OUP user menu

A giant family of short palindromic sequences in Stenotrophomonas maltophilia

Francesco Rocco , Eliana De Gregorio , Pier Paolo Di Nocera
DOI: http://dx.doi.org/10.1111/j.1574-6968.2010.02010.x 185-192 First published online: 1 July 2010

Abstract

The genome of Stenotrophomonas maltophilia is peppered with palindromic elements called SMAG (Stenotrophomonas maltophilia GTAG) because they carry at one terminus the tetranucleotide GTAG. The repeats are species-specific variants of the superfamily of repetitive extragenic palindromes (REPs), DNA sequences spread in the intergenic space in many prokaryotic genomes. The genomic organization and the functional features of SMAG elements are described herein. A total of 1650 SMAG elements were identified in the genome of the S. maltophilia K279a strain. The elements are 22–25 bp in size, and can be sorted into five distinct major subfamilies because they have different stem and loop sequences. One fifth of the SMAG family is comprised of single units, 2/5 of elements located at a close distance from each other and 2/5 of elements grouped in tandem arrays of variable lengths. Altogether, SMAGs and intermingled DNA occupy 13% of the intergenic space, and make up 1.4% of the chromosome. Hundreds of genes are immediately flanked by SMAGs, and the level of expression of many may be influenced by the folding of the repeats in the mRNA. Expression analyses suggested that SMAGs function as RNA control sequences, either stabilizing upstream transcripts or favoring their degradation.

Keywords
  • repeated DNA sequences
  • palindromic DNA
  • stem-loop structures
  • whole-genome analysis
  • RNA hairpins
  • microbiological diagnostic

Introduction

Stenotrophomonas maltophilia is a nonfermentative Gram-negative bacterium that is ubiquitous in nature. It constitutes one of the dominant rhizosphere inhabitants (Ryan et al., 2009; Taghavi et al., 2009), but is also increasingly being described as an important nosocomial pathogen in debilitated and immunodeficient patients, and has been associated with a broad spectrum of clinical syndromes. It has been isolated frequently from cystic fibrosis patients, and has emerged as a serious pathogen in cancer patients (Looney et al., 2009). Stenotrophomonas maltophilia displays an intrinsic resistance to many antibiotics, making the selection of optimal therapy difficult (Crossman et al., 2008). Whether the bacterium is a mere colonizer or an infectious agent often remains unresolved, and virulence factors are still ill-defined. The chromosomes of the clinical K279a (Crossman et al., 2008) and the environmental R551-3 (Taghavi et al., 2009) strains exhibit extensive synteny, but each is punctuated by about 40 different GEIs or genomic islands (Rocco et al., 2009). Whether pathogenicity may be associated in part with the maintenance of specific GEIs in the S. maltophilia population remains to be established.

Stenotrophomonas maltophilia is extremely heterogeneous at the genetic level (Coenye et al., 2004; Kaiser et al., 2009). We described a procedure to obtain a rapid genotyping of S. maltophilia isolates based on the measurement of length variations of genomic regions marked by arrays of palindromic sequences (Roscetto et al., 2008). In this paper, we describe the organization and the features of this peculiar class of repeats, called SMAG (Stenotrophomonas maltophilia GTAG), because they carry at one terminus the tetranucleotide GTAG. SMAGs are species-specific members of the superfamily of repetitive extragenic palindromes (REPs), sequences described earlier in Escherichia coli and other Enterobacteriaceae (Higgins et al., 1988; Versalovic et al., 1991; Bachellier et al., 1999) and later on in other prokaryotes (Aranda-Olmedo et al., 2002; Feil et al., 2005; Tobes & Pareja, 2005; Tobes & Ramos, 2005). SMAGs constitute the largest family of REPs described so far. A look at the structure and organization of SMAG elements provides information on the processes underlying the expansion and remodeling of REP families, and the functional role that REPs may play.

Materials and methods

In silico analyses

Searches were carried out on the genomes of the S. maltophilia strains K279a (http://www.ncbi.nlm.nih.gov/nuccore/NC_010943) and R551-3 (http://www.ncbi.nlm.nih.gov/nuccore/NC_011071) and the 50 contigs of the strain SKA14 (http://www.ncbi.nlm.nih.gov/nuccore/NZ_ACDV00000000). The K279a genome was searched for SMAG sequences using the fuzznuc program (http://mobyle.pasteur.fr/cgi-bin/portal.py?form=fuzznuc). Initial searches were performed using as a query the sequence described in Roscetto (2008), and selecting homologous sequences containing up to four mismatches. Sequence variants were subsequently used as queries for refined searches. Regions of interest in the R551-3 and SKA14 genomes were identified by blast.

Bacterial strains and PCR analyses

SMAG-negative regions were searched in the DNA of 25 S. maltophilia strains (92, 262, 527, 545, 549, 598, 616, 707, 714, 915, 1019, 1029, 1039, 1054, STM2, OBGTC3, OBGTC13, OBGTC16, OBGTC22, OBGTC28, OBGTC29, OBGTC30, LMG959, LMG10851 and LMG10871) by PCR and sequence analyses. The strains and PCR conditions were described previously (Roscetto et al., 2008).

RNA analyses

Reverse transcriptase-PCR (RT-PCR) analyses were carried out by reverse transcribing total S. maltophilia RNA by random priming, and amplifying the resulting cDNA using pairs of gene-specific oligonucleotides as described (De Gregorio et al., 2005). RNAse protection and primer extension assays were carried out as described (De Gregorio et al., 2005). The sequences of all the primers used are available upon request.

Results

The SMAG family

A thorough analysis of the chromosome of the S. maltophilia K279a strain revealed that the SMAG family is much wider than postulated initially (Roscetto et al., 2008). K279a DNA hosts 1650 SMAG repeats, all constituted by a stem-loop sequence (SLS) flanked, at one side, by the tetranucleotide GTAG. The genomic coordinates of all SMAGs are reported in the Supporting Information, Table S1. The elements can be sorted, on the basis of changes in the stem and loop residues, into 40 variants. For the sake of simplicity, they have been assigned to five major subfamilies (Fig. 1a). The large SMAG-1 subfamily includes all the repeats used for genotyping (Roscetto et al., 2008). SMAG-1 to SMAG-4 repeats have 8 bp stems and SMAG-5 repeats have 9 bp stems. The S. maltophilia genome contains hundreds of DNA tracts that partly resemble SMAG sequences. We discarded complementary sequences fitting the consensuses shown in Fig. 1a, but either located 5 bp away or more, or containing more than two mismatches. In the selection scheme adopted, GT pairing was allowed, because SMAGs may fold into secondary structures at the DNA as at the RNA level. In most repeats, stem sequences are fully complementary (Fig. 1b). An exception is SMAG-2 units, many of which have stems with one to two mismatches. In 50% of the stems with one mismatch, the first base pair is mutated. The folding ability of these elements is therefore impaired only slightly.

1

The SMAG family. (a) The consensus sequences of the five subfamilies of SMAG repeats identified in the K279a genome, and the relative abundance of each subfamily, are shown. Sequences are according to the IUB codes: A, adenosine; C, cytidine; G, guanosine; T, thymidine; H, A or C or T; D, A or G or T; K, G or T; M, A or C; N, any base; R, A or G; S, G or C; Y, C or T. Complementary residues are underlined. (b) Abundance of SMAG-1 to SMAG-5 units with zero, one or two mismatches.

Genomic organization of SMAG repeats

Only 20% of the SMAG family is comprised of solitary elements. Most repeats are grouped into a few predominant arrangements, described below.

Dimers >1/3 of the SMAG family is comprised of elements located at a close distance (<100 bp) from each other. On the basis of their relative position, these elements form head–head (HH) or head–tail (HT) or tail–tail (TT) dimers. Dimers range in size from 47 to 142 bp, the majority of them being ∼70–90 bp in size. Paired repeats belong to the same (homodimers) or different (heterodimers) subfamilies. In total, 228 HH, 55 HT and 26 TT dimers were identified in the K279a chromosome (Fig. 2). HH homodimers represent the most abundant category of paired elements. The differences among dimer categories shown in Fig. 2 are statistically significant (χ2=53.4, P=2.5 × 10−12). A main difference among the HH, TT and HT dimers is that repeats of the first two classes may fold, rather than into separate SLSs, into a large one (Fig. 2). According to analyses carried out at the mfold web server (Zuker, 2003), 70% of HH dimers may fold into large SLSs, with dG values ranging from −50 to −70 kcal mol−1. In none of the three classes of heterodimers could a preferential combination of specific subfamilies repeats be observed. In terms of homodimers, HH dimers are predominantly comprised of SMAG-1, SMAG-2 and SMAG-3 sequences. In contrast, TT dimers are predominantly comprised of SMAG-4 (Fig. 3).

2

SMAG dimers. The three classes of SMAG dimers, and the SLSs potentially formed by alternative folding of HH and TT dimers are depicted. The abundance of HH, TT and HT homodimers and heterodimers is shown.

3

Abundance of HH, TT and HT dimers among SMAG subfamilies.

Spacer sequences that separate dimer repeats are poorly homologous. An exception is the spacers of SMAG-3 HH homodimers, most of which (30/40) fit the consensus sequence nnCGCGCGCAGCGCGGn(16−19)GAAGAGC.

Trimers at 86 loci in the K279a genome, groups of three repeats can be found at a close distance from each other. Taking into account the relative position of each element, trimers can be viewed as dimers flanked by solo repeats. Twenty-eight trimers include SMAGs from one subfamily, 58 SMAGs belonging to two or three subfamilies.

Clusters 456 elements are clustered at 64 loci at a 10–150 bp distance from each other. Large clusters may include up to 22 repeats, and contain elements from different subfamilies. Most clusters contain 4–8 SMAGs, are comprised of repeats of one subfamily and result from tandem amplification of SMAGs (monomers or dimers), together with stretches of flanking DNA of variable lengths.

Interspersion of SMAGs with coding sequences

Many SMAG monomers, dimers and trimers are at a close distance from genes. We found 307 SMAGs located 1–20 bp from ORF stop codons, and 99 that overlap ORF stop codons. Nine of the overlapping repeats encode a few aminoacids and the stop codon; all the others provide only the stop codon, the terminal GTAG motif functioning as a UAG translational stop signal. Curiously, the stop codons of the convergently oriented ORFs Smlt0783–Smlt0784 and Smlt4197–Smlt4198, are contributed by interleaved SMAG dimers. The same holds for ORFs Smlt1380–Smlt1381 and Smlt0188–Smlt0189, the stop codons of each being contributed by interleaved SMAG trimers. Some SMAGs located between convergently oriented ORFs are at a close distance from the stop codons of both. Accordingly, the number of the ORFs immediately flanked by SMAGs is higher than the number of repeats (501 vs. 406). By contrast, we found only 81 SMAGs located 1–50 bp from ORF stop codons, and 16 that overlap ORF start codons and encode 4–29 aminoacids. About 1/3 of the ORFs flanked 5′ by SMAGs (26/97) carries SMAG sequences also at the 3′ end. K279a ORFs at a close distance from SMAGs are listed in Table S2.

Thirty SMAGs are entirely located within ORFs. These repeats can be sorted into two main groups. Sixteen out of 30 lie within ORFs encoding small hypothetical proteins that do not exhibit significant homology to ORFs encoded by either the S. maltophilia R551-3 or other prokaryotic genomes, and thus plausibly do not correspond to authentic gene products. Similar conclusions were reached for short ORFs interrupted by REPs in Pseudomonas syringae (Tobes & Pareja, 2005). The remaining 14 repeats are found at the same relative genome coordinates in the R551-3 DNA. However, only six interrupted ORFs are conserved in the two strains. SMAGs within ORFs are listed in Table S3.

On the whole, intergenic SMAGs are found at 747 loci. Of these, 370 separate unidirectionally transcribed ORFs, 343 convergently transcribed ORFs and only 34 divergently transcribed ORFs.

Conservation of SMAG sequences in other S. maltophilia strains

The size of repeated DNA families may vary among isolates. To gain a rough estimate of the size of SMAG families scattered in the other two sequenced S. maltophilia genomes, repeats perfectly matching the 40 SMAG sequence variants found in K279a DNA were searched in R551-3 and SKA14 DNAs. The relative abundance of the five SMAG subfamilies is comparable in the three genomes. However, their sizes varied, SMAG-2 elements being more abundant in R551-3 and SKA14 and SMAG-3 being predominant in K279a DNA (Fig. 4). The degree of conservation of SMAG sequences was checked by direct sequence comparisons. Thirty-two regions of the K279a chromosome containing SMAG-3 dimers were analyzed in R551-3. Dimers were conserved in 10 regions, missing in nine and replaced in 13 by SMAG-1 or SMAG-2 sequences (monomers or dimers). Fifty K279a intergenic regions containing SMAG-1 HH dimers were also checked in R551-3 DNA. Most (91%) of the K279a SMAG-1 fit the consensus WGCCGGCCgctGGCCGCCW, and have been called α units, and only 4% fit the consensus CGCCGGGCcatGCCCGGCG, and have been called β units (lowercase letters denote loop sequences). Consequently, most (88/99) K279a SMAG-1 HH dimers are comprised of α units. α dimers were conserved in 32/50 regions. Yet, the significant difference in spacer sequences makes it likely that some K279a dimers had been replaced by homologous dimers in R551-3 DNA or vice versa. α dimers were replaced by single β repeats in four regions, β HH dimers in five regions, β TT dimers in three regions and an SMAG-5 TT dimer in one region. SMAG sequences were not found in five regions. In three of them, 40–90-bp-long tracts with an almost perfect dyad symmetry were found.

4

Size variations of SMAG subfamilies in Stenotrophomonas maltophilia genomes. The height of stacked bars denotes the dimension of each subfamily in the three sequenced S. maltophilia genomes.

The changes observed arise from a recombination plausibly driven by the terminal GTAG sequences. The presence at several sites of either alternative SMAGs or unrelated palindromic sequences suggests that the functional role played by SMAG repeats is primarily associated with their ability to fold into secondary structures.

RNA analyses of SMAG-containing selected loci

The pattern of chromosomal interspersion suggests that many SMAG sequences may be passively transcribed into mRNA. Folding of these repeats into RNA hairpins may influence the level of expression of flanking genes. To investigate this issue, 14/50 K279a chromosomal regions containing SMAGs inserted between unidirectionally transcribed genes, and located at a short distance from both, were selected, and their lengths were measured in 25 S. maltophilia strains by PCR. The sizes of the amplimers suggested that SMAG sequences were conserved in most of the analyzed regions. Only two SMAG-negative regions were identified in two different strains, 545 and STM2, and the lack of SMAG DNA was confirmed by sequence analysis. Transcripts spanning the selected genes were detected by RT-PCR, and SMAG-negative regions functioned as a control. The detection of K279a transcripts encompassing both ORFs in each pair ensured that ORFs and interleaved SMAGs are transcribed from the same promoter (Fig. 5). For both gene pairs, upstream transcripts accumulated at higher levels than downstream transcripts in K279a, but not in the strains 545 and STM2 lacking SMAG sequences (Fig. 5). The 4076/4075 cDNA ratio did not change in strain 1029, in which ORFs are separated by a SMAG monomer (Fig. 5b). This suggests that, in a given RNA context, SMAG monomers and dimers function as RNA stabilizers with the same efficiency. We also analyzed a trimeric SMAG repeat located 4 bp downstream from the sensor kinase and the response regulator genes of the smeS–smeR two-component system (ORFs 4477 and 4478), and 13 bp upstream of ORF 4479, which encodes a hypothetical protein. The short distances suggest that the SMAG trimer is cotranscribed with flanking ORFs. We failed to identify strains lacking SMAG sequences in this region that could function as a control. RT-PCR experiments similar to those shown in Fig. 5 revealed that downstream 4479 transcripts accumulated at high levels, but upstream 4478 transcripts were almost undetectable (Fig. 6a). To clarify the issue, RNAse protection assays were carried out. Antisense RNAs of different lengths spanning 4478 and 4479 ORFs protected only 4479 transcripts (Fig. 6b). Intriguingly, protected bands included the SMAG repeat labeled as c in Fig. 6b. The same result was obtained in RNA extension experiments, in which bands of elongation extended over SMAG repeat c only (Fig. 6c). We hypothesize that repeats a and b fold into one large secondary structure, which is cleaved, and this promotes rapid 3′–5′ degradation of upstream 4478 transcripts.

5

RNA expression of genes flanked by SMAGs. Total RNAs (200 ng) derived from the K279a, STM2, 545 and 1029 Stenotrophomonas maltophilia strains were reverse transcribed using a mixture of random hexamers as primers. Transcripts homologous to ORFs 3592 and 3591 [(a)] and ORFs 4076 and 4075 [(b)] were measured by RT-PCR using pairs of gene-specific oligonucleotides. Lanes a and b show the reaction products obtained after 24 and 27 amplification cycles, respectively. Amplimers were detected only when samples were incubated with reverse transcriptase (RT+lanes) before PCR. The 911- and 1029-bp amplimers in the small autoradiogams in the left side of (a) and (b) correspond to K279a transcripts spanning SMAG and flanking ORFs detected with primers 1 and 4 (ORFs 3592 and 3591) and 5 and 8 (ORFs 4076 and 4075) after 35 PCR cycles. Single and double hairpins indicate SMAG monomers and dimers, and dotted lines indicate SMAG-negative intergenic regions. ORFs 3592 and 3591 encode the adenylosuccinate synthetase and a putative transmembrane protein, and are at a distance of 1 and 3 bp from SMAG sequences, respectively. The two ORFs are the last two cistrons of an operon also including ORFs 3596, 3595, 3594 and 3593. ORFs 4076 and 4075 correspond to the heat shock proteins HslV and HslU, and are at a distance of 2 and 25 bp from SMAG sequences, respectively.

6

Cleavage of SMAG RNA. (a) Transcripts corresponding to ORF 4479 (down) and ORF4478 (up) accumulated in K279a cells were detected as in Fig. 5. (b) Antisense RNA probes A and B (lanes 1 and 4) were hybridized to 20 μg of Stenotrophomonas maltophilia K279a (lanes 2 and 5) or yeast (lanes 3 and 6) RNA. T1 RNAse-resistant RNA hybrids were electrophoresed on a 6% polyacrylamide–8 M urea gel. (c) A primer complementary to ORF 4479 was hybridized to 10 μg of K279a RNA. Annealed primer moieties were extended by reverse transcriptase, and elongated products were electrophoresed on a 6% polyacrylamide–8 M urea gel. Major reaction products are marked by arrows. Numbers to the left of the autoradiograms indicate the size in nucleotides of coelectrophoresed DNA molecular size markers. SMAG sequences are shown as in Fig. 5. Filled circles indicate GTAG termini.

Discussion

The number of predicted SLSs is significantly higher in prokaryotic genomes existing in nature than in random sequences of comparable GC content (Petrillo et al., 2006). This implies that the ability of a variety of sequences to fold into secondary structures is positively selected in prokaryotic genomes and may have functional significance. A fraction of SLSs is represented by REPs, sequences shown or hypothesized to serve different functions. REPs are binding sites for the integration host factor, a protein required for site-specific recombination and DNA replication (Engelhorn et al., 1995). REPs are targets for the DNA gyrase (Espéli & Boccard, 1997), and repeats located between convergent genes may be a privileged target for the enzyme, in order to counteract the excess of positive supercoiling induced in the chromosome by DNA transcription (Moulin et al., 2005). As RNA elements, REPs may enhance the stability of 5′ proximal mRNA segments (Khemici & Carpousis, 2004). Finally, REPs induce innate immune system stimulation via TLR9, and could play a key role in the pathogenesis of Gram-negative septic shock (Magnusson et al., 2007).

Tobes & Ramos (2005) established that, for a palindromic sequence to be considered as REP, the following criteria should be met: (a) be extragenic, (b) range in size from 21 to 65 bp and (c) constitute >0.5% of the total intergenic space. SMAGs meet all these criteria, and constitute the largest set of REPs described so far. SMAGs correspond to the repeats identified by Nunvar (2010). SMAGs can be sorted into five distinct subfamilies, and come in different genomic formats. Single units make up only 1/5 of the SMAG family. The remaining elements are organized as dimers or are grouped in tandem arrays of variable lengths. Altogether, SMAGs and intermingled DNA occupy 13% of the overall intergenic space, and make up 1.4% of the total chromosome.

SMAG families residing in the environmental R551-3 and SKA14 S. maltophilia strains are comparable in size to the repeat family found in K279a. Yet, the sizes of some subfamilies vary, and K279a is enriched in SMAG-3. Most SMAG-3 are organized as HH dimers that feature conserved spacers, and may thus represent a relatively young sequence family variant. Changes in the abundance and chromosomal distribution may make SMAG-3 sequences suitable for use in accurate genotyping and epidemiological studies.

Also, the ∼500 REPs identified in the E. coli MG1655 strain have been sorted into subfamilies. Similar to SMAGs, single REPs represent only 20% of the family, the other elements being grouped in various configurations, all denoted as bacterial interspersed mosaic elements (BIME; reviewed in Bachellier et al., 1999). BIME-1 and BIME-2 correspond to SMAG TT and HH dimers. However, HH dimers are about 10 times more abundant than TT dimers. In contrast, BIME-1 (74 repeats) are three times more abundant than BIME-2 (24 repeats). Moreover, both BIME-1 and BIME-2 are invariably comprised of elements from different subfamilies (Bachellier et al., 1999; see also http://www.pasteur.fr/recherche/unites/pmtg/repet/index.html). The predominance of TT over HH dimers, and the composite nature of dimers, is also a distinctive feature of the abundant REP families found in Pseudomonas putida (Aranda-Olmedo et al., 2002) and P. syringae (Feil et al., 2005).

It has been hypothesized that REPs are mobilized by a transposase of the IS200/IS605 family, and the corresponding genes have been shown to be flanked by REPs in many species (Nunvar et al., 2010). Four genes encoding this transposase were identified in K279a DNA (ORFs 1101, 1152, 2816 and 4509), but only ORFs 1101 and 2816 are flanked by SMAGs. We believe that REPs are an ancient component of the genomes of Proteobacteria, which have been actively mobilized by transposition only early in their history. According to this view, REPs disappeared in time from most species, their dissemination being plausibly detrimental to the cell, and have been maintained only in species in which they could no longer transpose. This hypothesis is supported by the observation that SMAG sequences were found in none of the 41 species-specific GEIs, plausibly acquired by lateral gene transfer, which account for >10% of the K279a chromosome (Rocco et al., 2009). REPs are similarly restricted to core genome regions in P. syringae (Tobes & Pareja, 2005). In contrast to what was observed for REPs in other species (Tobes & Pareja, 2006), SMAGs are not targeted by mobile DNA. However, it is worth noting that a K279a GEI encoding type 1 pili (Rocco et al., 2009) is flanked by SMAG-2 dimers. About 1/7 of the ORFs of the K279a strain are flanked by SMAGs in a distance range that makes the presence of promoter or terminator sequences unlikely. It is plausible that most of these elements are transcribed into mRNA, and that their folding into RNA hairpins may influence the level of expression of flanking genes. The number of genes potentially controlled at the post-transcriptional level by SMAGs may be higher than estimated, because many repeats are inserted either upstream (17 elements) or downstream (150 elements) or within (44 elements) known or putative operons.

We analyzed genes transcribed in the same direction intermingled with SMAG sequences, and found that the repeats influence the segmental mRNA stability. Both monomers and dimers function as stabilizers of upstream transcripts, and work with comparable efficiency when embedded in the same RNA context (Fig. 5). RNA expression data are in line with the results of in silico analyses, indicating that some of the genes separated by HH dimers in K279a are intermingled with monomers, or HH, or even TT dimers of the same or different SMAG subfamilies in R551-3 DNA. This varied scenario shows that recombination may extensively reshape SMAG-positive regions without substantially altering the regulatory role of SMAGs. The distance between ORFs and SMAGs increased 10–15 bp in some R551-3 regions. This suggests that SMAGs may function as RNA elements over a relatively flexible distance interval. Some SMAGs may favor the degradation of upstream transcripts. This may correlate to the cleavage of large SLSs formed by alternative folding of SMAG dimers (Fig. 6). These structures resemble RNA hairpins formed by 100–170 bp repeats found in Neisseriae (De Gregorio et al., 2003) and Yersiniae (De Gregorio et al., 2006), which may be cleaved by RNAse III. Whether the hypothesized structures may be formed, whether they are cut by specific endoribonucleases or are resistant to cleavage is likely determined by the overall mRNA context in which SMAG dimers are embedded. Thorough analyses may eventually establish how SMAG sequences regulate the level of expression of different sets of S. maltophilia genes.

The dimensions and the complexity of the SMAG family make S. maltophilia an ideal organism to gain knowledge of the universe of small palindromic sequences, and clarify the roles that they may play in the lifestyle of the organisms in which they reside.

Supporting Information

Table S1. Sequences and chromosomal coordinates of the 1650 SMAG sequences found in K279a DNA.

Table S2. SMAGs that are close to, or overlap K279a ORFs, are listed.

Table S3. K279a ORFs containing SMAG sequences.

Acknowledgements

We are indebted to Raffaele Zarrilli for critically reading the manuscript, and Sergio Cocozza for statistical analyses. We thank one of the referees for hints and suggestions. Research was supported by a grant from the Italian Cystic Fibrosis Research Foundation (FFC) to P.P.D.N.

Footnotes

  • Editor: Roger Buxton

References

View Abstract