Peptide-mediated delivery of green fluorescent protein into yeasts and bacteria

Abstract

Stringent microbial cell barriers limit the application of many substances in research and therapeutics. Carrier peptides that penetrate or translocate across cell membranes may help overcome this problem. To assess peptide-mediated delivery into two yeast and three bacterial species, a range of cell penetrating and signal peptide sequences were fused to green fluorescent protein (GFP), expressed in Escherichia coli, partially purified and incubated with growing cells. Fluorescence microscopy indicated several peptides that mediated delivery. In particular, VLTNENPFSDP efficiently delivered GFP into Candida albicans and Staphylococcus aureus, while YKKSNNPFSD was most efficient for Bacillus subtilis and CFFKDEL for Escherichia coli. Carrier peptides may improve delivery of certain large molecular mass molecules into microorganisms for research and therapeutic applications.

Signal peptide, Candida albicans, Saccharomyces cerevisiae, Staphylococcus aureus, Bacillus subtilis, Escherichia coli

1 Introduction

Biological membranes are generally impermeable to large hydrophilic molecules, and this limits the use of many otherwise useful compounds. The methods used today to deliver bioactive macro-molecules into cells are: microinjection, electroporation and lipid-mediated uptake 1,2]. Unfortunately, these procedures often suffer from restricted species range, poor efficiency and toxic effects[3]. To overcome these limitations, one strategy is to develop peptides that can carry large molecules into cells. Such peptides have been pursued mainly for eukaryotic cells, but there also appears to be possibilities for microbial applications.

Cellular uptake experiments with cell penetrating peptides in eukaryotic systems demonstrated that peptides have a surprising capacity to deliver molecules into cells. For example, a 16-mer peptide derived from the Antennapedia protein homeodomain[4] and a 13-mer peptide from the HIV-1 Tat (transcriptional activator protein) protein[3] can deliver a range of molecules, even large proteins such as β-galactosidase, horseradish peroxidase or the Fab antibody fragment [5,,,8]. Also, antisense oligonucleotide delivery was significantly improved by coupling to transportan and other cell penetrating peptides [9,,11].

Microbial cell barriers stringently restrict foreign molecules, and peptides appear to be too large for passive uptake. Nevertheless, there is evidence for both receptor-independent and receptor-dependent delivery routes in microorganisms, and these mechanisms can be exploited to deliver foreign molecules. The first example of peptide-mediated delivery into microorganisms involved substrates for peptide permease transport mechanisms used to deliver large antimicrobials 10,12]. Recently, cell permeabilizing peptides were used to deliver antisense agents into bacteria[13]. In addition many delivery signal sequences from a classical receptor-mediated uptake mechanisms have been identified [14,,,17]. For example several well-defined peptides act as endocytosis signals for transport into the cytoplasm[18]. Therefore, there may be many carrier peptides available for microbial applications.

Here we aimed to assess several cell permeating and signal peptides for their potential to deliver foreign molecules into microorganisms. Selected peptide sequences were fused with the N-terminus of green fluorescence protein (GFP), expressed in Escherichia coli and partially purified for uptake studies. Several peptides, particularly signal peptides, increased the uptake of GFP into growing microorganisms, as seen by increased intracellular fluorescence.

2 Materials and methods

2.1 Strains and culture conditions

Candida albicans and Saccharomyces cerevisiae were grown in YPD medium (1% yeast extract, 2% bacterial peptone and 2% glucose) or minimal medium with or without uracil (25 mg l⁻¹) at 30°C with shaking at 200 rev min⁻¹. Staphylococcus aureus (RN4220), Bacillus subtilis (168) and E. coli (K-12) strains were grown in Mueller–Hinton (MH) broth (Sigma) at 37°C with shaking at 225 rev min⁻¹. Prior to use, overnight cultures were pelletted by centrifugation at 3000×g, rinsed and suspended in 1×PBS buffer.

2.2 Plasmid construction of peptide–GFP fusion proteins

Plasmid pMN406 containing the gfp gene was a generous gift from Dr. Michael Niederweis, Institut fur Mikrobiologie, Biochemie und Genetik, Germany. Restriction enzymes, T4 DNA ligase (Amersham Parmacia) and Taq DNA polymerase (Clontech) were used in accordance to the manufacturer's recommendations.

A peptide–GFP fusion was first constructed by PCR amplification of the sequence encoding GFP by using an upstream primer (5′-CG GAA TTC ATG GTG CTG ACC AAC GAA AAC CCG TTT TCT GAT CCC GGG TCG AAG GGC GAG GAG-3′), which provided an Eco RI site, a Sma I site and the 11-amino acid signal sequence (VLTNENPFSDP) and a downstream primer (5′-TCT CGG CTC GAT GAT CC-3′), which included a Hin dIII site. The 752-bp fragment was amplified by PCR using Taq polymerase (Clontech) and ligated into the pKK223–3 expression vector (Pharmacia Biotech). All remaining peptide–GFP constructions were obtained by cloning annealed and digested oligonucleotides into expression vector pKK233-3–GFP using the Eco RI and Sma I restriction sites (Fig. 1). The oligonucleotide sequences used in this study are listed in Fig. 1. The introduction of the signal peptide sequences were confirmed by Kpn I digestion site at a site introduced within the inserted sequences.

1

Structure of the expression vector used to produce peptide–GFP fusions. The figure shows a schematic diagram of the GFP expression plasmid and cloning strategies. The sequence encoding the peptide–GFP is under the control of a Ptac promoter followed by the sequence for the carrier peptide and gfp gene, with the Eco RI and Sma I restriction sites flanking the sequence for the peptide.

2.3 GFP expression and partial purification

The GFP protein was partially purified as described[19]. Bacterial cells expressing GFP or peptide–GFP were pelletted and suspended in Tris buffer (20 mM Tris–HCl, pH 8.0, 150 mM NaCl, 5 mM EDTA). The cells were disrupted by sonication on an ice-water bath and the insoluble material was removed by centrifugation at 4°C for 15 min at 16 000×g. Triethanolamine base and dry ammonium persulfate were added to the bright green supernatant to a final concentration of 100 mM and 1.6 M respectively. After 1 h of incubation on ice, precipitated proteins were removed by centrifugation for 20 min at 6000×g. Dry ammonium persulfate was added to a final concentration of 70% saturation. After 1 h, the peptide–GFP fusion proteins were collected by centrifugation as above and dialyzed against 1×PBS buffer. Protein concentration was estimated by using Bio-Rad Protein Assay Dye binding reagent.

2.4 Fluorescence microscopy and image analysis

Yeast cells were suspended in minimal medium supplemented with 25 mg l⁻¹ uracil and 5 mM MgCl₂. S. aureus, B. subtilis and E. coli cells were suspended in MH broth supplemented with 0.5% glucose and 5 mM MgCl₂. Cells were incubated with peptide–GFP (10 μg ml⁻¹ total protein) at 30°C for 1–3 h. After incubation, cells were centrifuged at 1500×g for 5 min. The pelleted cells were rinsed with 1×PBS buffer to remove extracellular GFP and background fluorescence. The cell pellets were suspended in 1×PBS buffer and fluorescence microscopy was performed using a Leica DMRXA microscope with a cooled frame CCD camera. GFP was visualized by using excitation and emission wavelengths of 488 and 507 nm respectively and the images were processed using Openlab™ software.

3 Results and discussion

3.1 Rationale

We aimed to screen a range of peptide sequences as potential carriers for the delivery of foreign substances into microorganisms. All selected peptides are proposed to enter cells either by cell permeabilization, endocytosis or membrane transport (Table 1, and references therein). Several of the selected peptides are known to carry foreign proteins and other molecules into mammalian cells, but none had been tested previously for their capacity to carry cargo molecules into microorganisms (Table 1). As there are many possible carrier peptides available, we first aimed to establish a convenient approach to assess the capacity for peptides to carry foreign molecules into several microbial species. Sequences encoding selected peptides were cloned as a cassette to provide an N-terminal fusion to the GFP. Fig. 1 illustrates the cloning strategy, inserted DNA sequences and the encoded peptides. The results indicated that a range of carrier peptides could be expressed as fusions with GFP and used directly for uptake studies in microbial systems. All fusions exhibited strong green fluorescence by fluorescence microscopy, although there were variations in expression efficiency or fluorescence between the different peptide–GFP fusion proteins. An important feature of this approach is that only partial purification of the peptide–GFP was required prior to uptake studies. Also, the approach lends itself to comparative analyses of a large number of carrier peptides in a range of cell types without the costs and complications associated with organic synthesis or column purification.

1

Peptide sequences fused to GFP

Carrier peptide	Charge	Native function or feature	Reference
Endocytosis
VLTNENPFSDP	2−	Receptor recycling signal	[20]
YKKSNNPFSD	1+	EH domain affinity	[25]
RSNNPFRAR	3+	EH domain affinity	[25]
AKTNPFRQQ	2+	EH domain affinity	[25]
CMVSCAMPNPF	0	Vesicle recycling signal	[26]
LLDLMD	2−	Receptor recycling signal	[27]
LMDLAD	2−	Receptor recycling signal	[27]
Non-endocytosis
RQIKIWFQNRRMKWKKa	7+	Nuclear localization	[28]
YGRKKRRQRRRb	8+	Nuclear localization	[28]
CKGGAKL	2+	Peroxisomes localization	[29]
CFFKDEL	1−	Endoplasmic reticulum localization	[29]
GASDYQRLGC	0	Trans-Golgi network localization	[29]

3.2 Peptide-mediated delivery of GFP into yeasts

Cells were incubated with free GFP and the peptide–GFP fusion proteins at 30°C for 3 h in minimal medium supplemented with 25 mg l⁻¹ uracil and 5 mM MgCl₂. Following incubation, cells were rinsed and examined for intracellular GFP distribution using fluorescence microscopy. As can be seen in Fig. 2, cells treated with free GFP showed only weak fluorescence. In contrast, several peptide–GFP fusion proteins appeared to accumulate within cells. The two yeast species used in this study, C. albicans and S. cerevisiae, are closely related, however, a somewhat different pattern of delivery was observed with the different peptide–GFP constructs. For C. albicans, the signal peptides VLTNENPFSDP, LLDLMD and LMDLAD that contain the NPF or LMD motif showed efficient delivery of GFP protein into cells, while for S. cerevisiae the peptides YKKSNNPFSD, RSNNPFRAR and GASDYQRLGC were most efficient. The results indicate that the signal peptides were most efficient as carriers for GFP delivery into yeasts (Fig. 2; Table 1). Also, delivery was more efficient with C. albicans relative to S. cerevisiae.

2

Microscopic analyses of peptide–GFP fusion protein internalization into yeasts. Following overnight growth, C. albicans and S. cerevisiae cells were harvested and incubated with different peptide–GFP fusions and control GFP (10 μg ml⁻¹ total protein) for 3 h. The cells were harvested, rinsed and examined for GFP fluorescence (total magnification 1000×); the images were captured using a CCD camera.

3.3 Peptide-mediated delivery of GFP into bacteria

Following the encouraging results obtained for peptide-mediated delivery into yeasts, the study was extended to assess peptide–GFP uptake by bacteria. Peptides have been used to carry cargo molecules into bacteria previously, as discussed above, however we are not aware of attempts to use the peptides described here for bacterial applications, and we know of no attempts to deliver proteins into bacteria using carrier peptides. Similar to the experiments carried out with yeast, freshly grown S. aureus, B. subtilis and E. coli were incubated with free GFP and peptide–GFP fusion proteins at 30°C for 1 h in MH broth supplemented with 0.5% glucose and 5 mM MgCl₂. Following incubation, cells were rinsed and examined for GFP distribution using fluorescence microscopy. The results illustrated in Fig. 3 showed only weak fluorescence in cells treated with free GFP. In contrast, bacteria treated with several of the peptide–GFP fusion proteins exhibited strong intracellular fluorescence. For S. aureus, the peptides VLTNENPFSDP and RSNNPFRAR efficiently delivered GFP, while YKKSNNPFSD was most efficient for B. subtilis and CFFKDEL was most efficient for E. coli.

3

Microscopic analyses of peptide–GFP fusion proteins internalization into S. aureus, B. subtilis and E. coli. Following overnight growth cells were harvested and incubated with different peptide–GFPs and with control GFP (10 μg ml⁻¹ total protein) for 1 h. The cells were harvested, rinsed and examined for GFP fluorescence (total magnification 1000×); the images were captured using a CCD camera.

3.4 Carrier peptide-mediated delivery mechanisms?

The peptides selected for this study are natural sequences known to be involved in cell uptake or membrane transport. Two of the peptides, Tat and homeotic protein Antennapedia, have been evaluated as carrier peptides for mammalian cells. Table 1 lists the selected peptides according to their native function or feature and indicates their putative mechanism for membrane transport.

The results for the NPFSD-containing peptide are consistent with a previous study showing that the NPF motif, a highly conserved cytoplasmic domain of several receptors, is involved in clathrin-dependent endocytosis[20]. In the present study the NPFSD-containing peptide mediated efficient internalization into S. cerevisiae and C. albicans, possibly through a conserved mechanism. In S. cerevisiae moderate internalization of GFP occurred with signal peptides containing the NPF motif. Also moderate uptake was observed with peptides having affinity to the Eps homology (EH) domain. One interesting feature of these peptides is that they contain aromatic amino acids, which are often present within internalized peptide domains, including most EH domains 21,22] and in the cytoplasmic domain of the α-factor pheromone receptor that internalizes the α-factor[23].

The peptides selected for this study are of eukaryotic origin and it is difficult to explain their capacity to deliver GFP into bacteria. The peptides that delivered GFP into bacteria are endocytosis or organelle localization signals that would seem irrelevant in bacteria. Nevertheless, there are possible explanations for our observations. One possibility is that the peptide–GFP may enter cells passively. However, this seems unlikely, as these peptides lack features common to bacterial permeabilizing peptides[24]. A second possibility is that the peptides may have affinity for microbial cell wall structures and following surface binding they could be internalized through the process of cell wall or membrane protein recycling.

Although the uptake mechanisms involved in this study are unclear, the carrier peptides that delivered GFP into microorganisms appear to enter cells actively rather than through passive diffusions. First, the permeable strain of S. cerevisiae, which is less restrictive to passive uptake of a range of large molecular mass substances, did not show greater peptide–GFP uptake. Second, efficient internalization was both energy- and time-dependent. Therefore, the results point to active receptor-based uptake, although questions regarding mechanism of uptake, peptide structure requirements, possible cargo molecules and delivery kinetics remain to be addressed in further studies.

Acknowledgements

This research was supported by the Swedish Science Council and Pharmacia Corporation. The authors would like to thank Michael Niederweis for providing plasmids.