form of PS. Both PSs are mainly found at the level of Endoplasmic Reticulum (ER) membranes.1–3 Of interest, recent findings indicate that PSs are enriched in ER membrane patches closely associated to mitochondria,4 the so called “Mitochondria-Associated Membranes (MAMs).”5

Apart from their function in Aβ pep-tides' production, PSs appear involved in several physiological processes within the cell, including Ca²⁺ homeostasis, protein transport and turnover, autophagy, cell adhesion, neurotransmitter release and axon guidance.6–11

The highly homologue PS1 and PS2 are commonly thought to have almost overlapping cellular functions, with minor differences, e.g., PS1 is considered the more abundant and the main protein involved in Aβ peptides' generation.2,12 However, we have recently described a novel cell function of PS2 not shared by PS1, i.e., the modulation of ER and mitochondria interactions and thus, of their Ca²⁺ crosstalk.13

ER and mitochondria are two major players in cellular Ca²⁺ homeostasis and their interaction is critical in cell physiology. Mitochondria are, in fact, known to massively take up Ca²⁺ when cytosolic microdomains of high Ca²⁺ concentrations (Ca²⁺ hot spots) are generated at the level of their outer membrane (OMM).14 This type of events occur upon opening of either plasma membrane voltage-gated Ca²⁺-channels or ER Ca²⁺ releasing channels.15,16 Mitochondrial functions are in turn tightly regulated by the intramitochondrial Ca²⁺ concentration that is known to modulate few key dehydrogenases and, when the increases are too large or prolonged, to trigger apoptosis.17,18 Thus, Ca²⁺ cross-talk between ER and mitochondria contributes to cell fate determination and alterations in the two organelles' axis can dramatically affect cell state.19

The number of proteins supposed to influence ER-mitochondria interactions is progressively increasing, among them the best studied include PACS-2,20 Mfn-2,21 IP3R-grp75-VDAC22 and Fis1-Bap3123 complexes, and Trichoplein/mitostatin.24 PS2 appears now as a novel member of this group:13 in fact, by PS2 overexpression or downregulation, we have demonstrated that this protein (but not its homologue PS1) modulates the ER-mitochondria tethering by increasing the number and/or the extent of their contact sites. In particular, PS2 overexpression facilitates mitochondrial Ca²⁺ uptake following ER Ca²⁺ release, while the opposite occurs upon its downregulation. PS2 overexpression is thus associated to a higher probability of cytosolic Ca²⁺ hot spots generation at the level of the OMM. Strikingly, expression of FAD-linked PS2 mutations leads to a larger increase in ER-mitochondria interactions, compared to the wt counterpart, as demonstrated both in cell lines and in primary cortical neurons, a finding that may be of major importance for interpreting the role of Ca²⁺ dysregulation in AD (Fig. 1).

As far as the molecular mechanism behind this PS2 effect is concerned, available data are still insufficient to draw a clear conclusion and only hypotheses can be made:

(1) Mfn-2,21 the protein that directly bridges, by homotypic interaction, ER and mitochondria, seems not to be directly involved since PS2 overexpression or downregulation does not modulate its protein level.13

(2) PS2 could be required for the proper docking of a protein—one of its substrates—that in turn acts as a tether between the two organelles. It could be speculated that under normal conditions, wt PS2, as the catalytic core of the γ-secretase complex, cleaves most of this substrate, while FAD-PS2 mutants, which are known to have a reduced enzymatic activity,25 leave un-cleaved a larger fraction of this tether, thus increasing the vicinity of the two organelles. However, the finding that pharmacological γ-secretase inhibition does not mimic this PS2 effect13 argues against a major role of substrate cleavage in the increased interaction between ER and mitochondria.

(3) Since mitochondrial dynamics within the cell is regulated by Ca²⁺,26 the increased ER-mitochondria interaction could be also due to an alteration of mitochondria movements due to the imbalance of ER Ca²⁺ handling caused by FAD-PS2 expression (reviewed in refs. 6, 7, 27 and 28). However, different treatments that alter ER Ca²⁺ concentration ([Ca²⁺]_ER) do not mimic the PS2 effect on organelles' juxtaposition.13

(4) The full length form of PS2 is likely involved in this phenomenon since a PS2 mutant devoid of endo-proteolytic maturation (PS2-D366A) is fully capable of increasing ER-mitochondria interactions.13

(5) Based on quantitative considerations,13 PS2 likely requires a partner protein to exert its effect in bridging ER and mitochondria; the nature of this partner is currently not clear, as it could be a molecule residing on ER and/or mitochondria membranes as well as a cytosolic component.

(6) Albeit highly homologous, the two PSs do not share this function and interestingly, the region with the highest variability between PS1 and PS2 is the big cytosolic loop, which is disrupted upon maturation;1 this suggests that the PS2 cytosolic loop could be the region of the protein involved in the interaction with the hypothetical molecular partner and thus required for modulating ER-mitochondria juxtaposition.

Concerning the importance and significance of this effect in cell physiology and pathology, several considerations are allowed (Fig. 2). PS2 decreases [Ca²⁺]_ER more potently than PS1,27,28 yet an additional effect on intracellular Ca²⁺ handling, i.e., the facilitation of ER-mitochondria Ca²⁺ transfer, was described only for PS2.13 A decrease in [Ca²⁺]_ER is generally considered protective against apoptosis since reduced [Ca²⁺]_ER would decrease the probability of mitochondria Ca²⁺ overload;17,19,29 on this line, the lower [Ca²⁺]_ER due to expression of FAD-linked PS2 mutants, compared to those of PS1, could explain the milder AD phenotype described in patients bearing PS2 rather than PS1 mutations, although both groups show amyloid pathology.1,2 On the other side, reduced [Ca²⁺]_ER could impair the physiological Ca²⁺ cross-talk between ER and mitochondria,30 that is required for the proper activity of the latter organelle.

This novel effect of PS2 on ER-mitochondria juxtaposition adds further complexity to the scenario of Ca²⁺ dysfunctions due to FAD-linked PS mutations: on one hand, the increased ER-mitochondria interactions caused by PS2 mutants could represent a compensatory phenomenon, ensuring proper Ca²⁺ signals towards mitochondria, despite the decrease in [Ca²⁺]_ER (Fig. 2B); on the other hand, this effect could overcome the reduction in [Ca²⁺]_ER and increase the probability of toxic Ca²⁺ transfer to mitochondria, with detrimental consequences on cell viability (Fig. 2C). Again, the observed milder AD phenotypes linked to FAD-PS2 mutations could argue against this second possibility that considers PS2 mutations as more noxious than those in PS1. Interestingly, a much larger number of mutations in PS1 than in PS2 have been so far identified in FAD patients;1 an odd difference for two such homologous proteins, suggesting that some PS2 mutations might be incompatible with life.

These different hypotheses can be now tested experimentally but nonetheless they highlight the complexity of Ca²⁺ dysfunction in neurodegeneration and the multifaceted nature of PSs.

Figures and Tables

Figure 1 ER-mitochondria juxtaposition and high Ca²⁺ microdomains on the cytosolic surface of OMM in SH-SY5Y cells. Cells were transfected with cDNAs coding for a FAD-PS2 mutant (PS2-T122R, B) or void vector (Control, A), together with those coding for a red mitochondrial and a green ER markers (mit-RFP and ER-GFP, respectively; upper cell in each part), in order to visualize ER-mitochondria juxtaposition sites (yellow pixels); increased number of interactions is observed in FAD-PS2 expressing cells. Alternatively, cells were transfected with cDNAs coding for a FAD-PS2 mutant (PS2-T122R, B) or void vector (Control, A) and the cameleon Ca²⁺ probe N33-D1cpv, targeted to the OMM,15 and then analyzed to detect the generation of high Ca²⁺ microdomains on the cytosolic surface of OMM (yellow-to-red spikes) upon ER Ca²⁺ release induced by bradykinin (lower cell in each part). At a similar average cytosolic Ca²⁺ rise, higher and more abundant Ca²⁺ microdomains are generated close to the OMM in the FAD-PS2 expressing cell compared to the control one. See reference 13 for details.

Figure 2 Model depicting the possible effects of FAD-PS2 mutations on ER-mitochondria interactions and Ca²⁺ cross-talk. In physiological condition (A), mitochondria receive constant Ca²⁺ signals from ER that ensure their proper activity.30 Expression of FAD-PS2 mutants (B and C) decreases [Ca²⁺]_ER27,28 and recruits mitochondria closer to ER Ca²⁺ releasing sites.13 As a consequence, closer ER-mitochondria juxtaposition may compensate [Ca²⁺]_ER reduction allowing mitochondria to receive appropriate Ca²⁺ signals that drive AT P production (B); alternatively, closer ER-mitochondria juxtaposition may expose mitochondria to excessive Ca²⁺ stimulation, triggering the apoptotic cascade29 by mitochondria Ca²⁺ overload, outer membrane permeabilization and release of pro-apoptotic factors (C).

Addendum to: