• Open Access

Synthesis of membrane proteins in eukaryotic cell-free systems

Authors


Abstract

Cell-free protein synthesis (CFPS) is a valuable method for the fast expression of difficult-to-express proteins as well as posttranslationally modified proteins. Since cell-free systems circumvent possible cytotoxic effects caused by protein overexpression in living cells, they significantly enlarge the scale and variety of proteins that can be characterized. We demonstrate the high potential of eukaryotic CFPS to express various types of membrane proteins covering a broad range of structurally and functionally diverse proteins. Our eukaryotic cell-free translation systems are capable to provide high molecular weight membrane proteins, fluorescent-labeled membrane proteins, as well as posttranslationally modified proteins for further downstream analysis.

Abbreviations
CFPS

cell free protein synthesis

EGFR

epidermal growth factor receptor

E-PCR

expression polymerase chain reaction

Grb2

growth factor receptor-bound protein 2

EGFRtrunc

EGFR with vIII deletion, endogenenous signal peptide substituted by melittin signal sequence

Mel

melittin signal peptide

PTM

posttranslational modification

RRL

rabbit reticulocyte lysate

Sf

Spodoptera frugiperda

WGE

wheat germ extract

1 Introduction

Membrane proteins are involved in essential physiological mechanisms ensuring structural and functional integrity of cells. They are the key components of many processes such as cell recognition, immune response, signal transduction, and transport of molecules. The aberrant and altered function of membrane proteins can lead to severe disorders. For example, mutations in integral membrane proteins are implicated in diseases associated with anomalous function of ion channels, such as heart dysfunction, migraine, or Alzheimer's disease [1, 2]. Furthermore, errors in signaling pathways can result in loss of control of cell division and lead to cancer development. Thus, it is not a coincidence that more than 50% of all drugs are targeting membrane proteins [3, 4]. Despite of the fact that membrane proteins cover up to 30% of total proteins encoded in the human genome [5], we know relatively little about their molecular interactions. The reason is not only the naturally low abundance of these proteins on cell membranes but also their high hydrophobicity which, upon extraction from their natural environment, often leads to precipitation and thus complicates structural and functional studies. A major challenge in membrane protein studies is the preparation of sufficient amounts of correctly folded proteins maintaining their physiological function. Preparation of membrane proteins using conventional cell-based methods often fails due to the increased integration of overexpressed protein into the host's membrane. Such an extensive incorporation of foreign molecules negatively affects the structural integrity of host membrane components, leading to altered cell metabolism or even cell lysis. Another problem that is often to be faced when trying to express membrane proteins in prokaryotic cultures is the formation of aggregated proteins (inclusion bodies).

These limitations can be circumvented using cell-free protein synthesis (CFPS), also termed in vitro translation, whose performance does not depend on the structural integrity of cells. CFPS uses cell lysates, rather than whole cells, as a source of the entire protein translational machinery. All molecular factors such as ribosomes, tRNAs, and translation factors are present in these lysates. Only essential amino acids and energy factors must be supplemented externally. Using such a system, a broad range of versatile proteins can be produced, including difficult-to-express membrane proteins or cytotoxic proteins [6]. Beneficial characteristics of cell-free systems such as the fast access to the protein of interest, low reaction volumes, and short reaction times enable high-throughput protein expression strategies. Furthermore, CFPS facilitates the expression of engineered proteins with inserted non-canonical amino acids [7]. For illustration, the principle of cell-free synthesis is depicted in Fig.1.

Figure 1.

Proteins can be expressed from a variety of DNA templates including circular and linearized plasmids as well as PCR products. In the linked transcription–translation procedure, an RNA aliquot of the initial transcription step is added to the cell-free extract. Extracts from prokaryotic and eukaryotic cells supply the entire translational machinery. EF, elongation factor; IF, initiation factor; RF, release factor.

The prerequisite of using any in vitro translation system is its potential to produce a functional protein. Since the folding and activity of many proteins is stabilized by covalent post-translational modifications (PTMs) such as phosphorylation, farnesylation, acetylation, palmitoylation, sulfide linkers, and glycosylation, it is of high interest to ensure PTM processing also in in vitro systems. In eukaryotic cells, the majority of PTMs are inserted into to the protein structure in the ER. In prokaryotic cells, PTM of oxidative folding takes place in the outer membrane. Thus, it is the presence of membrane compartments that is crucial to ensure the correct fulfillment of the genotype–phenotype pathway.

1.1 Lysates for CFPS

Commercially available cell-free systems are derived from Escherichia coli extracts, wheat germ extracts (WGE), rabbit reticulocytes (RRL), and lysates from insect cells (Spodoptera frugiperda cells Sf21, Qiagen EasyXpress Insect Kit II). Cell extracts from E. coli provide high protein yield [8], but a major drawback of these lysates is the absence of the glycosylation machinery. Nevertheless, it was shown that extracts from E. coli transformed with the entire protein glycosylation locus gene cluster from gram negative bacteria Campylobacter jejuni can perform N-linked protein glycosylation [9, 10]. In this case, the oligosaccharyl transferase from C. jejuni is able to transfer sugars posttranslationally to specific structures in proteins even in the absence of membrane organelles [11].

However, a better platform for the rapid and efficient synthesis of eukaryotic proteins is provided by eukaryotic lysates. RRL and WGE are widely used to characterize proteins and investigate mRNA translational mechanisms [12-15]. WGE are very popular since they are scalable and easy to operate. They lack inhibitors and provide high yield of proteins in their native form [16]. RRL, on the other hand, are expensive, difficult to prepare, and have not yet been optimized to provide optimal amounts of synthesized proteins. Although both of these lysates have the potential to yield full-length proteins with correct fold, they also hold significant drawbacks. WGEs are not able to incorporate mammalian co-translational modifications and PTMs such as signal peptide cleavage and glycosylation. Similarly, extracts from RRL must be enriched with heterologous membrane fractions to produce glycosylated proteins, yet this is associated with decreased protein yield [17]. Thus, despite of the fact that in vitro translation systems based on prokaryotic extracts and extracts from WGE or RRL do provide benefits over conventional in vivo systems, their use for production of glycosylated human proteins is limited. In this context, insect lysates represent an interesting alternative for the high-yield production of glycosylated proteins [18].

Particular insect cell lines, e.g. Sf21 cells, grow well in suspension cultures and these cells can be easily scaled up for the large-scale production of recombinant proteins. High-yield in vivo expression of biologically active recombinant proteins is frequently achieved in both Sf9 and Sf21 cells. These insect cells can be lysed mechanically and the homogenate can be further processed for the defined preparation of translationally active lysates [18, 19]. Mild treatment during cell lysis ensures intact subcellular membranous structures (microsomes) derived from the ER. These membrane vesicles are crucial prerequisites for the production of functionally active membrane proteins with their subsequent PTM.

1.1.1 Template generation

The design of an expression template has significant influence on the efficiency of cell-free protein production. Templates suitable for CFPS can be applied in translationally active lysates either in the form of DNA templates or as mRNA templates. DNA templates can be supplemented as cDNA template in a complex mixture. They can also be directly applied to the translational systems in a circular form, e.g. plasmid, engineered by cloning or in form of linear PCR products. Template generation is the first step in CFPS process and its design has a significant influence on the overall process. The template construct directly encodes not only the primary but also the secondary structure of mRNA and thus has a strong impact on translational initiation. It is worth to evaluate the level of expression efficiency of several constructs to compare the yield of a given target protein in its full-length form. A fast method to monitor the efficiency of different constructs is expression polymerase chain reaction (E-PCR) that amplifies a single gene within a two-step PCR. In this procedure, the DNA template is supplemented with all necessary structural elements for transcription [20]. An essential prerequisite for the successful production of the final transcript is the presence of a promoter sequence upstream of the gene of interest. The generation of transcripts requires the presence of several other elements, such as the Shine–Dalgarno sequence in prokaryotic systems. Requirements in eukaryotic systems are even more complex. They include the addition of a Cap structure at the 5′ end of the encoded sequence or an internal ribosome entry site (IRES) element upstream of the translation initiation site [21]. These sequence elements can be introduced into the initial DNA construct assuring that the produced mRNA is equivalent to mature mRNA in living eukaryotic cells in terms of stability, accuracy, and translation efficiency. Additionally, E-PCR enables fast introduction of tag sequences and desired mutations into the coding sequence of a gene. Tag sequences and positions of these tags may influence the yield and solubility of the cell-free synthesized protein dramatically [22, 23].

1.1.2 In vitro transcription

In vitro transcription can be performed in batch reactions and requires a purified DNA template, T7 polymerase that catalyzes the formation of mRNA, cap analogs for synthesis of 5′capped mRNA, which exhibits higher translation efficiency in eukaryotic cell-free systems, nucleotides that represent mRNA building blocks, and source of energy. Furthermore, RNase inhibitors are required to protect the messenger RNA from RNAse degradation. In linked cell-free systems, transcription reactions are performed in a separate step followed by protein translation after an intermediate step of mRNA purification. However CFPS can also be performed in a one batch reaction, where the reaction conditions are the result of compromise for getting the optimal yields for both the transcription and translation reaction. These cell-free systems are called coupled transcription–translation systems. Both procedures take only a few hours and can be directly followed by functional studies of the de novo synthesized target protein.

1.1.3 In vitro translation

Primary cell lysates contain the whole complex translational machinery including ribosomes, tRNAs, soluble enzymes, initiation and elongation factors, as well as all the enzymes and factors necessary for fashioning the PTMs. For a higher protein yield, the primary lysate is supplemented with essential amino acids, ribonucleoside triphosphates (NTPs), and NTP regenerating systems. Lysates from eukaryotic cells are usually pretreated with micrococcus nucleases, to support the translation of only supplied exogenous mRNA [19, 24]. The optimization of translation conditions, such as incubation time, temperature, and composition of reaction mixtures is necessary to reach a high yield of functional protein. Changes in ratio of ADP, ATP, and GTP have significant influence on protein initiation and elongation and their concentration is maintained by an efficient energy regeneration system [25]. The translation efficiency remains strongly regulated in cell-free systems, thus after a specific incubation time, a plateau of protein amount is reached. Main factors responsible for translational regulation in eukaryotic cell-free systems are eukaryotic initiation factor 2 (eIF2) and factor 4F (eIF4F) [26]. Changes in energy charge strongly influence initiation and elongation of protein synthesis. The activity of the protein synthesis machinery in particular depends on the phosphorylation status of initiation factor eIF2 [27-30].

1.2 Posttranslational modifications (PTMs)

A variety of co-traslational modifications and PTMs have been identified in cell-free systems using WGE and RRL in the presence of canine pancreatic microsomal membrane [31]. PTMs strongly affect the physicochemical properties of nascent polypeptide chains, thus influencing protein fold, stability, and activity [32]. From this point of view, it is fundamental to optimize the conditions for CFPS enabling maximum yield of native-like, posttranslationally modified and fully active protein.

In living cells, proteins that are intended for secretion, membrane insertion, or inclusion into the lumen of specific organelles, (such as ER, nucleus, apoplast, lysozyme) are tagged with a specific signal sequence that is responsible for protein translocation. Encoded signal peptide structural motifs are evolutionary conserved, thus signal peptides of different origin can still govern the de novo synthesized protein into the ER to be subjected to further modifications. After the translocation of protein, this sequence is cleaved off by specific enzymes called signal peptidases. Signal peptide cleavage has been observed in RRL and WGE if they were supplemented with canine microsomes [33-37]. Similarly, cleavage of an artificially inserted signal peptide of honey bee origin, the melittin signal peptide (Mel), was observed in insect cell lysates, suggesting that the functional integrity of the entire translation system as well as the translocation machinery is retained [18].

1.2.1 Glycosylation

Glycosylation is one of the most frequent PTMs and it also belongs to the most complex ones [38]. It corresponds to covalent attachment of oligosaccharide chain to -NH2 group of asparagines (N-glycosylation) or to -OH group of serine, threonine, or hydroxylysine (O-glycosylation). The N-glycosylation linkage is performed in ER in a co-translational manner when a core of oligosaccharide unit (Glc3Man9GlcNAc2) is transferred from a lipid donor to a specific asparagin residue [39]. The structure of the initial glycan block is conserved in yeast and plant as well as in animal cells, yet it differs very much in mature proteins, after the process of glycane truncating is terminated in the Golgi aparatus [40]. This probably reflects the effect of the initial glycane block on directing protein folding [41, 42]. The glycoproteins are generally believed to be located only outside of the cells or in the lumen of the organelles, yet it was shown that glycoproteins are present also in the cell cytosol [43-45]. However, their presence in the cytosol is poorly understood. The ability of in vitro systems to produce correctly glycosylated proteins is of high interest for basic and applied research. This is mainly due to the fact that glycoproteins are involved in a number of fundamental physiological processes, e.g. in immunological reactions [46, 47]. In insect cell-based lysates, glycoprotein synthesis was observed, which could be inhibited by tunicamycin, an antibiotic that specifically inhibits the synthesis of N-linked glycoproteins [18].

1.2.2 Phosphorylation, isoprenylation, myristoylation, palmitoylation

Covalent attachment of a phospho group to the amino acid residue of serine, threonine, or tyrosine is broadly used by cells as a mechanism to control the protein's activity. Phosphorylation has been readily observed in cell-free expression using RRL [48, 49] as well as WGE [50]. Other PTMs have been identified in cell-free synthesized proteins, e.g. farnesylation, isoprenylation, and myristylation. All of them are likely to play a role in membrane localization [51, 52]. The anchoring of peripheral membrane proteins to the membrane surface was demonstrated by the formation of a palmitoyl membrane anchor in a cell-free insect system [53].

1.2.3 Membrane proteins

Cell-free expression has emerged as a powerful and highly versatile technique for the production of posttranslationally modified and functional membrane proteins. The open nature of the cell-free system enables the detailed optimization of the expression environment to modulate the folding kinetics for membrane proteins, independent of their size, their species origin, their topology, and function [54]. A crucial step in membrane protein synthesis in eukaryotic cell-free systems is their cotranslational embedding into membranes. Usually, integral membrane proteins are translocated through a protein-conducting channel formed by a conserved heterotrimeric membrane complex termed Sec61 complex [55]. The insertion of a melittin signal sequence is a convenient way to facilitate the translocon-dependent insertion of type I transmembrane proteins into microsomal membranes in cell-free systems [18]. The translocation of de novo synthesized membrane proteins is easy to be followed when proteins are labeled with fluorescent probes. Therefore, integral membrane proteins can be fused to fluorescent proteins such as green fluorescent protein (GFP)–variants. Alternatively, noncanonical fluorescent amino acids can be introduced into protein sequences in cell-free systems [56]. Subsequent detection of proteins using fluorescence techniques can be used to characterize the efficiency of cotranslational translocation into membrane vesicles. Monitoring the quality and quantity of in vitro synthesized and translocated membrane proteins is an essential prerequisite to evaluate the performance of chosen constructs and translation systems.

2 Materials and methods

The expression of target proteins was performed using either pIX3.0 or pIX4.0-based plasmid templates or linear E-PCR products. Synthesis was performed according to template generation procedure described in previous publication [18]. Primer sequences were provided in supporting information. pIX3.0 and pIX4.0-based plasmid templates as well as templates generated by E-PCR were suitable for direct transcription and translation reaction application.

2.1 Lysate preparation

The insect cell extract lysate was derived from S. frugiperda (Sf21 cells). The cells were grown at 27°C in an animal component-free insect cell medium. Sf cells were collected in exponential growth phase by centrifugation at a cell density of approximately 4 × 106 cells/mL and washed with a HEPES-based homogenization buffer consisting of 40 mM HEPES—KOH (pH 7.5) and 100 mM KOAc. The generated Sf cell pellet was resuspended in homogenization buffer to a final cell density of approximately 2 × 108 cells/mL. Resuspended Sf cells were lysed mechanically and the homogenate was centrifuged at 10,000 × g for 10 min at 4°C to spin down the nuclei and debris. The resulting supernatant was applied to a Sephadex G-25 column and fractions with the highest RNA/protein concentrations were pooled. Aliquots of the Sf lysate were immediately frozen using liquid nitrogen and stored at −80°C to preserve maximum activity.

2.2 CFPS

Protein expression was performed in Sf lysates using the linked transcription–translation procedure with a final volume of 50 μL. Transcription reaction mixture (Qiagen EasyXpress Insect Kit II) containing plasmid DNA (final concentration: 60 μg/mL) or E-PCR product (final concentration: 8 μg/mL) was incubated for 2 h at 37°C. Generated mRNA was purified with DyeEx spin columns (Qiagen) according to the manufacturer's instructions and mRNA was added to the translation mixture at a final concentration of 250 μg/mL. The translation mixture further contained 25% (v/v) Sf  lysate, canonical amino acids (200 μM each), ATP (1.75 mM), GTP (0.45 mM), and the radioactive isotope 14C leucine (46.2 dpm/pmol). Random labeling was introduced using BODIPY-TMR-tRNA(Phe) (RiNA GmbH) at a final concentration of 2 μM in absence of 14C leucine. The translation mixture was analyzed after 90 min of incubation at 27°C.

The translation of SecY, E, and G was performed in a repetitive mode in such a way that after each translation, the reaction mixture was fractionated into pellet and supernatant fraction by centrifugation (16 000× g, 4°C, 10 min). The pelleted fraction containing the vesicles with cotranslationally translocated protein was reused for a new translation step using mRNA coding for the next protein of interest. For each new synthesis step, a fresh vesicle-depleted lysate was added (supernatant fraction after centrifugation, 16000× g, 4°C, 10 min).

2.3 SDS-PAGE and autoradiography

Five microliters aliquots of radioactively labeled standard translation mixtures were subjected to standard cold acetone precipitation procedure and left 30 min on ice. Precipitated proteins were collected by centrifugation (10 min at 20 000× g) and pellets were resuspended in 20 μL of 1× LDS sample buffer. Samples of synthesized proteins were analyzed by SDS-PAGE (Life Technologies) using precast 10% NuPAGE® Bis-Tris gels. SDS-PAGE was run at constant 200V for 35 min.

Gels with separated proteins were dried on Whatman paper for 60 min at 70°C (Unigeldryer 3545D, Uniequip) and 14C leucine-labeled proteins were visualized by phosphor imaging technique (Typhoon Trio+, GE Healthcare).

2.4 In-gel fluorescence

The detection of Mel-glyco-eYFP and BODIPY fluorophor-labeled membrane proteins was performed directly on a hydrated SDS-PAGE gel (Stacking gel: 15% Acrylamide, 0.4% Bis, 375 mM Tris (pH 8,8), 0.1% SDS, 0.1% APS, 0.04% TEMED; Resolving gel: 5% acrylamide, 0.13% bisacrylamide, 125 mM Tris (pH 6,8), 0.1% SDS, 0.1% APS, 0.2% TEMED) without staining. Components were purchased from Carl Roth or Merck and gel separation was performed as described in SDS-PAGE and autoradiography section. The gel was incubated for 30 min in a 50% methanol/water solution at room temperature. Fluorescence signals of synthesized proteins were visualized on a UV-table (UV-Systems, Intas) and BODIPY-TMR-tRNA (Phe) incorporation was additionally analyzed in the phosphoimager (Typhoon Trio+, GE Healthcare).

2.5 Protein yield quantification

The yield of 14C-labeled protein was determined by hot trichloroacetic acid (TCA) precipitation followed by radioactive counting. De novo synthesized protein was assayed after an incubation time of 90 min. Five microliters aliquots were withdrawn from the translation reaction mixture and mixed with 3 mL 10% (v/v) TCA containing 2% (w/v) casein hydrolysate. Proteins were left to precipitate for 15 min in a water bath at 80°C, followed by incubation on ice for 30 min. Afterwards, non-incorporated amino acids were removed from the translation mixture by filtration (filtration paper MN GF-3, Macherey-Nagel). Proteins trapped on filter paper were washed twice with 5% (v/v) TCA and two times with acetone. Dry filter papers with retained proteins were transferred into scintillation tubes, mixed with scintillation cocktail (Quicksafe A, Zinsser Analytic), and let gently shaking for at least 1 h on an orbital shaker. Amounts of 14C leucine incorporated into the protein structures were analyzed by LS6500 scintillation counter (Beckman Coulter).

2.6 Deglycosylation assay

Deglycosylation of Mel-glyco-eYFP was probed on 5-μL aliquots of the translation mixture using PNGase F (N-glycosidase F, NEB). The assay was performed according to standard manufacturer's instructions and the output was analyzed by SDS-PAGE followed by in-gel fluorescence section.

2.7 CLSM for fluorescence analysis

Fluorescence imaging was acquired on a LSM 510 (Zeiss) for confocal laser scanning microscopy (CLSM). Translation reaction mix containing the cell-free synthesized eYFP fused protein of interest was diluted twice with 10 μL PBS (phosphate buffered saline without Mg2+ and Ca2+, Biochrom AG) and analyzed on an untreated μ-slide (Ibidi). Excitation was induced at 488 nm by an argon laser and emitted light was detected in the range of 500–550 nm. Fluorescence pictures were depicted by Zeiss LSM Imaging software.

3 Results and discussion

3.1 Evaluation of insect cell lysate

In vitro synthesis of membrane proteins and posttranslationally modified proteins, was investigated in insect cell lysates. These lysates harbor ER-derived microsomal vesicles that offer the possibility to embed membrane proteins in their membranes or store de novo synthesized secretory proteins in their lumen. Additionally, maintained ER reaction conditions in microsomes enable proteins translocated into microsomes to undergo PTMs, e.g. glycosylation. Evaluation of the glycosylation potential of insect cell lysates is exemplarily illustrated on enhanced yellow fluorescent protein (eYFP), designed as a fusion construct with Mel sequence and containing an engineered consensus sequence for glycosylation -N-X-S/T- (Mel-glyco-eYFP). A scheme of engineering a glycoprotein using linear DNA template is shown in Fig.2. Melittin sequence that cotranslationally targets de novo synthesized proteins into microsomal vesicles was inserted at the N-terminus of the protein. Subsequently, cell-free glycoprotein expression, its translocation into the lumen of proteoliposomes, and its glycosylation in various lysate preparations were analyzed. Figure 3A illustrates the distribution of synthesized Mel-glyco-eYFP in translation mixture using CSLM since eYFP serves as a fluorescent probe. It can be seen that fluorescence signals of this protein are localized in the lumen of microsomal vesicles indicating successful translocation of the protein inside the vesicles. The glycosylation of Mel-glyco-eYFP was probed by treatment with endoglycosidase PNGase F and analyzed by SDS-PAGE combined with in-gel fluorescence. Results of deglycosylation assay of Mel-glyco-eYFP synthesized in two insect lysate batches are depicted in Fig.3B. In both cases, a decrease of protein molecular weight (MW) is observed upon the enzymatic treatment, suggesting efficient cleavage of N-linked sugar moieties. Similar glycosylation efficiency found in different lysate batches indicates full activity of the core protein glycosylation enzymes within a homogenous system without the need for supplementing the reaction with additional membrane vesicles.

Figure 2.

Creating synthetic glycoproteins. Expression polymerase chain reaction (E-PCR)-based introduction of an engineered glycosylation site into eYFP. E-PCR products are generated in a two-step amplification procedure, thereby introducing the melittin-signal sequence and a synthetic glycosylation site into the gene of interest. Linear templates are suitable for translation in eukaryotic cell-free systems. Mel, melittin signal peptide (amino acid composition: MKFLVNVALVFMVVYISYIYAD); NXS/T, glycosylation site (amino acid composition: DGELVSNGTVT); PS, protein synthesis; sec. structure, secondary structure (stem loop); green square, glycosylation.

Figure 3.

Glycoprotein synthesis in microsomal vesicles. (A) Localization of in vitro synthesized Mel-glyco-eYFP in microsomal vesicles by CLSM. (B) Mel-glyco-eYFP deglycosylation assay using PNGase F analyzed by SDS-PAGE and visualized by in-gel fluorescence. (−) before PNGase treatment, (+) after PNGase treatment. Glycosylated Mel-glyco-eYFP (G) and deglycosylated variant (D) are indicated by arrows.

3.2 Synthesis of membrane proteins

The high MW membrane protein with one helical transmembrane domain, epidermal growth factor receptor (EGFR), has been synthesized using in vitro systems. Wild-type EGFR is a 134-kDa protein belonging to the family of receptor tyrosine kinases (RTKs). EGFR cooperates in a highly complex manner with other receptors to regulate cell growth. Upon activation by ligand binding, EGFR undergoes dimerization that stimulates its intrinsic tyrosine kinase activity. The consequence is autophosphorylation of several tyrosine residues in C-terminal domain, launching downstream regulation, and signaling by interaction with growth factor receptor-boundprotein 2 (Grb2) adaptor protein. Mutations affecting its expression result in aberrant signaling pathways and are involved in many types of cancers. The EGFR receptor therefore has been classified as oncogene (reviewed in [57]). Up to now, nine different EGFR mutations have been described [58]. The best characterized is the mutation in its extracellular domain resulting in expression of EGFR with truncated amino acid residues 6–237 (vIII deletion) that includes EGF binding site. This form is constitutive active and triggers abnormal signaling leading to non-regulated cell growth and division in lung and breast cancer. Due to its crucial involvement in cancer development, EGFR is an important subject of anticancer therapies. The characterization of this protein and its binding partners can lead to the discovery or design of EGFR antagonist that could impede the misregulated events. Therefore, there is an urgent need to produce EGFR forms in sufficient amounts for functional binding studies. Table1 depicts different constructs of EGFR spanning high MW range from 100 up to 162 kDa, used for in vitro translation. They include wild-type EGFR as well as its fusion with Mel signal peptide (Mel-EGFR). Additionally, fluorescent eYFP has been fused to EGFR (EGFR-eYFP) and to Mel-EGFR (Mel-EGFR-eYFP). Furthermore, DNA constructs of the truncated form of EGFR (deletion vIII) have been generated in both forms, with eYFP (Mel-EGFRtrunc-eYFP) and without eYFP (Mel-EGFRtrunc). In the end, construct of adapter protein Grb2 in its fusion form with eYFP has been prepared (Grb2-eYFP). All proteins were labeled with radioactive 14C Leu to facilitate protein detection. The efficiency of cell-free synthesis using above-described EGFR constructs was analyzed by autoradiography. The results are shown in Fig.4A. Although the quantities of individually expressed proteins differ, they all exhibit a distinct band with the expected apparent MW. EGFR constructs, were produced successfully using in vitro translation in insect cell lysates.

Table 1. Overview of different EGFR constructs used for in vitro translation
ProteinSize (kDa)TMD
  1. TMD, number of transmembrane domains.

EGFR: epidermal growth factor receptor134.31
 (+eYFP) 162.9 
Mel-EGFR: endogenous signal peptide substituted by melittin signal sequence134.61
 (+eYFP) 162.2 
Mel-EGFRtrunc: EGFR with vIII deletion, endogenous signal peptide substituted by melittin signal sequence104.71
 (+eYFP) 132.3 
Grb2: growth factor receptor-bound protein 2(+eYFP) 51.6-
Figure 4.

Cell-free expression of different EGF receptor variants and cytosolic Grb2-YFP adaptor protein. (A) Autoradiogram of 14C leucine-labeled EGFR, Mel-EGFR, Mel-EGFRtrunc, and their YFP fusion constructs. Proteins were separated according to their apparent molecular weight by SDS-Page. (B) Fluorescence signal distribution in different fractions of the in vitro translation system. Cytosolic Grb2-YFP is detected in the supernatant (S), transmembrane Mel-EGFtrunc-YFP is localized in the in the vesicle fraction (V); T = total translation mixture.

Figure 4B shows the results of fluorescence analysis of successfully in vitro expressed cytosolic adapter protein Grb2-eYFP and Mel-EGFRtrunc-eYFP in different sample fractions obtained after the centrifugation of the translation mixture (T). From Fig.4B, the different localization of the two proteins is obvious. Grb2-eYFP as a soluble protein is mostly present in the supernatant fraction (S) while Mel-EGFRtrunc-eYFP, as a membrane protein, fused with melittin sequence governing the translocation into microsomal membrane, is mostly present in the pellet (vesicular fraction, V).

Furthermore, an integral membrane complex composed of subunits SecY, SecE, and SecG (Table2) has been synthesized using two different approaches. SecYEG is part of the prokaryotic translocon complex involved in translocation of bacterial protein into periplasm. The simultaneous synthesis of translocon subunits using all three DNA templates in one batch did not yield satisfactory results since SecG protein was missing in the final product (data not shown). Nevertheless, it is possible to express all subunits in separate reactions as depicted in Fig.5A. The presence of vesicles in insect lysates and the possibility of translocation of proteins into vesicles allows repeated cycles of synthesis using different templates for each translation reaction (repetitive synthesis). In this manner, all three subunits of the translocon were present in the final translation mixture as illustrated in Fig.5B. Repetitive synthesis and storing of subunit proteins in vesicles offer the advantage of progressive protein subunit enrichment in a well-defined set of vesicles.

Table 2. Overview of the membrane-embedded trimetric complex of SecY, SecE, and SecG expressed in this study
ProteinSize (kDa)TMDFunction
  1. TMD, number of transmembrane domains.

  2. Data source: http://www.uniprot.org, downloaded on 28th of March 2012.

SecY48.510Subunits of the protein translocation channel SecYEG
SecE13.63 
SecG11.32 
Figure 5.

Autoradiogram of 14C leucine-labeled SecY, SecE, and SecG. (A) SecY, SecE, and SecG synthesized in separate batches. (B) SecY, SecE, and SecG cell-free synthesized using repetitive protein enrichment in vesicles. T: total translation mixture; S, supernatant fraction; and V, vesicle fraction.

Degeneration of the genetic code enables changes of codon triplets without any change in the encoded amino acid sequence. Different organisms utilize different codon usage, thus codon optimization in autologous manner could lead to a significant increase of protein yield [59-62]. We have investigated the influence of codon optimization on the yield of several proteins spanning a broad range of structurally and functionally divergent transmembrane and intramembrane proteins (Table3). The efficiency of their in vitro synthesis was analyzed by SDS-PAGE followed by in-gel fluorescence analysis (Fig.6). Every membrane protein has been labeled during translation using BODIPY-TMR-tRNA (Phe) that randomly introduces fluorescent probe into protein sequence and allows radioactive-independent protein detection. Each protein has been synthesized using non-optimized and insect codon optimized constructs. As depicted in Fig.6, all cell-free expressed membrane proteins migrate at the expected MW and do not show any sign of fragmentation. It also demonstrates that codon optimization has in some cases led to increased protein yield (SERT, ZIP1, IRK1).

Table 3. Overview of membrane proteins used for random labeling
ProteinSize (kDa)TMDFunctionUniProtKB
  1. TMD, number of transmembrane domains. *, intramembrane protein; Data source: http://www.uniprot.org, downloaded on 28th of March 2012.

SERT: sodium-dependent serotonin transporter71.112Transport of serotoninP31645
ZIP1: zinc transporter35.18Zinc accumulationQ9NY26
IRK1: ATP-sensitive inward rectifier potassium channel 145.62K+ channelP48048
AQP5: Aquaporin 529.16Transport poreP55064
CAV1: Caveolin 121.3*Involved in T-cell activationQ2QLF2
TAP2: Transporter 2 ATP-binding cassette subfamily B (MDR/TAP)78.59ABC transporterQ5HY71
Figure 6.

Fluorescent labeling of cell-free expressed membrane proteins. SERT, ZIP1, IRK1, AQP5, CAV1, and TAP2 wild-type (wt) and codon-optimized (op) variants were expressed in the insect cell-free system in the presence of BODIPY-TMR-tRNA(Phe). Proteins were separated according to their apparent molecular weight by SDS-Page and in-gel fluorescence was analyzed using a laser-based fluorescent gel scanner (Typhoon Trio+, GE Healthcare).

4 Concluding remarks

The continuously growing diversity of options to design and modulate expression conditions in cell free-protein synthesis systems has enhanced in vitro translation as a unique tool to produce a broad range of proteins, including transmembrane proteins, secretory, and cytosolic proteins. Eukaryotic cell-free systems in particular, are designed to express complex membrane proteins as well as glycoproteins. The engineering of N-linked glycosylation sites into the templates for cell-free production facilitates the analysis of synthetic glycoproteins. Multispanning membrane proteins of different size, topology, and function as well as integral components of high molecular weight membrane complexes can be successfully synthesized in eukaryotic cell-free systems. The protein translational machinery in these systems accepts the BODIPY-Dye, a fluorophore with a high fluorescent quantum yield, enabling the sensitive, non-isotopic, fluorescence-based labeling of membrane proteins. Rapid detection of fluorescent bands from nanogram levels of in-vitro-produced membrane proteins facilitates many biotechnological applications including functional analysis of membrane proteins, drug discovery, and mutation screening. A variety of research applications is emerging in the field of CFPS, including rapid screening of mutants by direct translation of PCR products, the study of complex membrane protein assembly processes, and the site-directed labeling of membrane proteins. While cell-free technologies have now been widely accepted for the synthesis of functional membrane proteins, a clear future focus will be on the improvement of production efficiencies of posttranslationally modified membrane proteins for their detailed in vitro pharmacological analysis.

Acknowledgments

We would like to thank Iris Claußnitzer and Birgit Cech (RiNA GmbH) for preparing Bodipy-TMR-tRNA(Phe) and generating expression PCR constructs. This research is supported by the German Ministry of Education and Research (BMBF, No. 0313965A and No. 0312039).

The authors have declared no conflict of interest.

Ancillary