Interaction of the Pacemaker Channel HCN1 with Filamin A

Pacemaker channels are encoded by the HCN gene family and are responsible for a variety of cellular functions including control of spontaneous activity in cardiac myocytes and control of excitability in different types of neurons. Some of these functions require specific membrane localization. Although several voltage-gated channels are known to interact with intracellular proteins exerting auxiliary functions, no cytoplasmic proteins have been found so far to modulate HCN channels. Through the use of a yeast two-hybrid technique, here we showed that filamin A interacts with HCN1, an HCN isoform widely expressed in the brain, but not with HCN2 or HCN4. Filamin A is a cytoplasmic scaffold protein with actin-binding domains whose main function is to link transmembrane proteins to the actin cytoskeleton. Using several HCN1 C-terminal constructs, we identified a filamin A-interacting region of 22 amino acids located downstream from the cyclic nucleotide-binding domain; this region is not conserved in HCN2, HCN3, or HCN4. We also verified by immunoprecipitation from bovine brain that the filamin A-HCN1 interaction is functional in vivo. In filamin A-expressing cells (filamin+), HCN1 (but not HCN4) channels were expressed in hot spots, whereas they were evenly distributed on the membrane of cells lacking filamin A (filamin–) indicating that interaction with filamin A affects membrane localization. Also, in filamin– cells the gating kinetics of HCN1 were strongly accelerated relative to filamin+ cells. The interaction with filamin A may contribute to localizing HCN1 channels to specific neuronal areas and to modulating channel activity. Pacemaker channels are encoded by the HCN gene family and are responsible for a variety of cellular functions including control of spontaneous activity in cardiac myocytes and control of excitability in different types of neurons. Some of these functions require specific membrane localization. Although several voltage-gated channels are known to interact with intracellular proteins exerting auxiliary functions, no cytoplasmic proteins have been found so far to modulate HCN channels. Through the use of a yeast two-hybrid technique, here we showed that filamin A interacts with HCN1, an HCN isoform widely expressed in the brain, but not with HCN2 or HCN4. Filamin A is a cytoplasmic scaffold protein with actin-binding domains whose main function is to link transmembrane proteins to the actin cytoskeleton. Using several HCN1 C-terminal constructs, we identified a filamin A-interacting region of 22 amino acids located downstream from the cyclic nucleotide-binding domain; this region is not conserved in HCN2, HCN3, or HCN4. We also verified by immunoprecipitation from bovine brain that the filamin A-HCN1 interaction is functional in vivo. In filamin A-expressing cells (filamin+), HCN1 (but not HCN4) channels were expressed in hot spots, whereas they were evenly distributed on the membrane of cells lacking filamin A (filamin–) indicating that interaction with filamin A affects membrane localization. Also, in filamin– cells the gating kinetics of HCN1 were strongly accelerated relative to filamin+ cells. The interaction with filamin A may contribute to localizing HCN1 channels to specific neuronal areas and to modulating channel activity. The hyperpolarization-activated pacemaker current If/Ih has an established role in underlying generation and neurotransmitter-mediated modulation of cardiac pacemaker activity (1DiFrancesco D. Annu. Rev. Physiol. 1993; 55: 455-472Crossref PubMed Scopus (673) Google Scholar) and control of excitability and other functions, including integration of synaptic activity and modulation of synaptic strength, in several different types of neurons (2Pape H.C. Annu. Rev. Physiol. 1996; 58: 299-327Crossref PubMed Scopus (982) Google Scholar, 3Robinson R.B. Siegelbaum S.A. Annu. Rev. Physiol. 2003; 65: 453-480Crossref PubMed Scopus (911) Google Scholar). The molecular correlates of f/h channels are the hyperpolarization-activated cyclic nucleotide-gated channels, of which four isoforms (HCN1–4) are known. When expressed alone in heterologous systems, HCN channels elicit currents whose properties are qualitatively similar to those of native If/Ih currents, although they differ quantitatively in kinetics and cAMP sensitivity (4Viscomi C. Altomare C. Bucchi A. Camatini E. Baruscotti M. Moroni A. DiFrancesco D. J. Biol. Chem. 2001; 276: 29930-29934Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar, 5Altomare C. Bucchi A. Camatini E. Baruscotti M. Viscomi C. Moroni A. DiFrancesco D. J. Gen. Physiol. 2001; 117: 519-532Crossref PubMed Scopus (122) Google Scholar). Recently, several studies have addressed the problem of the distribution of HCN isoforms in different cell types. HCN1, the first HCN isoform to be cloned (6Santoro B. Grant S.G. Bartsch D. Kandel E.R. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 14815-14820Crossref PubMed Scopus (228) Google Scholar), is extensively expressed in the brain with a selective distribution in specific brain areas. It is expressed for example in the layer 5 neocortical neurons, in the CA1, and to a lower degree, CA3 hippocampal regions, in the molecular cell layer of the cerebellum, and in the superior colliculus (3Robinson R.B. Siegelbaum S.A. Annu. Rev. Physiol. 2003; 65: 453-480Crossref PubMed Scopus (911) Google Scholar, 7Santoro B. Tibbs G.R. Ann. N. Y. Acad. Sci. 1999; 868: 741-764Crossref PubMed Scopus (305) Google Scholar, 8Moosmang S. Biel M. Hofmann F. Ludwig A. Bio. Chem. 1999; 380: 975-980Crossref PubMed Scopus (205) Google Scholar, 9Kaupp U.B. Seifert R. Annu. Rev. Physiol. 2001; 63: 235-257Crossref PubMed Scopus (306) Google Scholar). Interestingly, in some of the tissues of expression such as the neocortex, the retina, the hippocampus, and the taste receptors, the HCN1 isoform was found to be present in specific cell types and/or in specific subcellular compartments (e.g. the inner segment of rods and the cortical dendrites of CA1 hippocampal neurons) both at the RNA and protein levels (10Magee J.C. J. Neurosci. 1998; 18: 7613-7624Crossref PubMed Google Scholar, 11Santoro B. Chen S. Luthi A. Pavlidis P. Shumyatsky G.P. Tibbs G.R. Siegelbaum S.A. J. Neurosci. 2000; 20: 5264-5275Crossref PubMed Google Scholar, 12Moroni A. Gorza L. Beltrame M. Gravante B. Vaccari T. Bianchi M.E. Altomare C. Longhi R. Heurteaux C. Vitadello M. Malgaroli A. DiFrancesco D. J. Biol. Chem. 2001; 276: 29233-29241Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar, 13Demontis G.C. Moroni A. Gravante B. Altomare C. Longoni B. Cervetto L. DiFrancesco D. J. Physiol. 2002; 542: 89-97Crossref PubMed Scopus (53) Google Scholar, 14Lorincz A. Notomi T. Tamas G. Shigemoto R. Nusser Z. Nat. Neurosci. 2002; 5: 1185-1193Crossref PubMed Scopus (340) Google Scholar). The correct functioning of various ion channels depends upon the interaction with auxiliary subunits (15Hanlon M.R. Wallace B.A. Biochemistry. 2002; 41: 2886-2894Crossref PubMed Scopus (88) Google Scholar) and with scaffolding proteins, which co-localize channels and regulatory components in subcellular compartments for improved efficiency of channel modulation (16Prat A.G. Cunningham C.C. Jackson Jr., G.R. Borkan S.C. Wang Y. Ausiello D.A. Cantiello H.F. Am. J. Physiol. 1999; 277: C1160-C1169Crossref PubMed Google Scholar, 17Petrecca K. Miller D.M. Shrier A. J. Neurosci. 2000; 20: 8736-8744Crossref PubMed Google Scholar, 18Eldstrom J. Doerksen K.W. Steele D.F. Fedida D. FEBS Lett. 2002; 531: 529-537Crossref PubMed Scopus (29) Google Scholar). The aim of this work was to investigate whether HCN1 channels interact with partner proteins in the brain and to check whether existing associations with any such proteins have a role in the specific cellular localization of the channels and modulation of channel function. We used the C terminus of mouse HCN1 (mHCN1), because this contains specific sequences for protein-protein interaction (including PDZ and SH3 1The abbreviations used are: SH, Src homology; aa, amino acid; MEM, minimal Eagle's medium; HA, hemagglutinin; m, mouse; h, human; CNBD, cyclic-nucleotide binding domain; GFP, green fluorescent protein; pS, picosiemens; pF, picofarad; domains), to screen a mouse brain cDNA library by yeast two-hybrid assay. We found that HCN1 associates with several proteins and specifically with filamin A, a high molecular mass cytoskeletal protein known to anchor different transmembrane proteins to the actin cytoskeleton and to act as a scaffold for various signaling used the two-hybrid to screen a library of mouse brain cDNA in with a of the C terminus of in as a The were identified by the for were in and The proteins filamin A A protein of the family of Y. N. S. Proc. Natl. Acad. Sci. U. S. A. 1999; PubMed Scopus Google B. 1999; PubMed Scopus Google Filamin A was for the filamin A-interacting of we several of C-terminal in as we with and and the was and and the for these were by the C terminus with and with and we a in The used were and and and and of was with and and the was and was with and the was this region and were in with a and with the was cloned the with the we used with the and and and and with the the C terminus of HCN1 was by from to a a and the was cloned the of The used for and were were by We used the to the yeast whereas the yeast was with A. The were and to the to the interaction found by the and and we used the with the and the In of the by the and of filamin A by were in and filamin A was in protein was by an in on to which cDNA and were of was for at or actin-binding and or cells C.C. PubMed Scopus Google Scholar) were in a bovine The cells were by in a and on at and cells were with or in a for at and the cells were by and at a lower for and as were with and for at in a to protein E. A. N. Proc. Natl. Acad. Sci. U. S. A. 1996; PubMed Scopus (90) Google Scholar). The cells were with at and with for 5 at this was by a with a for at were with a and on with cells were by with at and of bovine brain in of a with a and a a first for at the was and the was at for at The was of a and for on a for at the was and the used for the immunoprecipitation of membrane proteins were in of and with at on a The we of protein for h, and at for at The was with a 5 The proteins were by to and with for and cells used for immunoprecipitation were in as in and from the and from were in of and was as for brain the immunoprecipitation of proteins by the in were used alone or with filamin A for at for immunoprecipitation with and were as The was with a and to to the of and cells were in a fluorescent and at with fluorescent cells were for The was as the was to a high The a of in the of to by a at were to in currents at were to were to the is is and is the to and which were were by the current at and to the of to for a to were by to a function an C. Bucchi A. Camatini E. Baruscotti M. Viscomi C. Moroni A. DiFrancesco D. J. Gen. Physiol. 2001; 117: 519-532Crossref PubMed Scopus (122) Google Scholar), and the were and were to cell by a specific are as the was by a with a to cDNA by proteins the of the C terminus as a and found these proteins, some were check for we a yeast on a high which to the of proteins that were to to the HCN1 A a protein of the family of Y. N. S. Proc. Natl. Acad. Sci. U. S. A. 1999; PubMed Scopus Google Scholar), B. 1999; PubMed Scopus Google Scholar), and filamin A, and We on filamin A. The interaction of filamin A with HCN1 in yeast is in Filamin A, or actin-binding is a cytoskeletal protein an actin-binding at the and and and and A. A. 2001; PubMed Scopus Google Scholar). Filamin A as a because of specific as a scaffold example by binding a variety of proteins such as membrane (e.g. and (e.g. ion channels (e.g. (e.g. and proteins (e.g. and K. Miller D.M. Shrier A. J. Neurosci. 2000; 20: 8736-8744Crossref PubMed Google Scholar, Y. N. S. Proc. Natl. Acad. Sci. U. S. A. 1999; PubMed Scopus Google Scholar, A. A. 2001; PubMed Scopus Google Scholar, A. J. Biol. Chem. 1998; Full Text Full Text PDF PubMed Scopus Google Scholar, M. B. Biol. 2000; PubMed Scopus Google Scholar, R. K. N. P. R. Proc. Natl. Acad. Sci. U. S. A. 2001; PubMed Scopus Google Scholar, J. L. A. M. Nat. Rev. Biol. 2001; PubMed Scopus Google Scholar, J. J. Biol. 2001; PubMed Scopus Google Scholar). the of filamin A and the C-terminal segment by the HCN1 with the and from to whether filamin A and HCN1 also interact in in a and specifically in the both proteins are Using specific A we filamin A from of membrane proteins from a bovine The and membrane proteins were and to for In the filamin proteins a to the molecular mass of the HCN1 isoform A. Gorza L. Beltrame M. Gravante B. Vaccari T. Bianchi M.E. Altomare C. Longhi R. Heurteaux C. Vitadello M. Malgaroli A. DiFrancesco D. J. Biol. Chem. 2001; 276: 29233-29241Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar) as by a the was in the membrane proteins immunoprecipitation we verified the and of the filamin A in the filamin Also, HCN1 was not by the with a or by the HCN1 with in the of and in the filamin A-interacting of and to investigate whether HCN isoforms other HCN1 filamin A, we a yeast was by the yeast with filamin A and the with different C-terminal of or with the C of or and were in high yeast of interaction of the HCN C-terminal used to for interaction with filamin in a In a of yeast of filamin A and HCN C-terminal sequences is the C-terminal of used as a control for is the amino of the which interacts with filamin A. contains a conserved the HCN isoforms and in a known to SH3 domains and to be in some in filamin A binding (6Santoro B. Grant S.G. Bartsch D. Kandel E.R. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 14815-14820Crossref PubMed Scopus (228) Google Scholar, 17Petrecca K. Miller D.M. Shrier A. J. Neurosci. 2000; 20: 8736-8744Crossref PubMed Google Scholar), although filamin A not SH3 We first whether or not this be responsible for the interaction and filamin A and by amino acids with a known to filamin K. Miller D.M. Shrier A. J. Neurosci. 2000; 20: 8736-8744Crossref PubMed Google Scholar) The were to indicating that this is not for We also that the C-terminal of a conserved HCN isoforms for PDZ domains which are known to be in the of and protein scaffolding M. C. Annu. Rev. Neurosci. 2001; PubMed Scopus Google Scholar). check for a with filamin A in this we a from the C-terminal of the protein the PDZ in C and this not interact with filamin A, whereas indicating that the PDZ is not responsible for this We to the of from the control and four to a as in also that the and is for the sequences may contribute to protein-protein example with domains), we the to in the the not the interaction an of this in the binding with filamin A. The is not conserved in any of the other HCN isoforms HCN3, was verified by with as as a of sequences with and In with this the HCN2 and to the HCN1 control from the C-terminal of the cyclic-nucleotide binding downstream from the C of HCN2 or not interact with filamin A and The filamin A-interacting was downstream from the in a to the we also for protein-protein in and of the various to which a or was for In filamin A in with was an and was by as the control to was in and an and was as The proteins were and the immunoprecipitation was with the the of filamin A in the In similar were the as in the region amino acids and was the The from the yeast and the in are in investigate the role of the interaction filamin A and we and by the isoform with in a cell lacking filamin A or and the with those with the cell the filamin A or K. Miller D.M. Shrier A. J. Neurosci. 2000; 20: 8736-8744Crossref PubMed Google Scholar, C.C. PubMed Scopus Google Scholar). We first verified by immunoprecipitation that filamin A is expressed in but not in cells and that and filamin A, as in but not in cells The interaction was because no was cells were with the In and of and cells the protein are is a of protein expression in both cell types. In and cells were with for to protein and the and these the was in on the membrane of cells whereas in cells was evenly distributed on the membrane The was of in and in this as in and channels were expressed evenly on the membrane of both cell types. that filamin A to a channel that HCN1 channels are the the because the of filamin A not to the membrane of channels, the that this protein is not in HCN1 channel from the to the the of the A on the properties of channels, we currents by in both and cells. In the currents by to the to are for and cells the was in in cells and The cell not differ the pF, pF, In the for the types of cells are In cells lacking filamin A the was to the relative to filamin A-expressing cells by to the and and and for and of and were and and were in cells in cells. the and the in cells were similar to those for expressed in cells C. Bucchi A. Camatini E. Baruscotti M. Viscomi C. Moroni A. DiFrancesco D. J. Gen. Physiol. 2001; 117: 519-532Crossref PubMed Scopus (122) Google Scholar). whether the in current and in kinetics in were to the specific interaction channels and filamin A, we the properties of channels expressed in and cells. We not in any of the properties of were in and in and were and in and and in cells Also, the of were not different in and cells in the of to that the protein-protein interaction of channels with filamin A strongly affects the channel HCN channels are of a of ion channels which and cyclic-nucleotide channels. channels to this interact with and/or are by proteins, including to modulate ion channel activity and proteins in the of channels and is known that several channel isoforms interact with the or to currents similar to those in native Physiol. 1999; PubMed Scopus Google Scholar). Also, channels interact with whereas and channels have been to link to filamin A K. Miller D.M. Shrier A. J. Neurosci. 2000; 20: 8736-8744Crossref PubMed Google Scholar, 18Eldstrom J. Doerksen K.W. Steele D.F. Fedida D. FEBS Lett. 2002; 531: 529-537Crossref PubMed Scopus (29) Google Scholar, L. M. C. J. Biol. Chem. 2003; Full Text Full Text PDF PubMed Scopus Google Scholar). expression of HCN isoforms not in currents whose properties those of the native pacemaker currents (4Viscomi C. Altomare C. Bucchi A. Camatini E. Baruscotti M. Moroni A. DiFrancesco D. J. Biol. Chem. 2001; 276: 29930-29934Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar, C. B. C. R. C. Viscomi C. Moroni A. Baruscotti M. DiFrancesco D. J. Physiol. 2003; PubMed Scopus Google Scholar). of this may be to the of of channel a on such as of the channel protein or the interaction with for channel kinetics F. DiFrancesco D. M.R. C. J. Physiol. PubMed Scopus Google Scholar) and the of expressed channels G. DiFrancesco D. J. Physiol. 1997; PubMed Scopus Google has been that HCN channels are by interaction with the J. B. J. Z. Wang R.B. M.R. J. D. R. 2001; PubMed Google Scholar, N. F. R. K. 2003; Scopus Google Scholar), the of this interaction is C. B. C. R. C. Viscomi C. Moroni A. Baruscotti M. DiFrancesco D. J. Physiol. 2003; PubMed Scopus Google Scholar). The of a of HCN properties is by that some of the of a isoform in different expression J. Altomare C. Bucchi A. DiFrancesco D. R.B. J. Physiol. 2002; PubMed Scopus Google Scholar). in which intracellular channel function is protein-protein for HCN channels we used the yeast two-hybrid and a cDNA library of mouse a we used the HCN1 because of expression in various of the (e.g. neocortex, hippocampus, and B. Chen S. Luthi A. Pavlidis P. Shumyatsky G.P. Tibbs G.R. Siegelbaum S.A. J. Neurosci. 2000; 20: 5264-5275Crossref PubMed Google Scholar, 14Lorincz A. Notomi T. Tamas G. Shigemoto R. Nusser Z. Nat. Neurosci. 2002; 5: 1185-1193Crossref PubMed Scopus (340) Google and yeast the used to and that the C terminus interacts with filamin A, A, and of A Interestingly, of these proteins are in the of the actin cytoskeleton and has been that A interacts with of filamin A Y. N. S. Proc. Natl. Acad. Sci. U. S. A. 1999; PubMed Scopus Google Scholar, J. L. A. M. Nat. Rev. Biol. 2001; PubMed Scopus Google Scholar), the region that we have found to interact with The of the on the filamin A may that A and HCN1 that A, filamin A, and HCN1 are in a It is also to that of the of which A is a are in the modulation of channels such as and A Hofmann F. P. K. J. 1999; PubMed Scopus Google Scholar, C. Hofmann F. J. K. J. Neurosci. 2000; 20: PubMed Google Scholar). We found that the interaction HCN1 and filamin A is conserved in in as by of the proteins from bovine brain It is known that filamin A interacts with several membrane proteins, those are the channels and which to filamin A at C-terminal is to the K. Miller D.M. Shrier A. J. Neurosci. 2000; 20: 8736-8744Crossref PubMed Google Scholar, L. M. C. J. Biol. Chem. 2003; Full Text Full Text PDF PubMed Scopus Google Scholar). Although the is conserved in we found that is not responsible for the interaction filamin A and Using the yeast we the C terminus region with filamin A to a of 22 amino acids downstream from the region a which contribute to a protein-protein interaction with filamin A. by the that not the binding to filamin A A region to this was not found in the C of the HCN2 or isoforms as verified by the of interaction with filamin A in the yeast in and K. Although we not investigate the interaction and filamin A, no region to the HCN1 binding was found by or is the function of the interaction filamin A and The binding of filamin A be a the of HCN1 channels in of the cell by the in and although HCN1 channels are distributed in a in cells lacking filamin A, expression areas are and channels are evenly distributed on the was to the specific A because was not in filamin– or filamin+ cells with and was not to the lower of protein expression in relative to because no was with which was expressed in and cells but at a lower in cells as by the of filamin A and and has been to be to downstream signaling components R. K. N. P. R. Proc. Natl. Acad. Sci. U. S. A. 2001; PubMed Scopus Google Scholar), the that filamin A to the of subcellular in which receptors, and are in to a function in channel localization and from that filamin A a protein responsible for the of M. B. Biol. 2000; PubMed Scopus Google Scholar). a of membrane known to membrane proteins in proteins of the to in cardiac J. Biol. Chem. 2000; Full Text Full Text PDF PubMed Scopus Google Scholar). the four HCN isoforms HCN1 is the to and of the channel by cAMP are of the of a (4Viscomi C. Altomare C. Bucchi A. Camatini E. Baruscotti M. Moroni A. DiFrancesco D. J. Biol. Chem. 2001; 276: 29930-29934Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar, 11Santoro B. Chen S. Luthi A. Pavlidis P. Shumyatsky G.P. Tibbs G.R. Siegelbaum S.A. J. Neurosci. 2000; 20: 5264-5275Crossref PubMed Google Scholar, C. B. C. R. C. Viscomi C. Moroni A. Baruscotti M. DiFrancesco D. J. Physiol. 2003; PubMed Scopus Google Scholar, S. Wang J. Siegelbaum S.A. J. Gen. Physiol. 2001; 117: PubMed Scopus Google Scholar). we found that HCN1 filamin A, this to with the that interaction with filamin A the of of the cAMP in for improved channel tissues HCN channels isoform S. Biel M. Hofmann F. Ludwig A. Bio. Chem. 1999; 380: 975-980Crossref PubMed Scopus (205) Google Scholar), and be by of HCN1 and HCN2 S. Wang J. Siegelbaum S.A. J. Gen. Physiol. 2001; 117: PubMed Scopus Google Scholar, C. J. J. Biol. Chem. 2001; 276: Full Text Full Text PDF PubMed Scopus Google Scholar), or HCN1 and C. B. C. R. C. Viscomi C. Moroni A. Baruscotti M. DiFrancesco D. J. Physiol. 2003; PubMed Scopus Google Scholar). is that HCN1 subunits channels with an function whose aim is to channels in specific membrane whereas the function on the cAMP sensitivity of HCN2 or in C and also that the interaction with filamin A was not for channel from the to the the of protein by not an of in the of cells indicating that channels the membrane or in the of filamin A. The of interaction filamin and not a of the channel in the membrane but also a in the properties of the expressed In in filamin– cells the HCN was that in filamin+ and both channel and were In with the that HCN1 properties are by a specific interaction with filamin A, we found no in the properties of channels expressed in or cells. is with the that filamin A channel expression in specific of channels to the membrane and the channel the other we not have a for the of filamin A binding to channels. is established that the C terminus is in the kinetics of HCN channels (4Viscomi C. Altomare C. Bucchi A. Camatini E. Baruscotti M. Moroni A. DiFrancesco D. J. Biol. Chem. 2001; 276: 29930-29934Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar, M. B. Siegelbaum S.A. Tibbs G.R. 2001; PubMed Scopus Google Scholar, J. Chen S. Siegelbaum S.A. 2002; Full Text Full Text PDF PubMed Scopus Google Scholar), and the to the binding of the C terminus to filamin A an with the by which the C In showed that HCN1, but not HCN2 or to an isoform of filamin A. The interaction the C-terminal of HCN1 and the at the C-terminal of filamin A, the of the protein whereas the actin-binding domains of the protein to filamin A HCN1 channel kinetics and channels to of the cell the of channel expression and A function of HCN1 channel is to the efficiency of channel control by modulating We C. C. F. Cantiello J. of and Shrier for the of and cells and for and U. for in

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.