Single-particle cryo-electron microscopy requires a protein sample that is correctly folded, monodisperse and available in sufficient amount. For a membrane protein this requirement is the main experimental limit, because the protein must be extracted from the membrane lipid bilayer and kept stable in a detergent micelle. This thesis concerns this step for two targets studied in the laboratory: the human sweet taste receptor TAS1R2R3 and an ancestrally reconstructed TREK1 potassium channel (ancTREK1). For TAS1R2R3 the structural objective is the complex between receptor and the proteinic sweeteners, a mode of interaction for which no structures are available. For ancTREK1 the objective is the comparison of the channel at 4°C and at 40°C, to describe the structural basis of its response to the change of temperature. TAS1R2R3 was expressed as a split-GFP construct, in which a functional GFP is reconstituted only when the TAS1R2 and TAS1R3 subunits assemble. To establish a recombinant expression protocol, firstly, the protein expression was compared in HEK293 EBNA and HEK293F cells, using PEI-mediated transfection and baculovirus-mediated transduction. The detergent screening performed on these membranes showed that the receptor was eluted in the void volume of the FSEC profile, indicating that it was recovered as high molecular weight aggregates rather than a homogeneous detergent-solubilized particle. The solubilization was therefore identified as the limiting step, possibly related to aggregation. The TAS1R2R3 was then produced in Sf9 cells, where the same construct was co-expressed with the cognate G protein and the scFv16 antibody fragment. Neither this co-expression nor the optimization of the solubilization conditions and purification protocol solved this problem, and the receptor was recovered mainly in the flowthrough and the wash fractions of the IMAC and affinity chromatography steps. The constructs were therefore redesigned, replacing the wild type endogenous leader sequence with the HA leader sequence and expressed in HEK293 cells. In a small-scale purification the receptor was for the first time retained in the resin and the eluate gave an FSEC profile with a defined peak, clearly separated from the void peak, and the GFP-tagged heterodimer was detected in the same fraction by fluorescence analysis. The ancTREK1 channel was first expressed in Pichia pastoris, the system in which the modern murine TREK1 and other K2P channels had been produced. The detergent screening of membranes obtained after the scale-up of the culture showed that DDM and DM solubilized the channel with comparable efficiency and DDM was selected. The channel was then purified by affinity chromatography, exchanging the detergent from DDM to LMNG, and by SEC. The elution profile showed a broad and asymmetric peak, in contrast with the narrow peak observed during the screening. The expression was therefore moved to Expi293 cells, where the detergent screening was repeated under the same two thermal conditions: DDM and LMNG both preserved a symmetric and stable peak, and LMNG was selected for the purification. Two small-scale purifications were performed: the second gave a protein of good quality with a single, well-defined peak, and a good recovery. The work defined, for each target, the expression system, the construct and the solubilization conditions required to obtain a sample suitable for structural work. For both proteins the remaining step is the production at the scale required for cryo-EM analysis, which would describe how proteinic sweeteners are recognized by the human sweet taste receptor and how temperature controls the gating of an ancestral TREK1 channel.
La crio-microscopia elettronica a singola particella richiede un campione proteico correttamente ripiegato, monodisperso e disponibile in quantità sufficiente. Per una proteina di membrana, questo requisito è il principale limite sperimentale, perché la proteina deve essere estratta dalla membrana a doppio strato lipidico e mantenuta stabile in una micella di detergente. Questa tesi riguarda tale passaggio per due proteine studiate in laboratorio: il recettore umano del gusto dolce TAS1R2R3 e un canale del potassio TREK1 ricostruito per via ancestrale (ancTREK1). TAS1R2R3 è stato espresso come costrutto split-GFP, nel quale una GFP funzionale si ricostituisce solo quando le subunità TAS1R2 e TAS1R3 si assemblano. Per mettere a punto un protocollo di espressione ricombinante, la proteina è stata prodotta per essere confrontata in cellule HEK293 EBNA e HEK293F, mediante trasfezione via PEI e trasduzione con baculovirus ricombinante. Lo screening dei detergenti eseguito su queste membrane ha mostrato che il recettore eluiva nel void volume del profilo FSEC, indicando che veniva recuperato come aggregato ad alto peso molecolare e non come particella omogenea solubilizzata in detergente. La solubilizzazione è stata quindi individuata come il passaggio limitante e possibilmente legato all’aggregazione. TAS1R2R3 è stato perciò prodotto in cellule Sf9, dove lo stesso costrutto è stato co-espresso con la proteina G corrispondente e con il frammento anticorpale scFv16. Né questa co-espressione né l’ottimizzazione delle condizioni di solubilizzazione e del protocollo di purificazione hanno risolto il problema, e il recettore è stato recuperato principalmente nel flowthrough e nei lavaggi dell’IMAC e della cromatografia di affinità. A questo punto i costrutti sono stati ridisegnati, sostituendo la sequenza leader endogena wild type con la sequenza leader HA, ed espressi in HEK293. In una purificazione su piccola scala il recettore è stato per la prima volta trattenuto nella resina e l’eluato ha dato un profilo FSEC con un picco definito, nettamente separato dal picco del void volume, e l’eterodimero legato alla GFP è stato rilevato nella stessa frazione mediante analisi in fluorescenza. Il canale ancTREK1 è stato dapprima espresso in Pichia pastoris, il sistema nel quale erano stati prodotti il TREK1 murino moderno e altri canali K2P. Lo screening dei detergenti, eseguito sulle membrane ottenute dopo lo scale-up della coltura, ha mostrato come DDM e DM solubilizzavano il canale con efficienza comparabile, e il DDM è stato selezionato. Il canale è stato purificato prima per cromatografia di affinità, scambiando il detergente da DDM a LMNG, poi per SEC. Il profilo di eluizione ha mostrato un picco largo e asimmetrico, in contrasto con il picco stretto osservato durante lo screening. L’espressione è stata pertanto spostata in cellule Expi293, dove lo screening dei detergenti è stato ripetuto nelle stesse condizioni termiche: sia il DDM che LMNG hanno mantenuto un picco simmetrico stabile, e LMNG è stato selezionato per la purificazione. Sono state eseguite due purificazioni su piccola scala e la seconda ha dato una proteina di buona qualità e con una buona resa. Il lavoro ha definito, per ciascun bersaglio, il sistema di espressione, il costrutto e le condizioni di solubilizzazione necessarie a ottenere un campione adatto al lavoro strutturale. Per entrambe le proteine il passaggio che resta da compiere è la produzione sulla scala richiesta per l’analisi in crio-EM, che descriverebbe come le proteine dolci vengono riconosciute dal recettore del gusto dolce umano e come la temperatura controlla un canale TREK1 ancestrale.
Il recettore umano del gusto dolce TAS1R2R3 e il canale ancestrale TREK1 come bersagli per la crio-EM a singola particella: espressione ricombinante, purificazione e caratterizzazione.
DECANDIA, MANUEL
2025/2026
Abstract
Single-particle cryo-electron microscopy requires a protein sample that is correctly folded, monodisperse and available in sufficient amount. For a membrane protein this requirement is the main experimental limit, because the protein must be extracted from the membrane lipid bilayer and kept stable in a detergent micelle. This thesis concerns this step for two targets studied in the laboratory: the human sweet taste receptor TAS1R2R3 and an ancestrally reconstructed TREK1 potassium channel (ancTREK1). For TAS1R2R3 the structural objective is the complex between receptor and the proteinic sweeteners, a mode of interaction for which no structures are available. For ancTREK1 the objective is the comparison of the channel at 4°C and at 40°C, to describe the structural basis of its response to the change of temperature. TAS1R2R3 was expressed as a split-GFP construct, in which a functional GFP is reconstituted only when the TAS1R2 and TAS1R3 subunits assemble. To establish a recombinant expression protocol, firstly, the protein expression was compared in HEK293 EBNA and HEK293F cells, using PEI-mediated transfection and baculovirus-mediated transduction. The detergent screening performed on these membranes showed that the receptor was eluted in the void volume of the FSEC profile, indicating that it was recovered as high molecular weight aggregates rather than a homogeneous detergent-solubilized particle. The solubilization was therefore identified as the limiting step, possibly related to aggregation. The TAS1R2R3 was then produced in Sf9 cells, where the same construct was co-expressed with the cognate G protein and the scFv16 antibody fragment. Neither this co-expression nor the optimization of the solubilization conditions and purification protocol solved this problem, and the receptor was recovered mainly in the flowthrough and the wash fractions of the IMAC and affinity chromatography steps. The constructs were therefore redesigned, replacing the wild type endogenous leader sequence with the HA leader sequence and expressed in HEK293 cells. In a small-scale purification the receptor was for the first time retained in the resin and the eluate gave an FSEC profile with a defined peak, clearly separated from the void peak, and the GFP-tagged heterodimer was detected in the same fraction by fluorescence analysis. The ancTREK1 channel was first expressed in Pichia pastoris, the system in which the modern murine TREK1 and other K2P channels had been produced. The detergent screening of membranes obtained after the scale-up of the culture showed that DDM and DM solubilized the channel with comparable efficiency and DDM was selected. The channel was then purified by affinity chromatography, exchanging the detergent from DDM to LMNG, and by SEC. The elution profile showed a broad and asymmetric peak, in contrast with the narrow peak observed during the screening. The expression was therefore moved to Expi293 cells, where the detergent screening was repeated under the same two thermal conditions: DDM and LMNG both preserved a symmetric and stable peak, and LMNG was selected for the purification. Two small-scale purifications were performed: the second gave a protein of good quality with a single, well-defined peak, and a good recovery. The work defined, for each target, the expression system, the construct and the solubilization conditions required to obtain a sample suitable for structural work. For both proteins the remaining step is the production at the scale required for cryo-EM analysis, which would describe how proteinic sweeteners are recognized by the human sweet taste receptor and how temperature controls the gating of an ancestral TREK1 channel.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi Magistrale Manuel Decandia.pdf
non disponibili
Dimensione
36.2 MB
Formato
Adobe PDF
|
36.2 MB | Adobe PDF | Richiedi una copia |
È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per maggiori informazioni e per verifiche sull'eventuale disponibilità del file scrivere a: [email protected].
https://hdl.handle.net/20.500.14239/37426