OUP user menu

The insecticidal toxin Makes caterpillars floppy 2 (Mcf2) shows similarity to HrmA, an avirulence protein from a plant pathogen

Nicholas R. Waterfield , Phillip J. Daborn , Andrea J. Dowling , Guowei Yang , Michelle Hares , Richard H. ffrench-Constant
DOI: http://dx.doi.org/10.1016/S0378-1097(03)00846-2 265-270 First published online: 1 December 2003


The Photorhabdus luminescens W14 toxin encoding gene makes caterpillars floppy (mcf) was discovered due to its ability to kill caterpillars when expressed in Escherichia coli. Here we describe a homologue of mcf (renamed as mcf1), termed mcf2, discovered in the same genome. The mcf2 gene predicts another large toxin whose central domain, like Mcf1, also shows limited homology to Clostridium cytotoxin B. However, the N-terminus of Mcf2 shows significant similarity to the type-III secreted effector HrmA from the plant pathogen Pseudomonas syringae and no similarity to the N-terminus of Mcf1. HrmA is a plant avirulence gene whose transient expression in tobacco cells results in cell death. Here we show that E. coli expressing Mcf2 can, like E. coli expressing Mcf1, kill insects. Further, expression of the c-Myc tagged N-terminus of Mcf2, the region showing similarity to HrmA, results in nuclear localisation of the fusion protein and subsequent destruction of transfected mammalian cells. The Mcf1 and Mcf2 toxins therefore belong to a family of high molecular mass toxins, differing at their N-termini, which encode different effector domains.

  • Makes caterpillars floppy 2 toxin
  • HrmA
  • Insecticide
  • Photorhabdus luminescens

1 Introduction

Photorhabdus luminescens is an insect pathogenic bacterium found in association with nematodes that kill insects [1]. Photorhabdus bacteria are released from the gut of their nematode vector directly into the insect blood system whereupon they multiply and kill the insect host by producing a range of toxins [2]. Several of these toxins are high molecular mass and include the insecticidal Toxin complexes or Tcs [3,4] and Makes caterpillars floppy or Mcf [5]. The Mcf toxin (here renamed Mcf1) was discovered in a screen of recombinant Escherichia coli carrying cosmids from P. luminescens strain W14 via its ability to ‘make caterpillars floppy’ following injection of bacterial cultures expressing Mcf1 [5]. The mcf1 open reading frame (ORF) was identified within the cosmid via transposon mutagenesis and subcloning [5]. The Mcf1 protein kills insect hemocytes and causes rapid destruction of the insect midgut epithelium, the primary organ for osmoregulation in insects, causing the caterpillars to lose body turgor and become ‘floppy’[5].

The predicted Mcf1 protein shows little homology to other proteins in current databases, except limited similarity to a region of Clostridium difficile toxin B [6]. The N-terminal domain of several Mcf1 proteins do, however, show similarity to a consensus sequence for a BH3 domain, as revealed by a Prosite search [5]. As proteins that carry only a BH3 domain are proapoptotic [7], we have recently tested the ability of Mcf1 to promote apoptosis in mammalian cells [8]. Cells treated with Mcf1 in the tissue culture media show apoptotic nuclear morphology, DNA laddering, and the presence of cleaved PARP and active caspase-3 after 6 h [8]. These effects are inhibited by the apoptosis inhibitor zVAD-fmk. Transfection of cells with constructs expressing only the N-terminal 1280 amino acids of Mcf1 as a fusion with cMyc also triggered cell destruction. The expressed cMyc–Mcf1 fusion protein was concentrated into the Golgi apparatus of transfected cells prior to cell death [8]. These results support the hypothesis that Mcf1 promotes apoptosis but the precise mechanism of action of the Mcf1 toxin remains unclear.

Here we describe the discovery of an mcf1 homologue, mcf2, within the same P. luminescens W14 genome. The predicted N-terminal domain of Mcf2 lacks a BH3-like motif but instead has a region of similarity to the HrmA type-III secreted effector from the plant pathogen Pseudomonas syringae pv. syringae, which is a plant avirulence protein [9]. Plant cell death induced by recognition of HrmA triggers localised cell death, which limits bacterial infection, and confers avirulence on the virulent bacterium P. syringae pv. tabaci in all tobacco cultivars examined [10]. The avirulence function of hrmA is dependent on the hypersensitive response and pathogenicity (hrp) genes, some of which are involved in the regulation and assembly of the bacterial type-III secretion system or TTSS [911]. HrmA is secreted by the TTSS directly into plant cells and transient expression of HrmA in tobacco cells results in cell death [10]. Here we show that, like Mcf1, expression of Photorhabdus Mcf2 in E. coli is sufficient to kill caterpillars. Further transient expression of the N-terminus, the region showing HrmA similarity, directly inside mammalian cells is sufficient to cause cell death. The potential roles of Mcf1 and Mcf2 during insect infection are discussed.

2 Materials and methods

2.1 Identification and cloning of mcf2

The mcf2 gene was identified during random end-sequencing of a P. luminescens W14 cosmid library. The sequence derived from the mcf2 containing cosmid assembled with mcf1 sequences (Lasergene, Madison, WI, USA) but was clearly different at the nucleotide level. Following identification of the mcf2 gene, the cosmid library was re-screened using polymerase chain reaction (PCR) primers to isolate a cosmid W14-1C-G1 carrying the complete mcf2 ORF. The complete nucleotide sequence of this cosmid was then determined using transposon mutagenesis using an EZ-Tet transposon sequencing primers within the transposon (EpiCenter).

2.2 Insect bioassays

For insect bioassays both mcf1 and mcf2 were cloned into pBAD30. The mcf1 gene was cloned into pBAD30 as a PCR product using the primers 5′-AATATGAGCTCTTGCCTTTGACCCGATCAT-3′ and 5′-ATCAGTCTAGACTAGATGGCCCAAGGCAGC-3′. The mcf2 gene was also cloned into pBAD30 in the same fashion, using the primers 5′-AATATGAGCTCGTAAATACCAAAG-3′ and 5′-ATCAGTCTAGATTAGGACAGCGATG-3′. PCR was conducted with 60°C annealing temperature and a 3-min extension time. The resulting PCR products were cloned using SacI and XbaI restriction sites engineered into the PCR primers (underlined). Expression of Mcf1 and Mcf2 from the resulting clones was confirmed by SDS–PAGE analysis following gel staining with Coomassie brilliant blue. To compare the injectable activity of pBAD30–mcf2 with pBAD30–mcf1 we carried out injections into the waxmoth Galleria mellonella. Bacterial cultures were grown in Luria broth (LB) with 100 µg ml−1 ampicillin and 0.2% glucose, at 30°C overnight. From this overnight culture, the bacterial cells from 1 ml were pelleted and used to inoculate 100 ml of fresh LB containing 100 µg ml−1 ampicillin. When this culture reached an OD600 of 0.6, 0.2% arabinose (w/v) was added and the culture grown a further 3 h at 37°C. 10 µl of the resulting induced culture was then injected into 50 individual G. mellonella larvae. 10 µl contained 1.0×107 colony forming units (CFU) for Mcf1 and 1.5×107 CFU for Mcf2, as determined by plating dilutions onto LB ampicillin agar plates. Mortality, defined as an inability to react to poking with a needle, was scored as regular intervals.

2.3 Mammalian cell culture, transfection and immunofluorescense

To test the effect of making the N-terminus of Mcf2 in mammalian tissue culture cells, the N-terminal fragment of Mcf2 was cloned into pRK5myc. The N-terminus (Mcf2N359) was cloned by PCR using rTth DNA polymerase (Applied Biosystems) using DNA isolated from a cosmid containing full-length Mcf2. Mcf2N359 was generated by PCR using the sense primer AACTAGGATCCATGCCTAGTAACAGCT and TCGTAGAATTCTGAGCCGATTTTCACT as the anti-sense primer. PCR was conducted with 60°C annealing temperature and a 3-min extension time. The resulting PCR product was cloned into pRK5myc (a kind gift of Karen Knox, MRC) using EcoRI and BamHI restriction sites engineered into the PCR primers (underlined). The resulting clone Mcf2N359 was sequenced.

The mammalian cell line NIH 3T3 (Swiss Mouse Fibroblasts) came from the European Collection of Animal Cell Cultures (ECACC, Porton Down, Salisbury, UK). Cells were cultured in Dulbecco's modified Eagle's medium, (DMEM) supplemented with 10% foetal calf serum, 2% (10 ml) 1× penicillin and streptomycin and 1% non-essential amino acids (Sigma) and grown at 37°C, 95% air/5% carbon dioxide (v/v). For cell transfection, pRK5myc was used to express the c-Myc eptiope tag as a fusion protein with the N-terminus of Mcf2. For transfection, NIH 3T3 cells were seeded at a concentration of 1×105 cells ml−1 into six-well plates containing ethanol-sterilised borosilicate glass coverslips (BDH) and vectors co-transfected with an pEGFP-actin (Clontech) using GeneJuice Transfection Reagent (Novagen) according to the manufacturer's instructions.

For immunofluorescense, samples were fixed with 4% paraformaldehyde (w/v) in PBS, permeabilised with 0.2% Triton X-100 and then blocked with 10% normal donkey serum. Cells were stained with primary antibodies to the c-myc epitope (Invitrogen) followed with a Cy3-conjugated secondary antibody (Jackson Laboratories). Fluorescence images of the samples were obtained using a Zeiss LSM-510 confocal laser-scanning microscope (Zeiss LSM-510 system with inverted Axiovert 100M microscope).

3 Results

3.1 Identification of mcf2 by sample sequencing

As part of our ongoing comparative analysis of Photorhabdus genomes, we have been end-sequencing cosmid libraries. One of these end-sequences, from P. luminescens strain W14, was found to predict a section of protein similar to Mcf. We termed this gene mcf2 and correspondingly renamed mcf as mcf1. We re-screened the W14 cosmid library using PCR to find a cosmid, W14-1C-G1, which encompasses all 7167 bp of the mcf2 ORF and sequenced the complete cosmid. The mcf2 gene predicts another large (262 kDa) protein. Like Mcf1, amino acids 1015–1548 of Mcf2 are 39% similar and 20% identical to amino acids 867–1368 of C. difficile toxin B, but Mcf2 lacks the similarity to the C-terminus of an RTX-like toxin from Actinobacillus pleuropneumoniae carried by the C-terminus of Mcf1 (Fig. 1A). Since the original description of the Mcf1 gene, a new entry in Genbank shows that both Mcf1 and Mcf2 also show a second region of homology to an RTX-like cytotoxin (gene VVA1030) from Vibrio vulnificus strain YJ016 (amino acids 695–914 of Mcf2 are 30% identical and 48% similar to amino acids 3250–3496 of VVA1030). The predicted N-terminal domain of Mcf2 is shorter than that of Mcf1 and lacks a BH3-like domain (Fig. 1A). Instead, the N-terminus carries a region with significant similarity (54% similarity and 40% identity) to the C-terminus of the plant avirulence protein HrmA from P. syringae pv. syringae (Fig. 1B). Further examination of unfinished genome sequences shows that Mcf1 and Mcf2 are part of a family of toxins including two undescribed homologues from Pseudomonas fluorescens strain Pf01, ORFs 4315 and 4316, which are 28% identical and 46% similar to each other (Fig. 1A). A comparison of these predicted proteins with the Photorhabdus Mcf1 and Mcf2 proteins shows that the N-terminal domain is variable in this family of proteins, predicting a glycosyltransferase domain in the P. fluorescens proteins. The mcf2 gene lies in an operon adjacent to three genes encoding a type-I secretion apparatus, which is conserved in the two different Photorhabdus strains examined (Fig. 2).


Similarity of Mcf2 to other toxins and to the HrmA plant avirulence gene. A: Predicted structure of the Mcf2 protein compared to Mcf1 and other members of the same toxin family. All these high molecular mass toxins carry a central domain with similarity to C. difficile toxin B but vary at their N- and C-termini. Both Mcf1 and Mcf2 are encoded within the genomes of P. luminescens W14 and TT01. Mcf2 predicts a region within its N-terminus with similarity to the type-III effector HrmA whilst Mcf1 shows similarity to a consensus sequence for a BH3 (pro-apoptotic) domain. TcaB from C. difficile and a novel homologue from P. fluorescens (ORF 4315) both carry a putative glycosyltransferase domain. The C-termini of Mcf1 carries an RTX-like export domain whereas the C-terminus of TcaB carries a receptor-binding domain. Note that P. fluorescens ORF 4316 is not shown for brevity but that it is 28% identical to and 46% similar to ORF 4315 (see text). B: Alignment of part of the N-terminus of Mcf2 with the C-terminus of the HrmA type-III effector protein from P. syringae pv. syringae.


Genomic organisation of the mcf2 locus in the P. luminescens strain W14 and TT01 genomes. Note that the mcf2 gene lies within an operon containing three candidate type-I secretion genes. The double bars in the W14 mcf2 locus mark regions not currently assembled due to the presence of numerous repeats.

3.2 Comparison of the activity of Mcf1 and Mcf2

E. coli carrying the Mcf2 expressing cosmid W14-1C-G1 were injected into Manduca sexta caterpillars, and like E. coli expressing Mcf1, this cosmid also caused caterpillars to die with a ‘floppy’ phenotype (data not shown). To better compare the toxicity of Mcf1 and Mcf2 we cloned both genes into the arabinose-inducible and tightly regulated pBAD30 vector. Previous attempts to clone either gene away from their own native promoters into ‘leaky’ expression plasmids consistently failed, suggesting that both proteins may be detrimental to E. coli. Examination of cultures of E. coli expressing Mcf1 and Mcf2 by SDS–PAGE analysis showed Mcf2 to be a high molecular mass protein migrating faster than Mcf1, as expected (Fig. 3A). However, the Mcf2 protein shows a ladder of bands on the gel, suggesting either that it is degraded or that it may exist in a number of different post-translationally modified forms. Injection of 1.5×107 arabinose-induced mcf2 E. coli into G. mellonella caterpillars resulted in 50% mortality within 70 h (Fig. 3B). Similar results were achieved with 1.0×107 arabinose-induced mcf1 E. coli. Thus induced mcf2 carrying E. coli kill caterpillars at the same speed as E. coli carrying mcf1.


A: SDS–PAGE gel of Mcf1 and Mcf2 expressed in pBAD30 under arabinose induction (asterix in each lane) and stained with Coomassie brilliant blue. Lanes are uninduced (0) and 3 h after induction with arabinose (3). Note that only the high molecular mass proteins are shown. B: Time taken to kill G. mellonella larvae after injection of a fixed dose of Mcf1 and Mcf2 expressing E. coli. G. mellonella larvae were injected with 1.0×107 cells of pBAD–mcf1 and 1.5×107 cells of pBAD–mcf2. At these doses of recombinant bacteria both bacteria killed 50% of the insects in 60–70 h.

3.3 Expression of the Mcf2 N-terminus in transfected cells

Given the similarity of a region of the Mcf2 N-terminus to the plant avirulence protein HrmA, and the fact that transient expression of HrmA in plant cells results in cell death, we tested the ability of the N-terminus of Mcf2 to induce cell death following transfection of mammalian NIH3T3 cells. Expression of the first 359 amino acids of Mcf2 in NIH3T3 cells as a cMyc N-terminal fusion in the plasmid pRK5myc resulted in translocation of the fusion protein to the nucleus (Fig. 4A). Subsequently, expression of the N-terminus of Mcf2 resulted in a collapse of the cell body after 24 h, as revealed by co-transfection with GFP-actin.


Transfection of the N-terminus of Mcf2 into mammalian tissue cultures cells causes cell death. A,B: Cells co-transfected with a construct expressing only the c-Myc tag and an actin-GFP vector show normal cellular morphology, with the c-Myc tag expressed evenly in the cytoplasm. C–F: Cells co-transfected with the N-terminus of Mcf2 and actin-GFP show nuclear localisation of the c-Myc tag (panel C) and subsequent contraction of the cell body (panel E), resulting in cell death.

4 Discussion

The Mcf1 encoding gene was discovered due to its ability to kill insects when expressed in E. coli and injected into caterpillars. Here we have described the identification of a homologue of mcf1, termed mcf2, within the same P. luminescens W14 genome. The predicted amino acid sequence of Mcf2 is similar to Mcf1 but differs at its N-terminus. Comparison with TcaB from C. difficile and a predicted Mcf-like homologue (ORF 4315) from the genome of P. fluorescens suggests that the Mcf proteins may belong to the same super-family as large clostridial toxins. In this family of toxins (Fig. 1A), a central predicted transmembrane domain is conserved whereas both the N-termini and C-termini are variable. The N-terminus appears to encode the putative effector domain in all four proteins: a BH3-like domain in Mcf1, an HrmA-like domain in Mcf2, and a glycosyltransferase domain in both C. difficile TcaB and the novel homologue from P. fluorescens. The C-terminus of Mcf1 contains an RTX-like export domain whereas in TcaB there is a receptor-binding domain. The genomic context of mcf2, in an operon with type-I secretion machinery, suggests secretion of Mcf2 via the type-I pathway. Further, the low levels of expression achieved from cosmid W14-1C-G1, which contains a complete copy of this operon, suggests that the mcf2 gene is regulated by its native promoter in E. coli. It is interesting to note that HrmA is secreted by the type-III pathway in P. syringae pv. syringae and therefore that the HrmA-like domain of Mcf2 may effectively be a fusion of a type-III effector with the high molecular mass Mcf1-like toxin.

Expression of Mcf1 in recombinant E. coli allows the E. coli both to persist within and also to kill caterpillars [5]. It is therefore unexpected that Photorhabdus strains, both W14 and TT01, should require a second potent toxin also possessing the same properties. There are several potential explanations for this observation. First, as insect death is a prerequisite for the lifecycle of Photorhabdus, individual strains may employ ‘functional redundancy’ in toxins. In other words, they encode more than one toxin capable of killing the insect host. Secondly, Mcf1 and Mcf2 may have activities against different types of insects. We have only tested the Mcf1 and Mcf2 toxins against larval lepidoptera (butterflies and moths) and not, for example, coleoptera (beetles). Finally, and alternatively, Mcf1 and Mcf2 may have very different sites of action within the insect. The Mcf1 toxin can destroy the insect gut and hemocytes by triggering massive programmed cell death. However, the site and mode of action of Mcf2 is unknown and is currently under investigation.


This work is supported by a grant from the BBSRC Exploiting Genomics Initiative to R.ff.-C. A.J.D. is supported by a BBSRC CASE studentship with Syngenta, Jealotts Hill. We thank members of the D. Clark and S. Reynolds laboratories at Bath for suggestions and discussion and B. Reaves for the provision of antibodies and practical advice.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
View Abstract