Genetic Characterization of the Capsulation Locus of Haemophilus influenzae Serotype e

Abstract

The capsulation (cap) locus of Haemophilus influenzae type e (Hie) was characterized and sequenced. No IS1016 element was found to flank the locus. The 18.2-kb locus included 14 open reading frames (ORFs), which were grouped into three functional regions. Eight new ORFs (named ecs1 to ecs8) were identified in the Hie capsule-specific region II.

In the post-Haemophilus influenzae serotype b (Hib) vaccine era, concern about the potential emergence of non-vaccine-preventable strains has arisen (1, 17, 20, 23, 26). In encapsulated H. influenzae strains, the genes for the production of the polysaccharide capsules are organized in a capsulation (cap) locus, which consists of three different functional regions (11, 13). Regions I and III are common to all capsular types and contain genes necessary for transport and process of the capsular material, while region II contains serotype-specific biosynthesis genes (7, 10, 18, 19, 25).

Invasive disease caused by H. influenzae serotype e (Hie) strains has recently been observed in Italy, suggesting the importance of further molecular investigations on Hie cap locus (4, 5). It is recognized that the Hie capsule is a copolymer of the repeat unit of an N-acetylglucosamine and N-acetylmannosamine uronic acid (22, 24), but the genes involved in the polysaccharide biosynthesis have neither been identified nor characterized.

In the present study, we characterized the Hie cap locus for the first time. Eleven invasive Hie strains isolated in Italy during the period of January 2000 to December 2008 were analyzed. The strains were identified as type e by PCR capsular genotyping (6).

Location of the Hie cap locus within the chromosome.

PCR amplification of the 5′ and 3′ end junctions of the Hie cap locus was performed by using primer sets “capfSodC/bexBrev” and “hcsBfrw/HI1637,” respectively (Table 1). The resulting PCR products were sequenced and analyzed. All 11 Hie strains were found to have the cap locus in the identical chromosomal location as that of H. influenzae serotype f (Hif), associated with the same flanking genes (sodC at the 5′ end and HI1637 at the 3′ end), confirming previous investigations (19). Sequence analysis of the two end junctions also revealed that they contained no sequences reminiscent of the insertion element IS1016. It is well known that this element provides the molecular substrate for amplifications of the cap gene sequences (11). Most Hib strains, in which the cap locus lies between direct repeats of IS1016, possess a duplication of the capsule genes (11, 12). The finding that our Hie strains' lack of IS1016 sequences flanking the cap locus is remarkable, since reasonably the locus cannot be amplified.

TABLE 1.

PCR primers and products used for sequencing of the Hie capsulation locus

Primer set	Nucleotide sequence (5′ to 3′)	Region amplified	Size (bp)	Source or reference
capfSodC	CATGCGCATTTTCCACGCCAGC	sodC-bexB	1,581	19
bexBrev	TAGCGATTCAAGGGAGGGT	sodC-bexB	1,581	This study
bexBfrw	ACGCCCATAACGAGAGACT	bexB-bexD	2,004	This study
bexDrev	TCGCAGGTAAGACACCAGAG	bexB-bexD	2,004	This study
bexDfrw	AAAGACACCTCGTGGGTCA	bexD-regionIII	5,003	This study
e2	GCTTTACTGTATAAGTCTAG	bexD-regionIII	5,003	6
el	GGTAACGAATGTAGTGGTAG	regionIII-hcsA	9,045	6
hcsArev	ACTGACCGCACTTTACGACG	regionIII-hcsA	9,045	This study
hcsAfrw	GCACAAAGTGAGCGTCGTA	hcsA-hcsB	1,703	This study
hcsBrev	ATAGAAGTCTGCCTGGCGAG	hcsA-hcsB	1,703	This study
hcsBfrw	GATTGCTTATCGTGGCTCAGT	hcsB-HI1637	1,504	This study
HI1637	AAATTTCCATTATGGGAAACG	hcsB-HI1637	1,504	19

Sequencing of the Hie cap locus.

The complete cap locus from the strain Hie 274 (isolated from the cerebrospinal fluid of a patient with meningitis) was sequenced. To cover the entire Hie cap locus, overlapping amplicons ranging from 1,500 bp to 9 kb were obtained by PCR analysis using several primer pairs based on the published sequences of the Hib and Hif cap loci (GenBank accession numbers AF549213 and AF549211, respectively) (Table 1). Each amplicon was then subcloned into pCR4-TOPO (TOPO TA cloning kit or TOPO XL PCR cloning kit; Invitrogen, Milan, Italy). Both strands of the insert from each plasmid were sequenced by the primer walking service at Eurofins MWG Operon (Ebersberg, Germany). The nucleotide sequences were assembled and analyzed using DNAMAN sequence analysis software (version 5.2; Lynnon Corp., Quebec, Canada). Nucleotide and deduced amino acid sequences were compared to other known sequences databases by using the National Center for Biotechnology Information BLAST programs. The entire Hie cap locus was shown to be 18.2 kb in length. It contained 14 open reading frames (ORFs) which, by analogy with other capsule loci, were grouped into three distinct regions (I, II, and III) (Fig. 1). A comparison of the deduced proteins from the Hie cap locus genes with the corresponding gene products from Hib and Hif as well as with proteins from other bacterial species is shown in Table 2.

FIG. 1.

Genetic organization of the Hie capsulation locus of strain 274. The arrows indicate genes. Region I contains four genes called bexDCBA, homologous to those found in Hib and Hif (white arrows). Region II includes eight serotype-specific genes designated ecs1 to ecs8 (right-hatched arrows). Region III comprises two genes named hcsA and hcsB, homologous to those found in Hib and Hif (gray arrows).

TABLE 2.

Comparison of the deduced proteins from the H. influenzae serotype e capsulation locus of strain 274

Hie 274 deduced protein, no. of amino acids (aa)	Similar protein (source organism)	Accession no.	% Identity	% Similarity
Region I
BexA, 217 aa	BexA (H. influenzae serotype f)	AF549211	94	97
	BexA (H. influenzae serotype b)	AF549213	94	97
	CpxA (Actinobacillus pleuropneumoniae)	CP000687	82	91
	CtrD (Neisseria meningitidis)	EU038216	80	89
	CpxA (Mannheimia haemolytica)	AF170495	76	88
BexB, 265 aa	BexB (H. influenzae serotype f)	AF549211	97	99
	BexB (H. influenzae serotype b)	M33788	95	99
	CpxB (Actinobacillus pleuropneumoniae)	CP000687	77	90
	CpxB (Mannheimia haemolytica)	AF170495	74	86
	CtrC (Neisseria meningitidis)	EU038216	69	83
BexC, 377 aa	BexC (H. influenzae serotype f)	AF549211	98	98
	BexC (H. influenzae serotype b)	AF549213	94	96
	CpxC (Mannheimia haemolytica)	AF170495	76	88
	CpxC (Actinobacillus pleuropneumoniae)	CP000687	74	87
	CtrB (Neisseria meningitidis)	EU038216	59	79
BexD, 428 aa	BexD (H. influenzae serotype f)	AF549211	91	94
	BexD (H. influenzae serotype b)	AF549213	91	95
	CpxD (Actinobacillus pleuropneumoniae)	CP001091	73	85
	CpxD (Mannheimia haemolytica)	AF170495	71	84
	CtrA (Neisseria meningitidis)	AF520902	55	74
Region II
Orf1 (Ecs1), 374 aa	SacA (Neisseria meningitidis)	AL157959	71	85
Putative UDP-N-acetyl-d-glucosamine 2-epimerase	VIBHAR_00689(Vibrio harveyi)	CP000789	63	77
	VV0341 (Vibrio vulnificus)	BA000037	62	77
	WecB (Klebsiella pneumoniae)	CP000647	61	75
	WecB (Escherichia coli)	AE014075	61	76
Orf2 (Ecs2), 421 aa	WecB (Mannheimia haemolytica)	AF170495	67	79
Putative UDP-N-acetyl-d-mannosaminuronic acid dehydrogenase	EcbB (Pasteurella multocida)	AF302466	64	77
	ORF6 (Pseudomonas aeruginosa)	AF498407	63	77
	WecC (Escherichia coli)	CP000948	61	75
	SeD_A4308 (Salmonella enterica)	CP001144	61	74
Orf3 (Ecs3), 991 aa	BcbC (Pasteurella multocida)	AF169324	50	69
Putative glycosyltransferase	Msm_1297 (Methanobrevibacter smithii)	CP000678	34	55
	Msp_0219 (Methanosphaera stadtmanae)	CP000102	33	51
	Eco1C (Escherichia coli)	CP000946	31	50
	WaaV (Shigella sonnei)	CP000038	30	50
Orf4 (Ecs4), 486 aa	BcbD (Pasteurella multocida)	AF169324	46	65
Unknown function	KfoD (Escherichia coli)	AB079602	36	55
	ORFA (Yersinia enterocolitica)	AY653208	21	43
	EcbD (Pasteurella multocida)	AF302466	26	42
	CMU_015760 (Cryptosporidium muris)	XM_002140142	26	41
Orf5 (Ecs5), 240 aa	ORF5 (Actinobacillus suis)	AY253301	68	85
Unknown function	BcbE (Pasteurella multocida)	AF169324	65	82
	BcbE (Photobacterium damselae)	AB074293	49	67
	SamA (Shewanella amazonensis)	CP000507	46	64
	VC0395 (Vibrio cholera)	CP000627	45	63
Orf6 (Ecs6), 125 aa	ORF6 (Actinobacillus suis)	AY253301	77	91
Unknown function	BcbF (Pasteurella multocida)	AF169324	73	89
	SamA (Shewanella amazonensis)	CP000507	60	79
	VC0395 (Vibrio cholerae)	CP000627	60	78
	BAG50482 (Vibrio parahaemolyticus)	AB353134	64	78
Orf7 (Ecs7), 519 aa	BcbG (Pasteurella multocida)	AF302466	65	77
Unknown function	ORF7 (Actinobacillus suis)	AY253301	63	78
	BcbG (Photobacterium damselae)	AB074293	37	56
	VC0395 (Vibrio cholerae)	CP000627	36	55
	BcbG (Zymomonas mobilis)	AE008692	36	54
Orf8 (Ecs8), 325 aa	BcbI (Pasteurella multocida)	AF169324	64	78
Unknown function	ORF2 (Mannheimia haemolytica)	AF170495	63	74
	Fphi_1179 (Francisella philomiragia)	CP000937	54	73
	Neut_1976 (Nitrosomonas eutropha)	CP000450	40	63
	NE1334 (Nitrosomonas europaea)	AL954747	43	63
Region III
HcsA, 595 aa	HcsA (H. influenzae serotype b)	DQ368335	96	97
	HcsA (H. influenzae serotype f)	AF549211	95	97
	LipA2 (Actinobacillus pleuropneumoniae)	CP000687	62	75
	PhyA (Mannheimia haemolytica)	AF170495	59	72
	LipA (Neisseria meningitidis)	AM421808	56	71
HcsB, 420 aa	HcsB (H. influenzae serotype b)	DQ368335	91	95
	HcsB (H. influenzae serotype f)	AF549211	90	94
	PhyB (Actinobacillus pleuropneumoniae)	CP000687	65	78
	PhyB (Pasteurella multocida)	AF067175	64	77
	LipB (Neisseria meningitidis)	Z13995	55	68

Region I.

Overall, region I exhibited 90% and 96% sequence identity to the previously described region I from Hib and Hif, respectively (18, 19). Region I included four ORFs, which were named bexA, bexB, bexC, and bexD. Although the putative proteins of genes bexABCD were nearly identical (from 91 to 98% identity) to the region I corresponding gene products from both Hib and Hif (Table 2), some polymorphism at nucleotide sequence level was observed. The bexA gene from the Hie cap locus exhibited 95% identity to bexA from Hif but only 84% identity to bexA from Hib, in agreement with a previous study demonstrating bexA nucleotide sequence diversity among different H. influenzae serotypes (27).

Region III.

Overall, region III showed 91% and 93% sequence identity to the previously described region III from Hib and Hif, respectively (18, 19). Region III contained two ORFs, which were named hcsA and hcsB. Their deduced amino acid sequences exhibited high identity (from 90 to 96% identity) with the corresponding products from both Hib and Hif region III (Table 2). Recently, both HcsA and HcsB proteins have been demonstrated to be crucial for transport of capsular polysaccharide from the periplasm to the bacterial surface across the outer membrane (21).

Region II.

Overall, region II showed no sequence identity to the previously described specific capsular regions from other H. influenzae serotypes (7, 18, 19). On the contrary, high overall sequence identity (67%) was found with the capsule biosynthetic-specific region II from Pasteurella multocida B:2 (accession number AF169324), indicating that the genetic organization of the whole region is similar (2, 3). The G+C content of the DNA in the Hie cap locus region II is 31.3%, significantly different from that of both regions I and III (38% and 39.4%, respectively) and from the overall background for the H. influenzae species (38%), suggesting that region II might be more recently acquired. However, since the G+C content of DNA of P. multocida cap locus region II is 35%, this microorganism was probably not the direct source of the region II for Hie. Although we cannot rule out a common evolutionary origin of the two polysaccharide biosynthetic regions followed by a partial diversification of their DNA content, no data are available to support this hypothesis. Region II contained 8 ORFs, which were named ecs1 to ecs8 (for serotype e capsule-specific genes) (Table 2). The deduced products of ecs1 and ecs2 had homology with putative UDP-N-acetyl-d-glucosamine 2-epimerase and UDP-N-acetyl-d-mannosaminuronic acid dehydrogenase enzymes, respectively, which catalyze the two-step conversion of UDP-N-acetyl-d-glucosamine to N-acetyl-d-mannosaminuronic acid, as previously demonstrated with Escherichia coli (14). The encoded protein by the ecs3 gene showed similarity to glycosyltransferases (Table 2), which are involved in polymerization of the sugar monomers in several bacterial species (8, 9). Considering that the structure of the Hie capsular polymer is composed of repeating units of N-acetylglucosamine and N-acetylmannosamine uronic acid (22, 24), it is likely that the products of the ecs1, ecs2, and ecs3 genes play an essential role in the biosynthesis of serotype e polysaccharide. No specific putative functions were assigned to the remaining 5 ORFs (ecs4 to ecs8), although similarity with other deduced products in the database was detected, including the predicted products of the genes bcbDEFGI from cap locus region II from P. multocida (2), (Table 2). Further studies of functional activities of the Hie cap locus region II genes are required.

Although Hie strains belong to the phylogenetic division I of the encapsulated H. influenzae strains (15), the Hie cap locus shares two remarkable features of the division II cap loci: chromosomal location and lack of association with the IS1016 insertion element, confirming the previously described genetic distance of Hie from all other division I H. influenzae strains (16). The availability of the Hie cap locus sequences may be regarded as a powerful tool to be used in further investigations on molecular detection and characterization of the Hie isolates.

Nucleotide sequence accession number.

The nucleotide sequence for the Hie cap locus from this study has been deposited in the EMBL nucleotide sequence database under the accession number FM882247.