Universidad Autónoma de Madrid
Departamento de Biología Molecular
Regulation of mitochondrial
respiration in astrocytes: role of Ca 2+ , ATP demand and pyruvate production
Inés Juaristi Santos
Madrid, 2018
Departamento de Biología Molecular
Facultad de Ciencias
UNIVERSIDAD AUTÓNOMA DE MADRID
Regulation of mitochondrial respiration in astrocytes:
role of Ca
2+, ATP demand and pyruvate production
Memoria presentada por la Licenciada en Bioquímica Inés Juaristi Santos para optar al título de Doctor en Ciencias en la modalidad de formato clásico bajo la
supervisión de:
Directora de tesis:
Dra. Jorgina Satrústegui Gil-Delgado
Co-directora de tesis:
Dra. Araceli del Arco Martínez
Este trabajo ha sido realizado en el Departamento de Biología Molecular, Centro de Biología Molecular “Severo Ochoa” (C.S.I.C. – U.A.M.). La realización de esta Tesis ha sido
posible gracias a Ayudas para el Programa Predoctoral de Formación de Personal Investigador No Doctor del Gobierno Vasco.
3
Agradecimientos
Si la ciencia es buscar un gato negro en una habitación oscura sin estar segura de que haya gato, una tesis es rozar un bigote en esa oscuridad sin estar segura de que sea de gato, de ratón o de tigre.
En primer lugar, me gustaría expresar mi agradecimiento a Ina y a Araceli, sin las que, sin duda, esta tesis no existiría. Gracias por acogerme en vuestro laborato- rio todos estos años. Desde aquella entrevista en segundo de carrera no he parado de aprender y de disfrutar haciendo ciencia.
Ina, gracias por tener el despacho siempre abierto, por alegrarte con cada resul- tado positivo y buscar alternativas a los negativos. Gracias también por brindarme todas las oportunidades posibles. Gracias, por ser un ejemplo. Araceli, infinitas gra- cias por tus consejos, tus apoyos, tus manos y tus críticas. Por enseñarme que una banda borrosa puede dar mucha felicidad y que cuando ni siquiera sale borrosa solo hay que animarse y repetirlo. Gracias por decirme que todo no se puede, pero ponérmelo muy fácil para conseguirlo. Me siento muy afortunada por haber tenido unas directoras como vosotras, creo que no hay muchas.
También quiero dar las gracias al resto del 321, si cuando veo el CBM pienso en casa, es por vosotras. Paloma, por escucharme siempre, por venir corriendo a ayudarme, por decir que sí a casi cualquier plan que te proponga, por aguantar mi angustia y celebrar mi alegría. IreneR, por enmarronarte más que yo en cada cosa que empezamos, eres la compañera más digna. Gracias al resto del laboratorio, a Bea, Isabel y Bárbara y suerte a los recién llegados, Ana, Luis.
A las que ya no estáis en el laboratorio, pero estaréis en el cartel de la entrada para siempre. A Irene Llorente, porque mucha de esta tesis empezó cuando no sabíamos qué hacer con los astrocitos que tardaban cuatro semanas en crecer, sin ti esta tesis tampoco estaría. A Nacho, por enseñarme a sacar mitocondrias y medir a tope. A Carlos, porque recuerdo todos tus consejos.
Estos años no hubiesen sido tan fáciles sin mis amigas y compañeras de infi- nitas tardes por las calles de Madrid. A Lucas, porque me he hecho mayor a tu lado, porque hemos recorrido la ciudad juntas, con bata o sin ella, porque estos 10 años serían muy diferentes sin tu compañía. A Ángel, porque nos aguantas durante horas hablando de células y por abrazarme tanto. A Cris, por tu complicidad, tus
"todo va ir súper bien", tus "claro que sí", por todos nuestros momentos de hablar
5
solo de nosotras. A Sergio, por las discusiones acaloradas. A Montero, por tus uñas rojas. Gracias Arantxa, por ser la marida perfecta y tus mimos en Duque de Alba.
Mayte y Amaia, más embodied y embeded en esta tesis no podríais estar. Por ser fans de la mitocondria, porque tengo claro que, aunque vivamos en ciudades distintas, siempre habrá un fin de semana radom para encontrarnos, ya sea en Paris, Berlín, La Habana o, al menos, en Donosti.
Finalmente dar las gracias a mi familia. A mis tíos, Eloy y Elisa, porque vuestra casa es un refugio y un abrigo. Porque las fideuás de mi tío Eloy y los besos de mi tía Elisa curan cualquier pena o enfermedad. A mi abuelo, que siempre quiso tener una nieta científica y porque en lo único que tenía fe era en la ciencia y en la tecnología. A mis padres, Aita, Ama, eskerrik asko emandako laguntza guztiagatik, nigan beti izandako konfidantzagatik eta Madrilera ekartzeagatik. Ama, gracias por acompañarme todos estos años, aunque fuera desde Donosti. Nunca he hecho sola el camino a la renfe, no importa a qué hora. Almudena, elkarrekin egindako bidaien eta barreengatik.
Finalmente, gracias Manuel, por este viaje que hemos empezado juntas, porque me apoyas en todo, porque a tu lado todo es más fácil, viajar, leer, reír. Porque te emocionas y disfrutas sin límites.
6
Index
Figure index 11
Abreviations 15
Abstract 19
Resumen 23
1. Estudio de la falta de ARALAR en la producción del lactato en el cerebro 25
2. Regulación de la respiración mitocondrial en los astrocitos . . . 26
3. Estudio de la proteiína ARALAR sin capacidad de unión a Ca2+ . . . 28
Introduction 31 1. Brain Metabolism . . . 33
1.1. Metabolic differences between astrocytes and neurons . . . 34
2. Regulation of coupled respiration . . . 37
3. Extramitochondrial Ca2+ vs mitochondrial Ca2+ regulating mitochon- drial respiration . . . 38
Mitochondrial Ca2+ regulating mitochondrial respiration . . . 38
Ca2+ regulation from the external face of the inner mitochon- drial membrane . . . 39
4. Ca2+ regulation of mitochondrial respiration in neurons . . . 42
5. Astrocytes response to energy demand . . . 43
5.1. Calcium signals in astrocytes . . . 44
5.2. ATP . . . 45
5.3. Glutamate . . . 45
Glutamate transporters . . . 46
Glutamate receptors . . . 47 7
Glutamate metabolism . . . 47
Objectives 51 Materials and Methods 55 1. Animals . . . 57
1.1. Genotyping of mice . . . 58
2. Cell culture . . . 59
2.1. Primary neuronal and astrocyte cultures . . . 59
2.2. Mouse embryonic fibroblast primary culture . . . 60
3. Cell transfections . . . 60
4. Western Blot . . . 61
5. Experiments with astrocytes . . . 61
5.1. Measurements of glucose consumption and lactate production . 61 5.2. Determinations of Cytosolic glycerol 3-phosphate dehydroge- nase and mitochondrial glycerol phosphate dehydrogenase ac- tivity . . . 62
5.3. Cytosolic Na+, Ca2+ and pH imaging in primary astrocytes . . . . 63
5.4. Determinations of oxygen consumption . . . 64
5.5. Measurements of mitochondrial Ca2+, cytosolic pyruvate levels and ATP/ADP ratio “in vivo” using fluorescent protein sensors . . 65
5.6. Production of shRNA-encoding rAAVs for MCU silencing . . . 66
5.7. MCU knockdown by siRNA . . . 67
5.8. Mitochondrial calcium measurement with luminescence . . . 67
6. Experiments with mouse embryonic fibroblasts (MEFs) . . . 68
6.1. MEFs immortalization . . . 68
6.2. RNA isolation and RT-PCR analysis . . . 68
6.3. Mitochondria isolation from cultured mouse embryonic fibroblast 69 6.4. Proteasome inhibition treatment in MEFs . . . 69
6.5. Protease inhibition treatment in MEFs and freshly isolated mito- chondria . . . 70
6.6. Analysis of membrane protein complexes by Blue Native PAGE . 70 7. Statistical analysis . . . 70
8. Key resources table . . . 71 8
Results 75 1. ARALAR deficiency does not produce a primary increase in brain lactate 77
1.1. ARALAR-KO cultured astrocytes do not produce more lactate than WT astrocytes . . . 77 1.2. Glycerol phosphate shuttle may be the main NADH-shuttle in
astrocytes . . . 80 2. Regulation of mitochondrial respiration in astrocytes . . . 81 2.1. Regulation of mitochondrial respiration by extracellular ATP . . . 81 ATP-induced Ca2+ transients increase astrocyte respiration . . . 81 1 mM ATP induces a robust stimulation of astrocyte respiration . 84 Role of mitochondrial Ca2+ regulation systems and Ca2+ reg-
ulation of pyruvate supply in the acute stimulation of astrocyte respiration by ATP . . . 86 2.2. Regulation of mitochondrial respiration by glutamate in astrocyte 91
Glutamate stimulates mitochondrial respiration in astrocytes in a [Ca2+]o independent manner . . . 92 Glutamate acting through metabotropic receptors does not in-
crease respiration in cultured astrocytes . . . 95 Glutamate-stimulated respiration in astrocytes meets energy
demands of increased [Na+]I . . . 95 The increase in respiration induced by glutamate is not due to
glutamate consumption by astrocytes . . . 96 Glutamate-induced upregulation of glycolytic pyruvate produc-
tion stimulates astrocytic respiration . . . 98 Glutamate-induced upregulation of respiration occurs in the pres-
ence of elevated extracellular K+ . . . 100 3. Study of ARALAR Knock-In mice . . . 101
3.1. Inhibition of proteasome does not increase 4mut-ARALAR pro- tein levels . . . 104 3.2. Involvement of mitochondrial AAA-metaloproteases in mutant
ARALAR protein degradation . . . 105 3.3. TIM22 complex, CII and CIII are upregulated in ARALAR KI MEFs 106
9
3.4. Up-regultation of Tim22 protein does not increase mutated AR- ALAR levels in MEFs . . . 108
Discussion 111
1. ARALAR deficiency does not produce an increase in brain lactate levels 113 1.1. NADH shuttles in astrocytes . . . 114 2. Regulation of mitochondrial respiration in astrocytes . . . 114 2.1. Extracellular ATP implication in mitochondrial regulation . . . 115 2.2. Glutamate increases mitochondrial respiration in acute manner 119 3. Lack of 4mutARALAR protein levels in the brain and MEFs . . . 122 3.1. Upregulation of mitochondrial proteins in 4mut-ARALAR MEFS . 123
Conclusions 127
Conclusiones 131
References 135
Appendix 157
10
Figure index
INTRODUCTION FIGURES
Figure 1. Ca2+ modulation of mitochondrial respiration.
Figure 2. Schematic diagram of Ca2+-dependent mitochondrial carrier (CaMC) secondary structure.
Figure 3. Malate-aspartate NADH shuttle (MAS): role of the AGC in Ca2+ activa- tion of MAS activity.
Figure 4. Ca2+ regulation of mitochondrial respiration.
Figure 5. Generation ARALAR with disrupted calcium binding domains.
Figure 6. Scheme of glutamatergic signal that may activate mitochondrial func- tion.
Figure 7. Scheme of glutamate’s pathway inside the astrocytes.
MATERIALS AND METHODS FIGURES
Figure 8. ARALAR-KI mice.
Figure 9. Genotyping of ARALAR-KI mice.
Figure 10. Scheme of fundamental values of mitochondrial respiration.
RESULTS FIGURES
Figure 11. Glucose consumption and lactate production rates and oxidative metabolism in neurons and astrocytes.
Figure 12. Effect of 100 μM ATP on [Ca2+]I, [Ca2+]m, [Na+]I, cytosolic ATP/ADP ratio and mitochondrial respiration in astrocytes.
Figure 13. Effect of 1 mM ATP on [Ca2+]I, [Ca2+]m, [Na+]I, cytosolic ATP/ADP ratio and mitochondrial respiration in astrocytes in the presence or absence of external calcium.
Figure 14. OCR response to extracellular ATP in wild type and SCaMC-3-KO, ARALAR- KO mouse astrocytes.
Figure 15. Effects of MCU silencing using rAAV on [Ca2+]m entry in astrocytes.
Figure 16. Effects of MCU silencing siRNA-mediated on [Ca2+]m entry in astro- cytes.
Figure 17. Cytosolic pyruvate levels in response to extracellular ATP.
Figure 18. Effect of 200 μM glutamate on [Na+]I, [Ca2+]I, [Ca2+]m, and mito- chondrial respiration in astrocytes in the presence or absence of external calcium.
13
Figure 19. Activation of metabotropic glutamate receptor with DHPG (50 μM) does not stimulate mitochondrial respiration.
Figure 20. Effect of 200 μM glutamate on [Na+]I, [Ca2+]I and mitochondrial respiration in astrocytes in the presence or absence of external sodium.
Figure 21. Effect of glutamate stimulation on cytosolic pyruvate and cytosolic ATP/ADP ratio.
Figure 22. Effects of high K+ and glutamate on mitochondrial respiration.
Figure 23. Analysis of ARALAR protein levels in ARALAR KI tissues and mice growth in C56BL/6 background.
Figure 24. Analysis of ARALAR protein levels in ARALAR KI tissues and mice growth in CD1 background.
Figure 25. Analysis of 4mut-ARALAR mRNA and protein levels in ARALAR KI MEFs from CD1 background.
Figure 26. Inhibition of proteasome with MG132 in WT and ARALAR-KI MEFs.
Figure 27. Inhibition of metalloproteases with 1,10-Phenanthroline (P9375) in wild-type (WT) and ARALAR-KI (KI) MEFs.
Figure 28. Analysis of TIM22 complex protein levels and changes of mitochon- drial respiratory complexes.
Figure 29. Overexpression of TIMM22 protein in wild-type (WT) and ARALAR-KI (KI) MEFs.
DISCUSSION FIGURES
Figure 30. Ca2+ regulation of glycolysis.
Figure 31. Regulation of mitochondrial respiration in astrocytes.
14
Abreviations
ABREVIATIONS
AGC: Aspartate glutamate carrier AGK: Acylglycerol kinase.
Ant: Antimycin.
ANT: ATP/ADP translocase.
ADP: Adenosine diphosphate.
Asp: L-aspartate.
ATP: Adenosine triophosphate.
BN-PAGE: Blue Native Polyacrylamide Gel Electrophoresis.
[Ca2+]I: cytosolic Ca2+.
[Ca2+]mit: mitochondrial Ca2+. [Ca2+]o: outer Ca2+.
CNS: Central nervous system.
CNQX: 6 - cyano - 7 - nitroquinoxaline - 2,3 - dione.
D-asp: D-aspartate.
DHPG: (S)-3,5-dihydroxyphenylglycine.
DIV: Days in vitro.
DNP: 2,4-dinitrophenol.
EMRE: Essential MCU regulatory ele- ment.
ER: Endoplasmic reticulum.
Glc: Glucose.
Glu: L-glutamate.
GLUT: Glucose transporter.
GPCR: G-protein coupled receptor.
GPS: Glycerol 3-phosphate shuttle.
cGPD: cytosolic glycerol 3-phosphate dehydrogenase.
mGPD: mitochondrial glycerol 3- phos- phate dehydrogenase.
HBS: Hanks-buffered solution.
IMM: Inner mitochondrial membrane.
IP3: Inositol 1,4,5-triphosphate.
IP3R: Inositol triphosphate receptor.
MEF: Mouse embryonic fibroblast.
MCUC: Mitochondrial calcium uniporter complex.
MCU: Mitochondrial calcium uniporter.
mGluR: Metabotropic glutamate recep- tor.
MK-801: (5S,10R) - (+) - 5- Methyl - 10,11 - dihydro - 5H - dibenzo[a,d]cyclohepten - 5,10 - imine hydrogen maleate.
[Na+]I: cytosolic Na+. [Na+]o: outer Na+.
NCLX: Na+/Ca2+/Li2+ exchanger.
NCX: Na+/Ca2+ exchanger.
O/N: over night.
OCR: Oxygen consumption rate.
Olig: Oligomycin.
OMM: Outer mitochondrial membrane.
OXPHOS: Oxidative phosphorylation.
P2Y: Purinergic metabotropic receptor.
P2X: Purinergic ionotropic recepor.
PND: Postnatal days.
PDH: Pyruvate dehydrogenase com- plex.
PDK: Pyruvate dehydrogenase kinase.
PDP: pyruvate dehydrogenase phos- phatase.
PMCA: Plasma membrane Ca2+- ATPase.
Pyr: Pyruvate.
17
rAAV: Recombinant adeno-associated virus.
Rot: Rotenone.
RyR: Ryanodine receptor.
SCaMC: Short calcium binding carrier.
shRNA: Encodign small hairpin RNA.
siRNA: Small interfering RNA.
SOCE: Store-operated calcium entry.
TIM: Translocase Inner Membrane.
TOM: Translocase Outer Membrane.
Abstract
ARALAR, Ca2+-regulated aspartate-glutamate mitochondrial carrier, is an important component of the NADH malate aspartate shuttle (MAS). ARALAR-deficiency is a rare dis- ease causing a severe phenotype (OMIM ID #612949). The animal model, ARALAR-KO mouse, recapitulates the major findings in humans. The aim of the work was understand- ing the impact of ARALAR-deficiency in brain lactate levels; finding that lactate production upon mitochondrial blockade depends on upregulation of lactate formation in astrocytes rather than in neurons. However, ARALAR-deficiency decreased cell respiration in neurons, not astrocytes, which maintained unchanged respiration and lactate production. As the primary site of ARALAR-deficiency is neuronal, this explains the lack of accumulation of brain lactate in ARALAR-deficiency in humans and mice.
Astrocytes are central in regulation of synaptic signaling and neurotransmitter recy- cling, which are energy consuming processes. The aim of the work was to address the role of respiration and calcium regulation of respiration as part of the astrocyte response to these workloads, particularly to those caused by extracellular ATP and glutamate, two neurotransmitters produced by neurons and astrocytes. Extracellular ATP (100 μM-1 mM) caused a Ca2+-dependent workload and fell of the cytosolic ATP/ADP ratio, which acutely increased astrocytes respiration. Part of this increase is related to a Ca2+- dependent upregulation of cytosolic pyruvate production. Conversely, L-glutamate (200 μM) caused a Na+, but not Ca2+, dependent workload even though glutamate-induced Ca2+ signals readily reached mitochondria. This workload is due to Na+-dependent glutamate transport activity and triggers a rapid fall in the cytosolic ATP/ADP ratio and stimulation of respiration.
D-aspartate produced similar effects. Glutamate-induced increase in respiration is linked to a rapid increase in glycolytic pyruvate production, exceeding that caused by extracellu- lar ATP. The results suggest that both signaling molecules cause an increase in astrocyte respiration fueled by workload-induced increase in glycolysis and pyruvate production. As stimulation of respiration by ATP and glutamate are similar and pyruvate production was smaller than in the first case, the results suggest that the metabolic response to extracellu- lar ATP is a Ca2+-dependent upregulation of respiration added to upregulation of glycolysis.
The global contribution of these astrocyte respiratory responses to brain oxygen consump- tion is an open question.
A mouse model with mutations in Ca2+ binding domains of ARALAR, but with intact transporter domain (4mut-ARALAR) has been studied. Surprisingly, 4mut- ARALAR protein levels were mostly absent in brain mitochondria whereas mRNA levels were the same as WT. An analysis of the possible causes of 4mutAralar protein degradation is presented.
Resumen
1. Estudio de la falta de ARALAR en la producción del lactato en el cerebro
ARALAR (AGC1/Slc25a12) es el transportador de aspartato-glutamato de la mem- brana interna mitocondrial mayoritario en cerebro. ARALAR presenta dominios de unión a Ca2+ del tipo manos EF orientados hacia el espacio intermembrana que me- dian su activación por señales de Ca2+ citosólicas. Mutaciones en ARALAR (OMIM ID #612949) originan una enfermedad monogenética y recesiva caracterizada por un fenotipo muy severo de hipomielinización, epilepsia y retraso en el neurode- sarrollo (Falk et al., 2014; Wibom et al., 2009). El modelo murino deficiente en ARALAR, ARALAR Knock Out (ARALAR-KO) recapitula la mayoría de las caracterís- ticas fenotípicas (Jalil et al., 2005). En los pacientes y en los ratones ARALAR-KO se detectan la mayoría de las características asociadas a las enfermedades mito- condriales (Finsterer, 2008; Haas et al., 2007), y sin embargo, no se ha detectado un aumento en los niveles de lactato en el cerebro. En este trabajo nuestro primer objetivo fue estudiar la causa de la falta del aumento de lactato en el cerebro, es- tudiando la producción de lactato en situación basal y bajo inhibición mitocondrial tanto en neuronas como en astrocitos en cultivo. Además, medimos la captura de glucosa y la producción de lactato de las neuronas y astrocitos, obtenidas de ratones silvestres (WT) o ARALAR-KO y también el consumo de oxígeno mitocon- drial. En este estudio concluimos que aunque las neuronas ARALAR-KO mostraban un descenso del 40% en la respiración mitocondrial (Llorente- Folch et al., 2013), consumen y producen la misma cantidad de glucosa y lactato que las neuronas WT.
Para los astrocitos, WT y ARALAR-KO también determinamos que consumen y pro- ducen la misma cantidad de glucosa y lactato mostrando además valores similares de respiración basal. Los resultados obtenidos indican que, aunque la proteína AR- ALAR está presente en los astrocitos en cultivo no es necesaria para el correcto funcionamiento mitocondrial, como sí ocurre en neuronas. Además, cuando se in- hibe la respiración en astrocitos con rotenona (inhibidor del complejo respiratorio I) y antimicina A (inhibidor de complejo respiratorio III), se aumenta la producción de lactato tanto en los astrocitos WTs como en los ARALAR-KO. Las neuronas, sin embargo, en ningún caso son capaces de aumentar el flujo glucolítico. Nuestros datos indican que la falta de aumento de lactato en el cerebro de los pacientes
deficientes en ARALAR la es debida a que sólo se altera la función mitocondrial en las neuronas y no en astrocitos, siendo estos los únicos capaces de incrementar la producción de lactato una vez inhibida la mitocondria.
2. Regulación de la respiración mitocondrial en los astrocitos
En el cerebro se consume alrededor de un 20 % de la energía producida por el organismo (Mink et al., 1981). La mayoría de esa energía es requerida por la neu- rona post-sináptica (Alle et al., 2009; Attwell and Laughlin, 2001; Llorente-Folch et al., 2013), para regular los niveles de iones y recuperar las concentraciones basales. En nuestro laboratorio se ha descrito que en las neuronas el Ca2+ regula la respiración mitocondrial a través de ARALAR y su participación en la lanzadera malato-aspartato (MAS) (Llorente-Folch et al., 2013; Llorente-Folch et al., 2015;
Rueda et al., 2014), principalmente con estímulos que producen una pequeña o mediana demanda de ATP como el carbacol y el KCl. La activación a través de ARALAR-MAS representa una forma alternativa de regular la respiración mitocon- drial por el Ca2+ como molécula señalizadora diferente de la entrada de Ca2+ a la matriz mitocondrial. El Ca2+ entra a la mitocondria por el complejo del uniportador de calcio (MCUC), esto activa las deshidrogenasas del ciclo de Krebs aumentando así la función mitocondrial (Glancy and Balaban, 2012).
Las neuronas no están solas en el cerebro (Allen and Barres, 2009) y los astroc- itos están posicionados en muchos casos muy cerca de la hendidura sináptica y reciben muchas de las señales que llegan a la neurona post-sináptica. Los astroci- tos son células gliales que cumplen funciones claves en el cerebro como son: tam- ponar los iones H+ y K+, captar el glutamato liberado de las neuronas, administrar metabolitos a las neuronas o la gliotransmision. La respuesta de las mitocondrias a la demanda de trabajo requerida en estos procesos es poco conocida para los astrocitos.
En general, los astrocitos se han definido como células glicoliticas, sin embargo, desde hace tiempo se sabe que las mitocondrias de los astrocitos están posi- cionadas incluso en los procesos estrechos de los astrocitos cercanos a la sinapsis (Agarwal et al., 2017; Jackson et al., 2014; Jackson and Robinson, 2018; Stephen
26
et al., 2015). En este trabajo usamos el ATP extracelular y glutamato como estímu- los, ambos neurotransmisores son liberados tanto por las neuronas como por los astrocitos (Covelo and Araque, 2018; Parpura et al., 2017). Nos hemos centrado en la implicación del Ca2+ en la regulación de la respiración en células intactas tras la estimulación con estos agonistas.
Los astrocitos tienen receptores para ATP purinérgicos metabotropicos (P2Y) y receptores ionotrópicos (P2X) (Franke et. al; 2012). Los receptores P2Y van acoplados a proteínas G. Las proteínas Gqα producen a través de la activación de PLC la liberación de Ca2+ del retículo mediante los receptores IP3. Además, la activación de los receptores P2X puede permitir la entrada directa de Ca2+ a la célula. Se han utilizado dos concentraciones distintas de ATP (100 μM y 1 mM), para conseguir una mayor y una menor entrada de Ca2+ al citosol. Nuestros resultados indican que la estimulación de la respiración es dependiente de Ca2+ en el caso de 100 μM ATP, así como la bajada en la razón de ATP/ADP. Para 1 mM ATP, la estimulación de la respiración por ATP es igual en presencia y ausencia de Ca2+, sin embargo, en ausencia de Ca2+ se produce la entrada de Na+, incrementando la demanda de ATP, como sugiere que la razón ATP/ADP disminuya más que en presencia de Ca2+. Debido a eso concluimos que la estimulación con 1 mM ATP produce un aumento de la respiración que podría deberse a señalización por Ca2+, puesto que a mayor demanda de ATP (situación sin Ca2+) se observan niveles similares de estimulación que en presencia de Ca2+.
Hemos descartado que el incremento en la regulación de la respiración por Ca2+
fuese debido a los transportadores dependientes de Ca2+: ARALAR y SCaMC-3. No obstante, intentamos disminuir la entrada del Ca2+ a la mitocondria, silenciando la expresión del gen que codifica para el canal del MCUC (MCU), sin embargo, a pesar de lograr reducir los niveles de la proteína, no fue suficiente para regular la captura de Ca2+ por la mitocondria. Por tanto no podemos descartar que MCU juegue un papel en la regulación de la respiración en astrocitos. Medimos también la producción de piruvato citosólica a tiempo real (San Martin et al., 2014) determi- nando que la producción de piruvato era dependiente de entrada de Ca2+ externo en el caso de la estimulación con 1 mM ATP. Así pues, podemos concluir que el ATP extracelular regula la respiración en astrocitos principalmente a través de la demanda de ATP producida por la entrada de Ca2+ y la regulación del Ca2+ a la
glucolisis aumentando el suministro de piruvato a la mitocondria.
El glutamato tiene muchas dianas en el astrocito, pero principalmente es intro- ducido a la célula por los transportadores dependientes de Na+. Los transporta- dores mayoritarios son GLAST/Slc1a3 y GLT-1/Slc1a2 (Robinson M.B. and Jackson J.G., 2016). En este trabajo hemos comprobado que los astrocitos aumentan la respiración tras ser estimulados con glutamato. Este aumento de la respiración es dependiente de Na+ externo, pero no del Ca2+. Utilizando D-aspartato, un agonista de los transportadores de glutamato no metabolizable, también se produce un au- mento de la respiración. Tanto la estimulación con glutamato como con D-aspartato aumentan de manera similar la producción de piruvato y causa una bajada en la razón ATP/ADP. Por lo que concluimos que el glutamato aumenta la respiración principalmente por la regulación de la demanda de ATP producida por una entrada de Na+. En paralelo se produce un aumento de la producción de piruvato, may- oritariamente glucolitico, dirigido también por la demanda de ATP. En resumen, el glutamato regula la respiración tanto por el cambio de la razón de ATP/ADP como por el aumento del suministro de piruvato a la mitocondria.
3. Estudio de la proteiína ARALAR sin capacidad de unión a Ca
2+Como último objetivo nos propusimos estudiar la función de la activación por Ca2+ de ARALAR, sin interrumpir la función transportadora basal (Contreras and Satrústegui, 2009). Para ello se generó una línea de ratón con mutaciones en los dominios de unión a Ca2+. Los dominios EF1 y EF2 son necesarios para la unión del Ca2+ a ARALAR (del Arco and Satrústegui, 2004), aunque estudios estructurales posteriores han descrito que la unión directa del Ca2+ ocurre exclusivamente en EF2 (Thangaratnarajah et al., 2014). La línea de ratón producida tiene cuatro susti- tuciones en aminoácidos pertenecientes a los dominios de manos EF1 y EF2 (pro- teína 4mut-ARALAR), que impiden la activación por Ca2+ de ARALAR (Marmol et al., 2009). Los ratones que expresan esta proteína 4mut- ARALAR en homocigosis (ARALAR-KI) tienen una menor vida media y un peso corporal disminuido respecto a los ratones con la proteína silvestre. Además, aunque en estos ratones los niveles de mRNA no están afectados, la proteína 4mut-ARALAR apenas es detectable en el
28
cerebro y se detecta a niveles más bajos en el corazón y en el músculo esquelético.
Para poder profundizar en las causas de la ausencia de la proteína 4mut-ARALAR en el cerebro usamos como modelo celular fibroblastos embrionarios de ratón (MEF). Los MEFs, a su vez, tampoco presentan niveles detectables de la proteína, pero tienen los mismos niveles de expresión del gen 4mut-Aralar que los MEFs sil- vestres para el gen WT. Debido a que las mitocondrias tienen distintos mecanismos homeostáticos para regular los problemas con proteínas mal formadas (Baker et al., 2011), tratamos de inhibir esos procesos para aumentar los niveles de 4mut- ARALAR. Uno de los mecanismos es el proteasoma (Bragoszewski et al., 2013).
Inhibimos el protesoma, sin embargo, esta aproximación no consiguió incrementar la presencia de la proteína, a pesar de aumentar los niveles generales de ubiqui- tinación de los MEFs. Otro mecanismo muy importante de la membrana interna mitocondrial relacionada con la homeostasis proteica son las metaloproteasas de- pendientes de ATP (proteasas-AAA) (Baker et al., 2011; Wasilewski et al., 2017).
Usando un quelante de zinc y hierro (1,10 fenantroleno) tratamos a las células y las mitocondrias aisladas de MEFs silvestres y ARALAR-KI. En ningún caso pudimos recobrar los niveles de 4mut-ARALAR. A pesar de ello, seguimos sin poder descar- tar la implicación de las proteasas-AAA, ya que carecemos de un control positivo referente a que el quelante funcionase. Hemos determinado que se requiere pro- fundizar más para conocer la causa de la falta de la proteína 4mut-ARALAR en el cerebro y en los MEFs.
Introduction
1. Brain Metabolism
Brain consumes about 20 % of the energy produced in the body in humans un- der basal conditions, and about 7-8 % in mice and rats (Mink et al., 1981). The main fuel in the brain is glucose (Sokoloff, 1981), but under conditions as fasting ketone bodies may support brain energy. In basal conditions glucose is oxidized near completely with a ratio of glucose consumption to oxygen consumption close to 1: 6. The complete oxidation of one molecule glucose produces 30-34 molecules of ATP (Nicholls, 2013) and requires 6 molecules of O2. However, it has been cal- culated that near 10 % of the consumed glucose is released from the brain in form of lactate, the origin of which is mentioned below. Interestingly, this percentage of glucose consumption released in form of lactate changes over the different parts of the brain (Vaishnavi et al., 2010) and during development (Brekke et al., 2015).
The discovery Fox and co-workers (Fox et al., 1988) that physiological stimula- tion significantly increased both cerebral blood flow and cerebral metabolic rate of glucose consumption (CMRglucose), while cerebral metabolic rate of oxygen con- sumption (CMRO2) was minimally increased. This may suggest that a fast increase in glycolysis would be enough to maintain this neuronal activity, producing lactate as final product and 2 molecules of ATP from 1 molecule of glucose. This possibility is controversial and has prompted a number of studies which have focused on a possible role of lactate not as waste product but as a metabolite with a dual role, an end product of glycolysis in one cell type (astrocytes) and a fuel for oxidative phosphorylation (OXPHOS) on the other (neurons) (Barros, 2013; Magistretti and Allaman, 2018). This role requires a continuous flux of metabolites among these cell types. Alternatively, the existence of a direct consumption of lactate by a mito- chondrial lactate dehydrogenase, avoiding the requirement of transcellular lactate fluxes has been also proposed (Schurr, 2018).
Calculations by different groups (Attwell and Laughlin, 2001; Llorente-Folch et al., 2013) described that the brain requires a large energy budget. Synaptic activi- ties as ion pumping and reestablishing ion gradients are very expensive processes in terms of ATP consumption. Alle and co-workers described (Alle et al., 2009) that 80 % of required energy in the brain is consumed by the post-synaptic neuron.
It is of great interest to know the mechanisms whereby brain workloads are cou-
pled to energy supply and the differences between neurons and glia cells (mainly astrocytes) in response to neurotransmitter stimuli.
1.1. Metabolic differences between astrocytes and neurons
In the brain, neurons are not alone, glial cells, and particularly astrocytes are their neighbors and they are positioned near neuronal elements (Allen and Barres, 2009). Astrocyte functions are several: supplying metabolites to neurons, buffering extracellular H+- and K+- ion concentrations, clearing up glutamate released from neurons or other neurotransmitter to restore and maintain ambient conditions fol- lowing neurotransmitter exocytosis, and more recently, gliotransmission. Because of the enumerated functions, astrocytes are part of the so called tripartite synapse (Araque et al., 1999).
The differences in glycolytic capacity and oxidative phosphorylation in astro- cytes and neurons have been extensively studied and debated (Hertz et al., 2007;
Magistretti and Allaman, 2015; Yellen, 2018). However, there is a consensus in that neurons are more oxidative than astrocytes. Both cells types express the enzymes required for glycolysis and oxidative phosphorylation (Cahoy et al., 2008; Zhang et al., 2014), but based on the absence of immunolabeling with anti-cytochrome oxidase antibodies (Wong-Riley et al., 1989) and the small size of astrocyte mito- chondria, early studies reported a virtual absence of mitochondria in astrocytes.
Nowadays, it is clear that mitochondria are present even in fine processes of as- trocytes (Agarwal et al., 2017; Jackson and Robinson, 2018; Lovatt et al., 2007;
Oheim et al., 2018; Stephen et al., 2015). However, there are enzymatic differ- ences in these metabolic pathways that substantiate the assumption of a higher glycolytic capacity of astrocytes.
One interesting difference pointed out by different groups (Almeida et al., 2001;
Herrero-Mendez et al., 2009; Porras et al., 2004) is that neurons, but not astrocytes, have a limited capacity of upregulating glycolysis. Neurons are unable to upregu- late 6-phosphofructo-1-kinase (PFK1) activity (Almeida et al., 2001; Almeida et al., 2004), one of the main regulators of glycolysis. In resting astrocytes, PFK1 spe- cific activity is around fourfold that found in neurons (Almeida et al., 2004). This is due to lower levels of PFK1 positive allosteric activator, fructose-2,6-bisphosphate (F2,6P2). The enzyme responsible for the synthesis of F2,6P2 is 6-phosphofructo-
34
2-kinase/fructose-2,6-bisphosphatase (PFKFB) a bifunctional enzyme, functioning as kinase and phosphatase. There are four isoforms and the most abundant one in astrocytes is isoform 3 (PFKFB3). PFKFB3 has a high kinase versus bisphosphatase activity ratio (Pilkis et al., 1995) favoring F2,6P2 synthesis. Silencing of PFKFB3 abolishes the ability of astrocytes to up-regulate glycolysis when mitochondria is inhibited with nitric oxide (NO) (Almeida et al., 2004). In neurons, however, there is a post-translational regulation of PFKFB3 by degradation. PFKFB3 undergoes con- tinuous degradation by the ubiquitin-proteasome pathway. PFKFB3 has a target se- quence for ubiquitination by the anaphase-promoting complex/cyclosome (APC/C)- Cdh1 (Herrero-Méndez et al., 2009). In neurons, the pathway of PFKFB3 degra- dation via APC/C Cdh1 is very active causing very low levels of PFKFB3 (Herrero- Méndez et al., 2009). Limited glycolysis has been proposed to drive glucose to the pentose phosphate pathway producing NADPH (Herrero-Méndez et al., 2009) required to maintain the antioxidant responses in a highly oxidative cell as the neuron (for review Bolaños (2016)).
Another enzymatic difference between cell types is pyruvate dehydrogenase (PDH) activity. PDH catalyses the conversion of pyruvate into Acetyl-CoA which enters in the Krebs Cycle (TCA cycle) that oxidizes completely pyruvate into CO2. Comparing the activity of pyruvate dehydrogenase complex (PDH) in astrocytes and neurons (Halim et al., 2010) it is found that that the ratio of inactive, phospho- rylated PDH to total PDH is higher in astrocytes than neurons, in correlation with a lower PDH activity in astrocytes. Astrocytes have higher levels of PDH kinase (PDK) 4 and 2, that maintain PDH inactive, but also higher levels of the Ca2+ reg- ulated PDH phosphatase (PDP1). As described further on, this may suggest that astrocytes have more ability to regulate PDH, even if in basal situation PDH is less active.
The other main enzymatic pathways that are key to regulate the fate of pyru- vate are the NADH shuttles. When glycolysis ends with pyruvate the oxidation of cytosolic NADH is needed in order to obtain NAD+ to maintain glycolysis. This can be performed by lactate dehydrogenase producing lactate from pyruvate. NADH can also be oxidized by mitochondria through cytosolic-mitochondrial NADH shut- tles. Inner mitochondria membrane is impermeable to NADH and to oxidize cytoso- lic NADH in mitochondria, the cells may use two main strategies: malate-aspartate
shuttle (MAS) (Cheeseman and Clark, 1988; Satrústegui et al., 2007) and glycerol phosphate shuttle (GPS) (McKenna et al., 2006). Our group described MAS as the key Ca2+ regulator of mitochondrial respiration in neurons (described in section 3).
However, the relevance of both NADH shuttles in astrocytes has been in disputed (Kohler et al., 2018; McKenna et al., 2006; Ramos et al., 2003). Actually, the role of these two shuttles in astrocytes is one the focus of the present study.
The facts outlined above support the notion that upon neuronal activity astro- cytes upregulate glycolysis and release lactate for neurons to use as oxidative substrate. Indeed, lactate can be used by neurons and becomes neuroprotective in a way dependent on MAS activity to oxidize cytosolic NADH produced by LDH (Llorente-Folch et al., 2016). This transcellular mechanism, the astrocyte-neuronal lactate shuttle (ANLS) (Magistretti and Allaman, 2015; Magistretti and Allaman, 2018; Pellerin and Magistretti, 1994) was first described to operate in response to glutamatergic stimulation (Pellerin and Magistretti, 1994). For ANLS to work there is need of lactate concentration gradient between astrocytes and neurons, which indeed has been found (Machler et al., 2016) and a higher NADH/NAD+ ratio in astrocytes (Mongeon et al., 2016), allowing lactate release from astrocytes and neuronal conversion into pyruvate for oxidative use. Following this scheme, neu- rons could couple energy supply in response to their activity through astrocytic activation of glycolysis (Magistretti and Allaman, 2015; Schurr, 2014). In ANLS monocarboxylate transporters (MCT) of plasma membrane play an important role.
Astrocytes have higher amount of MCT4 and MCT1 isoforms than neurons (Sharma et al., 2015), to be able to extrude lactate.
In contrast to these ideas, the relevance of lactate as an alternative energetic substrate has also been questioned. Dienel proposed lactate as a brain substrate when it is present in high concentrations, but indicated that most research evi- dence supports that glucose is the major energy source in normal activated brain (Dienel, 2012). Other studies showed that neurons are able to up-regulate glycol- ysis (Diaz-Garcia et al., 2017; Patel et al., 2014) for review Yellen (2018) and that activated neurons release rather than take up lactate (Díaz-García et al., 2017). Re- lated to those results and in contrast to ANLS, a neuron-astrocytes lactate shuttle NALS has been proposed, where neurons up regulate glycolysis capturing glucose and releasing lactate into the media (Mangia et al., 2009; Simpson et al., 2007).
36
During the past years little attention has been paid to mitochondrial activity in astrocytes (for review (Jackson and Robinson, 2018). As described above most studies focused on the glycolytic capacity of astrocytes. In the studies where res- piration was measured (Hertz and Hertz, 1979; Hertz and Hertz, 2003; Hertz et al., 2007) were done with a technology lacking reproducibility and adequate controls.
This work was aimed to assess the regulation of mitochondrial function under Ca2+
signals in astrocytes. One of the standards for mitochondrial function is respirome- try and this requires a homogeneous cell population to assign variations in oxygen consumption rate (OCR) to a given cell type. Therefore, this work was carried out in cultured astrocytes.
Figure 1: Ca2+ modulation of mitochondrial respiration. Ca2+-signals can increase the ATP demand, or regulate the supply of substrates to mitochondria. Once Ca2+ enters the matrix Ca2+
can regulate the dehydrogenases of the Krebs cycle and the ATPase. I, II, III and IV represents the respiratory complexes of the electron transport chain (ETC). Modified from (Glancy and Balaban, 2012; Llorente-Folch et al., 2013; Nicholls (2013).
2. Regulation of coupled respiration
The oxygen consumption rate of mitochondria is regulated to match ATP produc- tion through generation and consumption of the proton electrochemical gradient, as stated by the principles of the chemiosmotic theory (Mitchell, 1961). However, many cells types have a really strong mechanism to maintain ATP levels constant, therefore matching ATP production to ATP utilization, surprisingly without major changes in ADP and Pi (Balaban et al., 1986). Ca2+ regulation of mitochondrial ATP production has been proposed as a mechanism that would work to support OXPHOS
at a constant ADP/ATP ratio (Glancy and Balaban, 2012) (Figure 1).
Our laboratory has focused on the ways whereby Ca2+ may regulate mitochon- drial respiration (Fig. 1). The first is by directly increasing ATP demand rather than ATP production. This is because any increase in cytosolic Ca2+ would imply an in- crease in ATP demand to restore the basal concentrations values (Llorente-Folch et al., 2013). The other ways of regulating mitochondrial respiration are by direct ac- tion of Ca2+ as a signaling molecule Llorente-Folch et al., 2015; Rueda et al., 2014;
Rueda et al., 2015).
3. Extramitochondrial Ca
2+vs mitochondrial Ca
2+regulating mitochondrial respiration
Ca2+ has two different ways of regulating mitochondria: Ca2+ entering into the matrix through mitochondrial calcium uniporter complex (MCUC), or Ca2+ from the external side of the inner mitochondrial membrane.
Mitochondrial Ca2+ regulating mitochondrial respiration
Matrix Ca2+ is able to activate dehydrogenases of the Krebs cycle and the AT- Pase (Glancy and Balaban, 2012). The most studied mitochondrial Ca2+ uptake mechanism is MCUC, a channel transporting Ca2+ across the inner mitochondrial membrane driven by electrochemical gradient across. Functional and pharmaco- logical studies revealed that MCU is a Ca2+ selective channel (Kirichok et al., 2004):
the MCU complex or MCUC, which is composed by several subunits: MCU, MICU, EMRE (Mammucari et al., 2018). The channel itself is made by the mitochondrial calcium uniporter protein (MCU) (Baughman et al., 2011; De Stefani et al., 2011).
MCU alone is necessary and sufficient to transport Ca2+ across the membrane.
However, MCU Knock Out cells have a residual mitochondrial Ca2+ uptake capacity (Hamilton et al., 2018; Pan et al., 2013). Interestingly, MCU protein has two trans- membrane domains, but no Ca2+ sensing domain. MCUb is an isoform of MCU that shares a 50 % of homology with MCU, but is a negative regulator of MCU; overex- pression of MCUb decreases mitochondrial Ca2+ uptake capacity (Raffaello et al., 2013). The proteins that provide Ca2+ sensitivity are MICU proteins, with different isoforms. MICU1 and MICU2 being the most studied. Both interact with MCU and
38
help the opening of the channel upon Ca2+ binding (Kamer and Mootha, 2014).
The essential MCU regulator (EMRE) is a 10 kD inner membrane protein (Sancak et al., 2013). EMRE Knock Down or Knock Out totally abolishes mitochondrial Ca2+
uptake even when MCU is overexpressed.
Once in the matrix, Ca2+ can activate dehydrogenases of the Krebs cycle, pyru- vate dehydrogenase, α-ketoglutarate dehydrogenase and isocitrate dehydroge- nase (Denton and McCormack, 1986). The most studied one is pyruvate dehy- drogenase (PDH). PDH is regulated by phosphorylation, the active state is non- phosphorylated and the inactive, phosphorylated (P-PDH). Regulation by phos- phorylation happens in E1 catalytic subunit. There are two isoforms of the PDH phosphatases, PDP1 and PDP2, PDP1 is regulated by Ca2+ and is preferentially expressed in astrocytes, but PDP2, which is the major isoform in neurons, is not Ca2+-sensitive (Halim et al., 2010; Sharma et al., 2015). Pyruvate dehydroge- nases kinases (PDK) fall in four isoforms, and all are negatively regulated by ADP, NAD+ and pyruvate. All four isoforms are present in neurons and astrocytes, but PDK2 and PDK4 levels are higher in astrocytes in accordance with the fact that P- PDH/PDH ratio is higher in astrocytes than in neurons (Halim et al., 2010), having also a lower PDH activity (Halim et al., 2010).
Ca2+ regulation from the external face of the inner mitochondrial membrane
Mitochondrial solute carriers (MC) make up the superfamily of solute carriers number 25 (SLC25) (Palmieri, 2008; Palmieri and Monne, 2016), including the mi- tochondrial transporters of most metabolites, one relevant exception being the mitochondrial pyruvate carrier (MPC) which does not belong to SLC25 superfamily (Bricker et al., 2012; Herzig et al., 2012). There is a special group of MC, the Ca2+
dependent mitochondrial carriers (Del Arco et al., 2016; Satrústegui et al., 2007) which have N-terminal extensions with Ca2+-binding EF-hands facing the intermem- brane space. This group is composed by aspartate-glutamate carriers (AGCs) and ATP-Mg/Pi carriers (SCaMCs) (Fig. 2).
Figure 2: Schematic diagram of Ca2+-dependent mitochondrial carrier (CaMC) secondary structure. In both CaMCs, the carboxy-terminal half corresponds to the mitochondrial carrier (MC) homology region, and the amino-terminal extension harbors Ca2+-binding EF-hand motifs. The MC homology region, 300 amino acids long, is represented according to its homology with that of X-ray structure of ANT-1. TM1-TM6 corresponding to the six transmembrane helices characteristic of all MC, TM-1, -3 and -5 are kinked by the presence of conserved prolines. The amino-terminal extensions of aspartate-glutamate carriers (AGCs) are longer than those of short CaMCs (SCaMCs), and the overall distribution and sequence of EF-hand motifs are unrelated to that of SCaMCs, which is very similar to calmodulin. From Satrústegui et al. 2007.
ATP-Mg –Pi Carriers, SCaMCs (del Arco and Satrústegui, 2004; Fiermonte et al., 2004). These carriers transport ATP4−-Mg2+or ADPH2−in exchange of Pi2−, and they can work in both directions (Amigo et al., 2013; Rueda et al., 2015; Traba et al., 2012). As the only other adenylate carrier is the ADP/ATP translocase (ANT) which exchanges ATP with ADP, SCaMCs are the only carriers that can change the net content of adenine nucleotides in the mitochondria (Aprille, 1988; Aprille, 1993;
Traba et al., 2011). There are three main isoforms in mice and SCaMC-3 (Slc25a23) is the main isoform in the brain (Amigo et al., 2013; Rueda et al., 2015). Activation by Ca2+ of SCaMC-3, occurs with a Km of about 3-4 μM (Amigo et al., 2013), i.e, within the range of that of the calcium uniporter (Satrústegui et al., 2007).
Aspartate glutamate carriers (AGCs) extrude one molecule of aspartate in exchange of one molecule of glutamate with one H+, making this reaction elec- trogenic, and essentially unidirectional in polarized mitochondria (LaNoue and Tis-
40
chler, 1974). There are two isoforms ARALAR (AGC1/Slc25a12) (del Arco and Satrústegui, 1998; Palmieri et al., 2001) and CITRIN (AGC2/Slc25a13) (Del Arco et al., 2000;
Kobayashi et al., 1999). The main isoform in the brain is ARALAR (del Arco and Satrústegui, 1998). AGCs form a dimer (Thangaratnarajah et al., 2014) by the interaction of EF hand 4-8. del Arco and Satrústegui (2004) showed that Ca2+
binding was abolished by removal of EF hands 1 and 2, and structural analysis showed that AGCs only bind one Ca2+ by the canonical EF2 hand (Thangaratnara- jah et al., 2014). Ca2+ binds ARALAR in the brain with a Km 300 nM leading to an increase Vmax without changes in the affinity for glutamate (Contreras and Satrústegui, 2009), however ARALAR still functions in a Ca2+-free form. ARALAR is part of malate-aspartate shuttle (MAS, Figure 3). AGCs activity is responsible for a unidirectional shuttle in energized mitochondria.
Figure 3:Malate-aspartate NADH shuttle (MAS): role of the AGC in Ca2+ activation of MAS activity. MAS is made up of four enzymes: mitochondrial and cytosolic aspartate aminotransferases (AAT) (1), malate dehydrogenases (2), and two mitochondrial carriers, the aspartate/glutamate car- rier (AGC) and the α-ketoglutarate/malate carrier (OGC). The site for Ca2+activation of AGC is shown.
Mitochondrial matrix Ca2+, entering through the Ca2+ uniporter (CU), activates pyruvate dehydro- genase (PDH) (3), isocitrate dehydrogenase (IDH) (4), and α-ketoglutarate dehydrogenase (α-KGDH) (5). AcCoA, acetyl CoA; Asp, aspartate; Glut, glutamate; Isoc, isocitrate; αKG, α-ketoglutarate; Mal, malate; OAA, oxalacetic acid; Pyr, pyruvate; SuccCoA, succinil-CoA. From Satrústegui et al. 2007.
4. Ca
2+regulation of mitochondrial respiration in neurons
Our group has studied the roles of ARALAR-MAS, and SCaMC-3 in upregulating oligomycin-sensitive respiration in cerebral cortex neurons in response to work- loads produced by increases in Na+ and/or Ca2+ and robust-to-small Ca2+ signals (Llorente-Folch et al., 2013; Llorente-Folch et al., 2015; Rueda et al., 2014). Both CaMC types have Ca2+ regulation domains facing the intermembrane space (IMS), which has the same Ca2+ concentration as cytosol. These two carrier families re- spond to different ranges of cytosolic/IMS Ca2+. In neurons the response to large Ca2+ signals and workloads (such as due to veratridine administration) ends up in oxidative stress and activation of PARP1. There is a drastic decrease in mi- tochondrial ATP caused by PARP1 activation (Rueda et al., 2015). This decrease is prevented through Ca2+-activated SCaMC-3, which imports cytosolic ATP-Mg or ADP in mitochondria, preventing PTP opening and maintaining workload-stimulated respiration (Rueda et al., 2015).
ARALAR in neurons is involved in the response to small calcium signals and workloads (Llorente-Folch et al., 2013; Pardo et al., 2006; Satrústegui et al., 2007).
The ARALAR-MAS pathway appears to be the main mechanism of OCR stimulation by regulation of pyruvate supply to mitochondria. It plays an outstanding role in the response to smaller workloads, being the only Ca2+-regulation mechanism respon- sible for upregulation of respiration in response to the small Ca2+ signals produced by carbachol and in response to KCl depolarization, which induces large Ca2+ sig- nals in mitochondria. This is explained through the Ca2+-activation of ARALAR in the intermembrane space leading to supply pyruvate to mitochondria. ARALAR- MAS activation by small Ca2+ signals leads to fall in cytosolic NADH/NAD+ which allows to shift pyruvate away from lactate and into mitochondria (Fig. 4).
In contrast to these functions of Ca2+in activating neuronal oxidative metabolism several surveys (Contreras and Satrústegui, 2009; Bak et al., 2006 and Satrústegui and Bak 2015) have proposed a mechanism whereby high Ca2+ signals as achieved during neuronal stimulation lead to an impaired MAS function which may block ox- idative glucose consumption. The mechanism is based on a matrix Ca2+-dependent decrease in the efflux α-ketoglutarate from mitochondria (Contreras and Satrústegui, 2009).
42
Figure 4: Ca2+ regulation of mitochondrial respiration: A: Pull mechanism: Ca2+ in the ma- trix could “pull” pyruvate into mitochondria independently of ARALAR-MAS. Ca2+ in the matrix could promote pyruvate dehydrogenase (PDH) complex activation by dephosphorylation. The increased ox- idation of pyruvate will “pull” pyruvate into mitochondria by mass action ratio on the mitochondrial pyruvate carrier. B: Push mechanism: extramitochondrial Ca2+-activation of ARALAR-MAS changes the cytosolic redox state of NAD/NADH and may promote the conversion of lactate to pyruvate. The increase in cytosolic pyruvate will function to “push” pyruvate into mitochondria by mass action ratio on the mitochondrial pyruvate carrier. Pyruvate entry would induce inhibition of PDH kinases which might be necessary for the full activation of PDH and respiration in intact neurons. PDH, pyruvate dehydrogenase (active); PDK, pyruvate dehydrogenase kinase; PDH-P, phosphorylated pyruvate de- hydrogenase (inactive); PDP-1 and PDP-2, calcium dependent and Ca2+-independent pyruvate dehy- drogenase phosphatase, respectively; Pyr, pyruvate; PyrC, pyruvate carrier; MAS, malate–aspartate shuttle. From Rueda et al. (2014).
As described, an important component of this interpretation is the Ca2+ sensi- tivity of MAS which is proposed to be critical in activating MAS in response to small and medium Ca2+ signals. One of the aims of the present Thesis was to verify this possibility by engineering a mouse mutant in which the Ca2+ binding domain of ARALAR was selectively disrupted leaving the remaining of the protein and its promotor region intact (Fig. 5). This part of the Thesis work, (ARALAR-KI mouse) is described in the methods and results sections.
5. Astrocytes response to energy demand
The structural situation of astrocytes makes those cells the target of many signals similar to the ones that neurons receive. In section 2, there were listed the ways whereby Ca2+ may couple ATP production to ATP demand, in parallel to ATP/ADP ratio change.
Figure 5: Generation of ARALAR with disrupted calcium binding domains. Alignment of the N-terminal domain of representative sequences of ARALAR (from human, orangutan, and mouse).
Boxes indicate the position of structure loops of amino acids that made the predicted Ca2+coordina- tion cage. Residues at the 27 and 29 positions EF-hand 1 and residues at the 65 and 67 positions of EF-hand 2 are subsituted to disable functional Ca2+binding. From Marmol et al. 2009.
5.1. Calcium signals in astrocytes
In vitro studies have long time ago described Ca2+ signals and Ca2+ waves in astrocytes, however the role and the origin of many calcium signals in astrocytes described in vivo is to be determined. The development of gene encoded Ca2+
probes (GECIs) targeted to plasma membrane revealed abundant types of Ca2+
signals, with different effects depending the brain region and affecting different astrocyte domains. Ca2+ signals in microdomains are random, seconds timescale, miniature, ‘spotty’ and transmembrane nature (Khakh and McCarthy, 2015; Shige- tomi et al., 2010). However, this work is carried out in cultured astrocytes and focused in global Ca2+ signals measured with Fura-2 (Ca2+ organic probe) (Fiacco et al., 2009) and mitochondrial Ca2+.
The variability and the relevance of Ca2+ signals in astrocytes is clear (Khakh and McCarthy, 2015). Interestingly, there is increasing evidence that mitochondria are localized in small astrocytes processes (Oheim et al., 2018; Stephen et al., 2015) and that these mitochondria are localized near synapses. Mitochondria move along processes (Stephen et al., 2015) and their position is regulated by Ca2+, glutamate uptake and Miro1 protein (Jackson et al., 2014; Jackson and Robinson, 2018; Stephen et al., 2015) suggesting a role in the response to neuronal signals.
We have used cultured astrocytes to study of the respiratory responses to relevant 44
agonists (ATP and glutamate) released by activated neurons and by astrocytes (Covelo and Araque, 2018, Parpura et al., 2017). We have specifically addressed the type of workloads imposed on astrocytes by these agonists and the role of calcium in the final respiratory response.
5.2. ATP
ATP is a universal signal in the brain, released by neurons and astrocytes. ATP was described as a signaling molecule in 70’ (Burnstock, 1972). ATP is a co- trans- mitter that acts on metabotropic (P2Y) and ionotropic purinergic receptors (P2X), both of which are present in astrocytes (Franke et al., 2012). Neuronal activity, K+ and ATP itself induces ATP release from astrocytes (Anderson et al., 2004; Covelo and Araque, 2018; Scemes and Spray, 2012).
The concentration of extracellular ATP can vary from around 100 μM when re- leased from neurons by exocytosis (Richardson and Brown, 1987) to reach patho- logical concentrations after epileptic seizures. However, extracellular ATP concen- trations are uncertain because of the action of ectonucleotidases which metabolize ATP, to ADP or adenine or adenosine which also have receptors in astrocytes as neurons.
P2Y receptors: ATP sensitive metabotropic purinergic receptors coupled to G proteins. P2Y can interact with Gqα proteins which activate phospholipase C pro- ducing IP3witch releases Ca2+ from internal stores after activation of IP3receptors.
In astrocytes the main IP3 receptor is IP3R2 (Petravicz et al., 2014; Srinivasan et al., 2015), but other IP3 receptors appear to exist (Sherwood et al., 2017). P2X receptors: ATP sensitive ionotropic receptors are ATP -gated ion channels allowing entry of Na+, Ca2+ and K+. There are seven different subunits and they can form homotrimeric or heterotrimeric channels (Nicke et al., 1998; Saul et al., 2013).
5.3. Glutamate
Glutamate is the most abundant excitatory neurotransmitter in the brain. It is released by presynaptic neurons and also by astrocytes (Covelo and Araque, 2018;
Navarrete and Araque, 2008; Navarrete and Araque, 2010; Parpura and Haydon, 2000; Perea et al., 2016).
Figure 6: Scheme of glutamatergic signal that may activate mitochondrial function. Ca2+
signaling in astrocyte processes stop mitochondria near synapse activity. Increases in neuronal activ- ity result in increases Ca2+in astrocytes. Calcium is removed from the cytosol into the mitochondria via MCU. Increases in matrix [Ca2+] stimulate dehydrogenases in the TCA cycle Ca2+accumulated by the mitochondria may be released via transient opening of the permeability transition pore or by the mitochondrial Na+/Ca2+-exchanger (NCLX). Glutamate can be transferred into the mitochondria and subsequently converted to a- ketoglutarate (a-KG) for entry into the TCA cycle. Na+/Ca2+exchanger (NCX), glutamate transporters GLT1 and GLAST, glucose transporter (GLUT1). Mitochondrial calcium uniporter (MCU), permeability transition pore (PTP), Na+/Ca2+ mitochondrial exchanger (NCXL), α- ketoglutarate (α-KG), Krebs cycle (TCA). From Jackson and Robinson (2018).
Glutamate transporters
In astrocytes glutamate is captured by Na+ dependent glutamate transporters.
There are five isoforms, GLAST/EAAT1/Slc1a3, GLT-1/EAAT2/Slc1a2, EAAC1/Slc1a1, EAAT4/Slc1a6, and EAAT5/Slc1a7 proteins (Robinson M.B. and Jackson J.G., 2016).
Glutamate transporters introduce one glutamate molecule with three Na+, one H+ in exchange of one K+ molecule (stoichiometry: 1 glutamate- 3 Na+-1 H+/ 1 K+).
Most of glutamate clearance is done by GLT-1 (or EAAT2) and GLAST (or EAAT1).
GLAST is expressed by astrocytes and oligodendrocytes in the nervous system (Domerck et al., 2005; Rothstein et al., 1994). GLT-1 expression is much higher in astrocytes than in other cells in the brain (Rothstein et al., 1994; Sutherland
46
et al., 1996). It has been suggested that GLT-1 and GLAST are part of a multi- protein complex that also includes mitochondria (Bauer et al., 2012; Genda et al., 2011; Robinson M.B. and Jackson J.G., 2016; Rose et al., 2009). Describing that Na+/K+ ATPase, different glycolytic enzymes and two different mitochondrial pro- teins co-immunoprecipitate with GLAST (Bauer et al., 2012). Similar results have been described for GLT-1 (Rose et al 2009, Genda et al) (Fig. 6).
Glutamate receptors
Astrocytes have ionotropic NMDA receptors (Jimenez-Blasco et al., 2015) and metabotropic receptors, which mobilize Ca2+ stores via IP3 receptors (Sun et al 2013). Signaling through glutamate metabotropic receptors is relevant in neonatal tissue, but not in adult brain astrocytes (Sun et al., 2013).
Glutamate metabolism
Glutamate-glutamine cycle, glutamate recycling: It is the pathway to return neurotransmitter glutamate to the neuron. Glutamate is taken up by astro- cytes and transformed to glutamine by glutamine synthetase, an ATP consum- ing enzyme present mainly in astrocytes (Norenberg and Martinez-Hernandez, 1979). Glutamine is released from astrocytes and captured by neurons.
Glutamate oxidation: Complete oxidation and partial oxidation of glutamate.
in vivo 13C-NMR labeling experiments have shown that the rate of glutamate synthesis is equal to 20-25% of that of glucose oxidation in the brain (Cotrina et al., 2000; Guthrie et al., 1999) reviewed in Dienel and McKenna (2014).
As glutamate levels are constant, this entails that glutamate oxidation rate should be also 20-25 % of the glucose oxidation rate in the brain. Both reac- tions occur mainly in the astrocytes (Karaca et al., 2015; Nissen et al., 2017;
Pardo et al., 2011). The complete oxidation of glutamate requires pyruvate recycling, a pathway discovered by Cerdan et al. (1990), in which a four car- bon Krebs Cycle intermediate is converted into pyruvate to be decarboxylated and reenter in the Krebs cycle as Acetyl-CoA. For this complete degradation of glutamate, a degradation product of α-ketoglutarate must exit from the cycle and be decarboxylated to form pyruvate that can be converted to Acetyl CoA
