Fused Fluorescent Tags Are Useful Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Fused fluorescent tags are usefull to determine protein localization in cell biology. However, occasionally fused protein constructs disturb topology and function of the gene of interest making alternative methods helpful to complement fluorescence-tagged protein analysis.

> We have applied a modified cytosolic yeast-two-hybrid Sos-recruitment system (SRS) to test the potential membrane localization and the orientation of the PEPINO/ PASTICCINO2 protein.

> Alternative N-terminal and C-terminal fusions of the human Son-of-sevenless (hSos, orthologous to yeast cdc25) gene to PEP/PAS2 both rescue yeast temperature sensitive cdc25 mutants at restrictive temperature conditions.

> The data suggest that the N- and C-terminus respectively resides on the cytosolic side of the plasma membrane. This implies that the first trans-membrane domain (Tm) harbouring a signal peptide is not cleaved off and that an anti-phosphatase pocket domain is oriented towards the outer lumen. As a further corollary, PEP/PAS2 has an even Tm number confirming algorithms predicting four strong Tms. Most notably, according to this model, the localization of four amino acids, which constitute essential core elements of the 3-hydroxyacyl-CoA dehydratase activity of yeast PHS1 have the identical membrane position and orientation in its plant ortholog PEP/PAS2.

Key words: Anti-phosphatase, Arabidopsis, cell proliferation, cotyledons, embryo, 3-hydroxyacyl-CoA dehydratase, lipid biosynthesis, plant organogenesis, PEPINO/PASTICCINO2.


The Yeast-Two-Hybrid (Y2H) Sytem is a versatile tool for investigating protein interactions. The classical Y2H system analyses the interaction in the environment of the cell nucleus (Fields and Song, 1989). Though very powerful, this sytem is biased against certain classes of proteins as for instance, transcriptional activators and repressors. In addition, some cytoplasmic and membrane-associated proteins might not adopt the correct conformation due to lack of modifications, which are performed in the cytoplasm but not in the nucleus. Also, some proteins might be toxic to yeast in the conventional Y2H system.

Several alternative systems have been developed to overcome these limitations, examples are the split-ubiquitin system (Johnsson and Varshavsky, 1994; Laser et al., 2000; Möckli et al., 2007), the expressed transactivator system (Hirst et al., 2001), the G-protein fusion assay (Ehrhard et al., 2000), the Sos-recruitment system (SRS) and variants of it including the (reverse) Ras-recruitment (RRS) system (Aronheim et al., 1997; Broder et al., 1998; Hubsman et al., 2001; Kruse et al., 2006). SRS and RRS use the activation of the mitogenic Ras pathway in S. cerevisiae as a selection mechanism. In this signaling pathway the key proteins CDC25 (the yeast homolog of human Son-of-sevenless gene, hSos) and Ras need to be localised at the membrane in order to relay the mitogenic signal (Fig.1A). The guanyl nucleotide exchange factor (GEF) CDC25 promotes the activation of Ras by GDT/GTP exchange. Yeast CDC25 can be replaced by human Sos (hSos), provided that hSos is localized at the membrane. When hSos is expressed in a temperature-sensitive yeast cdc25 mutant and successfully localized at the membrane, tha activity of hSos permits the growth at the restrictive temperature of 37°C and rescues the strain. In the SRS membrane localisation of hSos is achieved by protei-protein interaction. The protein of interest (bait) is fused to hSos in the pSOS-vector whereas the target protein (prey) is fused to a membrane localisation signal (e.g. a myristoilation signal) encoded in the pMYR-vector. In case of interaction hSos is localised to the membrane and activates yeast Ras. A yeast strain with a temperature sensitive cdc25-2 mutation will then be capable to grow even under restrictive temperatures. SRS (and RRS) have been successfully used to analyze protein interactions in humans, animals, yeast and recently in plants (Kruse et al., 2006; Kim et al., 2006; Frischmuth et al., 2004).

The presented work explores the applicability of the SRS to analyse protein (integral) membrane association by using only one part of this system (Fig.1B). We followed the observation that artificial targeting of Sos to the plasma membrane, by fusing it to myristoylation or farnesylation sequences, results in sustained receptor-independent Ras activity in cells (Aronheim et al., 1994; Gureasko et al., 2008). We have used SRS to analyse the Arabidopsis PEPINO/PASTICCINO2 protein. This gene has been shown to be essential for coordinated proliferation during development and for fertility in Arabidopsis thaliana (Haberer et al., 2002; Bellec et al., 2002). Subsequent analysis showed that this gene is involved in disparate molecular processes, as interactor of CDKs in cell cycle control and as 3-hydroxyacyl-CoA dehydratase involved in fatty acid synthesis (Da Costa et al., 2006; Bach et al., 2008). It has also been demonstrated, that mutations in PHS1/YJL097w, the PEP/PAS2 homolog of yeast, can be rescued by the plant protein (Bellec et al., 2002). The PEP-sequence was predicted to harbour highly conserved domains with plant and animal proteins with significant trans-membrane spanning domains (Haberer et al., 2002). However, PEP/PAS2-GFP/YFP fusion proteins have shown quite different cellular localizations including diffuse cytosolic distribution in Arabidopsis thaliana cells (Da Costa et al., 2006; Bach et al., 2008).

Here, we show the ability of N-terminal and C-terminal fusions of Arabidopsis thaliana PEPINO/PASTICCINO2 with hSos to rescue cdc25H mutant S. cerevisiae at restrictive temperature conditions. According to these data PEP/PAS2 has a cytosolic N- and C-terminus and an even number, probably four, of trans-membrane domains (Tms). Although evolutionary different to the yeast ortholog PHS1, which has six Tms, in PEP/PAS2 seven distributed core functional amino acids have the identical position and orientation in/at the membrane respectively. This demonstrates the applicability of the Sos-recruitment system to test for membrane localisation in cases where protein-GFP fusions might be critical in showing correct localization, orientation and topology respectively. It also enables to test the potential outer- or intracellular facing of N-terminal and C-terminal domains of transmembrane proteins and to show whether they have an even or uneven number of transmembrane domains respectively.

Results and Discussion

This work aimed to obtain information on PEP/PAS2 localization and topology using an alternative approach to GFP-tagged protein localization analysis. The SRS used takes advantage of Ras-activation by membrane localized Sos (SRS) in yeast (Aronheim et al., 1997; Kruse et al., 2006). Recently, SRS has also been applied to assess protein interaction in plants, e. g. Arabidopsis tonoplast proteins TIP1 and 2 vs. CMV1 replication protein and movement protein BC1 of Abutilon with itself respectively (Kim et al., 2006; Frischmuth et al., 2004). PEP/PAS2 should be particularly well suited for this test since it is capable to rescue mutations in its yeast homolog YJL097w/PHS1 (Bellec et al., 2002).

PEP/PAS2-Sos constructs recue cdc25H yeast cells under restrictive temperature conditions

For transformation yeast cdc25H cells were grown on plates with SD medium including glucose and lacking leucine. The constructs pSOS-PEP-NcoI/SacI as well as pSOS-PEP-HindIII both led to colonies under restrictive temperature (Fig. 2). The transformation experiments were repeated for each construct with independently raised yeast competent cells (Table 1, Fig.2). The generation of competent cdc25H yeast cells is critical and prone for reversion of the temperature sensitive point mutation E1328K (Petitjean et al., 1990). Therefore as a precaution, aliquots of cells used for transformation were always assessed for possible revertants in the preculture (see Materials and Methods). Most clones carrying PEP-SOS fusion constructs and growing under restrictive temperature truned out to be non-revertant. Successfully rescued cells, i. e. colonies with functional membrane-localized Sos and thus capable to grow under restrictive temperature, should still possess the cdc25H temperature sensitive point mutation. Consequently, PCR analysis using alternative primer pairs produced the amplification product of the mutant or the wild-type allele of CDC25 respectively (Fig. 3). This confirmed 39/40 clones tested to be non-revertants rescued by the PEP-SOS fusion protein (about 20 clones tested for each construct). Selected clones were further confirmed to be non-revertant by sequence analysis.

Notwithstanding these results, it was conspicuous that the number of colonies obtained under restrictive temperature is much lower than under permissive temperature although the selection marker on the pSOS plasmid vector allows transformed yeast cells to grow (Table1). This probably reflects an intrinsic property of the system, depending on critical requirements as for instance a sufficient number of correctly folded and membrane-integrated fusion proteins in each cell. We therefore transformed aliquots of the same yeast competent batches transformed with pSos-PEP constructs, with the complete Sos-recruitment system. Here, two different vectors are used one carrying the hSos gene (pSOS) and a second carrying the myristoilation peptide sequence (pMYR). In pSOS, the gene hSos is driven by the constitutive ADH promotor. In addition pSOS carries the LEU gene. Under permissive conditions (25°C) yeast cells harbouring pSOS grow on SD medium including glucose but lacking leucine. In pMYR, the myristoilation peptide gene (and the fused gene of interest) is driven by the GAL promotor. In addition pMYR carries the URA3 gene. The GAL promotor is activated in the presence of galactose but is repressed in the presence of glucose. Thus, under permissive conditions (25°C) yeast cells harbouring pMYR grow on SD medium including galactose but lacking uracil. The cells do not grow if the medium contains glucose. Under restrictive conditions (37°C) cdc25H yeast cells grow only if hSos is localised to the plasma membrane. In the complete Sos recruitment system this is the case when the prey protein fused to the myristoilation signal interacts with the bait protein fused to the hSos protein. Successfully co-transformed cdc25H yeast cells, with interacting hSos-prey protein, are able to grow at 37°C on galactose but not glucose medium.

Repeated tests were performed in parallel to the mentioned experiments using the corresponding yeats cells and the positive vs. negative control protein fusion constructs pSOSMAFB, pMYRMAFB, pSOSCollagenase and pMYRLaminC respectively. These constructs were tested under permissive and restrictive temperature on glucose and galactose media, which always lacked leucine and uracil respectively (Fig. 4). Only the pair of constructs carrying MAFB, a bZIP transcription factor gene should enable growth under restrictive temperature since MAFB forms homodimers (Kataoka et al., 1994). This was actually the case and the transformation results showed a lower number of surviving cells under restrictive temperature as compared to the transformation under permissive temperature (Fig. 4). The frequencies were comparable to those in the experiments with single pSOS-PEP constructs supporting the assumption that post-transcriptional demands on correct protein modification, folding and/or integration respectively have to be fulfilled regardless wether Sos is associated to the membrane by direct fusion to a protein with transmembrane domains or indirectly through bait-prey interaction with the prey being myristoilated (Table 1).

PEP/PAS2 has an even number of Tms with cytosolic amino- and carboxy-termini

The PEP/PAS2 protein sequence was analysed with the aid of the web-tool aramemnon (htpp://aramemnon.botanik.uni-koeln.de/). This tool performs analyses of proteins for transmembrane domains with a dozen algorithms. The prediction from each analysis is given as well as a consensus calculated from all combined results with a probability score for each predicted transmembrane domain. The „rigidity" of the algorithm may affect the number of predicted transmembrane spans. For instance, the version TmHMM_v1 predicts five and the version TmHMM_v2 predicts four transmembrane domains respectively. The consensus of all analyses indicated five possible transmembrane domains in PEP with different probability scores (Fig. 5). The domains span the aminoacid no. 10 to 32, 51 to 73, 100 to 121, 140 to 162 and 179 to 201. The probability scores are relatively high except that for the second possible transmembrane span (score 0.2). The first predicted domain harbours a signal peptide, which according to a further Aramemnon tool (seventeen compared algorhythms) most likely directs the protein to a secretory pathway (the scores for subcellular localisation given, are 0.0 for chloroplast, 8.1 for mitochondrion and 20.6 for secretory pathway).

Both the N-terminal and C-terminal fusion of Sos to PEP succeeded in rescuing thermo-sensitive cdcd25H yeast cells. This fact indicates that both termini of PEP potentially face the cytosolic side of the cell (Fig. 6). As a corollary of this, PEP should possess an even number of transmembrane domains. From the five Tm-domain scores predicted by the Aramemnon algorhythms the score for the second Tm-domain is significantly weak (Fig. 5). Together with the transformation data obtained, this favours a model where this latter Tm is not realized leading to a membrane spanning protein with four Tms (Fig. 6). This has several consequences for the understanding and further biochemical analysis of PEP/PAS2, some of which become evident by comparing PEP/PAS2 to PHS1.

There are some obvious differences to the yeast homolog but also some intriguing similarities. PHS1 in yeast has also cytosolic N- and C-termini and an even number of transmembrane domains. However, it has been assigned to have six Tms (Kihara et al., 2008). This might be due to evolutionary divergence, which is also reflected by the altered phosphatase-like domain in S. cerevisiae. In PEP/PAS2 this domain is identical to the anti-phosphatase signature motif in STYX (Wishart et al., 1995; Wishart and Dixon, 1998). Although the name suggests an antagonistic function to phosphatases this domain could also have a different function as a docking motif to associate with other proteins (Hunter, 1998). In any case, a point mutation in this motif strongly indicated a functional role in Arabidopsis thaliana (Haberer et al., 2002). The corresponding region in yeast is partly integrated into the membrane and oriented towards the cytosol whereas according to the described model in Arabidopsis this domain is localized in the luminal space. This does not exclude that this domain associates with the membrane since it is weakly hydrophobic. Alternatively, a similar orientation to yeast PHS1 could only be achieved if one would assume that the third Tm in the Aramemnon consensus is not realized in favour of the much weaker second Tm. The PEP/PAS2 region with the last two Tms exhibit striking topological similarities in comparison to PHS1. First, the functionally relevant catalytic amino acids in the fifth Tm of the yeast protein, Arg-141, Tyr-149, Gly-152 and Glu-156 have identical positions and orientation. Arg-141 is localized at the cytosol to membrane border while Tyr-149, Gly-152 and Glu-156 respectively are embedded within the lipid bilayer. Second, the same is also true for the highly conserved Pro-189 (in the sixth Tm in yeast). Third, the highly conserved Gln-201 and Arg-202 at the very C-terminus (Kihara et al., 2008) are also localized in the cytosol in PEP/PAS2 (Fig. 6).

There are two further corollaries of this data. First, they suggest that the N-terminal domain is not cleaved off. This is not unlikely since it is known that not all signal peptides are separated from their original proteins as for instance shown for phytohemagglutinin (PHA), the major seed lectin in common bean, alpha-zein in the maize floury2 mutant, legumin-1 in maize endosperm (Tague and Chrispeels,1987; Coleman et al., 1995; Yamagata et al., 2003). Second, as mentioned before in case where the signal peptide (amino acids 10 to 32) enters the lipid bilayer the following anti-phosphatase domain (amino acid 45 and amino acids 64 to 72) must completely or at least partly reside on the non-cytosolic side of the membrane. It is not clear how and whether this relates to any of the assigned molecular functions of PEP/PAS2 since this impedes direct interaction with cytosolic components (however see alternative model mentioned above). Together this fosters the proposed membrane localization in accordance with the 3-hydroxyacyl-CoA dehydratase function of PEP/PAS2 while it less likely suggests a function, which would require a direct interaction with components of the cell cycle machinery. However, PEPINO/ PASTICCINO2 has been shown to have fundamental influence on plant development with direct or indirect effects on cell proliferation processes (Haberer et al., 2002; Bellec et al., 2002). This implies a further as yet (in detail) unknown molecular participation of PEP/PAS2 in Arabidopsis development.

The presented results show that the Sos -recruitment system can be a valuable tool for determining potential protein membrane localisation in a case where, due to incompatibilities in GFP-fusions or due to multiple subcellular localizations, it is difficult to unambiguously show the membrane localisation and orientation of a protein respectively. In case of PEP, further more detailed biochemical analyses are necessary to analyze its Tm-domains. However, SRS has provided a valuable model of the PEP/PAS2-membrane localization providing valuable targets by further tests.