Role of CCAA Nucleotide Repeats in Regulation of Hemoglobin and Hemoglobin-Haptoglobin Binding Protein Genes of Haemophilus influenzae

Abstract

Haemophilus influenzae utilizes hemoglobin and hemoglobin-haptoglobin as heme sources. The H. influenzae hemoglobin- and hemoglobin-haptoglobin binding protein genes, hgpA, hgpB, and hgpC, contain lengths of tetrameric CCAA repeats. Using an hgpA-lacZ translational gene fusion, we demonstrate phase-variable expression of lacZ associated with alteration in the length of the CCAA repeat region.

Many pathogenic bacteria reversibly vary their array of cell surface components with high frequency from one generation to another, a phenomenon termed phase variation (12, 17). One genetic mechanism mediating phase variation involves changes in the number of repeated nucleotides in mononucleotide (homopolymeric) tracts or tandemly iterated oligonucleotides (12, 17).

Haemophilus influenzae is able to bind hemoglobin, and we have cloned three genes, hgpA, hgpB, and hgpC, encoding heme-repressible hemoglobin- and hemoglobin-haptoglobin binding proteins from H. influenzae type b strain HI689 (7, 8, 11, 13). Variable expression of these three proteins has been observed. Some hemoglobin binding protein affinity experiments yielded a single band at 120 kDa (HgpA) (7), while others yielded two bands at 120 and 115 kDa (8, 13). An hgpA mutant exhibited loss of a 120-kDa protein and increased expression of the 115-kDa protein (HgpB) (13). Furthermore, the hgpA hgpB double mutant exhibited a faint band at approximately 120 kDa, which was not observed in either the hgpA or hgpB single mutant and was identified as HgpC (13). Passage of the hgpA hgpB double mutant through a medium in which hemoglobin-haptoglobin was the sole heme source resulted in increased isolation of HgpC on a weight-per-weight basis (13).

The nucleotide sequences of hgpA, hgpB, and hgpC reveal multiple repeats of tetrameric CCAA units immediately following the sequence encoding the signal peptide (7, 8, 11, 13). The H. influenzae Rd KW20 genome contains four open reading frames (ORFs) with CCAA repeats, encoding proteins of high homology to HgpA (3, 8). This study investigates the potential role of the CCAA repeats in variable expression of the hemoglobin- and hemoglobin-haptoglobin binding proteins of H. influenzae.

Construction of an hgpA-lacZ translational gene fusion.

An hgpA-lacZ fusion was constructed in H. influenzae Rd KW20 (Table 1). The gene fusion was initially constructed in Escherichia coli with subsequent transformation into the H. influenzae chromosome (Fig. 1). In the fusion construction, the codons for mature LacZ are in the same translational frame as those for mature HgpA, resulting in an in-frame HgpA-LacZ fusion protein (Fig. 2A). Strain Rd KW20 was selected for construction of the fusion strain partly because it is more readily transformed than strain HI689 from which hgpA was cloned (unpublished observation). In addition, Rd KW20 does not contain hgpA (8). A 6.6-kbp DNA fragment, containing the ORFs of HI0588, HI0589, HI0590, HI0591, and HI0592 in Rd KW20, has apparently been replaced by hgpA in strain HI689 (8). Thus, insertion of the hgpA-lacZ fusion in strain Rd KW20 at this locus avoids interruption of the heme acquisition pathway of the pathogenic strain HI689. To construct the hgpA-lacZ fusion, a 1.3-kbp DNA fragment was amplified by PCR using the primer pair Phgpfus1 and Phgpnot1 (Table 2) with pHFJ2 (7) as the template. The reaction was performed in 50 μl containing 2 mM MgCl₂, 0.2 mM (each) deoxynucleoside triphosphate, 10 pM (each) primer, and 2 U of Pfu DNA polymerase (Stratagene, La Jolla, Calif.). Thirty cycles of PCR were performed (one cycle consists of denaturation at 95°C for 1 min, annealing at 60°C for 1 min, and extension at 72°C for 1 min) before a final extension step of 10 min at 72°C. This PCR product, encompassing a region upstream of the hgpA promoter, the promoter itself, the signal peptide coding region, and the CCAA repeat region, was cloned into the pCR-Blunt cloning vector (Invitrogen, Carlsbad, Calif.) to yield pFusNot7. The primer Phgpnot1 was designed to incorporate a NotI restriction site at the end of the PCR product. The ExSite PCR-based site-directed mutagenesis kit (Stratagene) was used to construct a NotI site in the hgpA leader sequence, using pFusNot7 as the template, the primers Psdm1 and Psdm2 (Table 2), and an annealing temperature of 52°C, yielding pFusNotII. The two engineered NotI restriction sites, flanking the CCAA repeat region, facilitated excision of the CCAA repeat region from chromosomal DNA for direct sizing of the repeat region.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Relevant characteristic(s)a	Source or reference
Strains
E. coli
TOP10	F−mcrA Δ(mrr-hsdRMS-mcrBC) φ80lacZΔM15 ΔlacX74 recA1 deoR araD139 Δ(ara-leu)7697 galU galK rpsL (Strr) endA1 nupG	Invitrogen
GM2929	F−dam-13::Tn9 (Cmr)	1
H. influenzae
Rd KW20	Capsule-deficient type d derivative; strain used in sequencing project	3
HI1718	Rd KW20 containing the hgpA-lacZ translational gene fusion	This study
HI1719	Rd KW20 containing the hgpA-lacZ translational gene fusion without CCAA repeats	This study
Plasmids
pCR-Blunt	Plac, lacZ-ccdB fusion gene, T7 promoter, Kanr, Sh ble (Zeor), ColE1 origin	Invitrogen
pLKC480	AmprlacZ lacY, lacA interrupted by Tn5 Kanr gene, no Plac	16
pHFJ2	pUC19, carrying a 4.2-kbp EcoRI fragment from H. influenzae, the original hgpA clone	7
pDown1	pCR-Blunt, carrying a 1.0-kbp PCR product from H. influenzae Rd KW20 as the downstream flanking DNA	This study
pFusNot7	pCR-Blunt, carrying a 1.3-kbp PCR product from pHFJ2 and containing the hgpA promoter and its upstream region, the signal peptide coding region, and the CCAA repeat region with one NotI site incorporated after the CCAA repeats	This study
pFusNotII	pFusNot7 with an extra NotI site constructed in the hgpA leader sequence	This study
pRepLac1	pLKC480, carrying the 1.3-kbp fragment from EcoRI/HindIII-digested pFusNotII, hgpA and lacZ genes in the same translational frame	This study
pFusion-LacI6	pRepLac1, carrying a 1.0-kbp PCR product from H. influenzae Rd KW20 as the downstream flanking DNA	This study
pFusion-LacII	pFusion-LacI6, with deletion of the CCAA repeat region	This study

^aRb^r, ribostamycin resistance (15 μg/ml for H. influenzae); Cm^r, chloramphenicol resistance (50 μg/ml for E. coli); Kan^r, kanamycin resistance (50 μg/ml for E. coli); Amp^r, ampicillin resistance (50 μg/ml for E. coli); Zeo^r, zeocin resistance (50 μg/ml for E. coli); F⁻, F episome negative. 

FIG. 1.

Construction of the hgpA-lacZ translational gene fusion in H. influenzae Rd KW20. The insert of the original hgpA clone pHFJ2 (7), the genomic maps of strains HI689 and Rd KW20 at the hgpA locus, and the gene fusion construct pFusion-LacI6 are shown. Each open box indicates an ORF, the direction of transcription is indicated by an arrow, and gene names where they have been assigned are given. Numbers are the numerical designations given to ORFs by Fleischmann et al. (3) in the Rd KW20 genome sequencing project; hgpA does not have an assigned number, since it does not exist in the Rd KW20 genome (8). The shaded area represents the hgpA-lacZ gene fusion construct. The solid bar is the CCAA repeat region of hgpA. The dashed lines indicate the homologous regions, used as the flanking regions for recombination.

FIG. 2.

hgpA-lacZ translational gene fusion in H. influenzae HI1718. (A) Genomic location of the hgpA-lacZ gene fusion. Each open box and an arrow indicate an ORF and its direction of transcription as designated by Fleischmann et al. (3), and numbers are the numerical designations given to ORFs by Fleischmann et al. (3). The shaded area represents the hgpA-lacZ gene fusion construct. The DNA sequence at the fusion junction is shown, starting with the ATG initiation codon for the first amino acid of the HgpA signal peptide. The codons encoding LacZ, beginning at the proline (10), are in the same frame as the codons encoding mature HgpA. The solid bar is the CCAA repeat region of hgpA. The relevant restriction sites are underlined. Southern analyses were performed with the 1.0-kbp DNA fragment containing the downstream flanking region (B) and the 6.3-kbp SmaI fragment from pLKC480 containing the lacZ-Kan^r cassette (C) as probes. Lanes 1, labeled λ digested with HindIII; lanes 2, H. influenzae Rd KW20 chromosomal DNA digested with BamHI and EcoRI lanes 3, the hgpA-lacZ recombinant H. influenzae HI1718 chromosomal DNA digested with BamHI and EcoRI.

TABLE 2.

Primers used in this study

Primer	Sequencea
Pprof	5′-TTTATAGGACTAAATATGACC-3′
Placr	5′-CAAGGCGATTAAGTTGGG-3′
Phgpfus1	5′-ATTAGCATGAATTCCGAACACAACTATCCCATTC-3′
Phgpnot1b	5′-TTTGTTCTAAGCTTGCGGCCGCATTACTATTTTGGTTGG-3′
Psdm1	5′-GCGGCCGCATAACGGAATAGGCAAGC-3′
Psdm2	5′-TAACGGCAAGTGTTGCTTATGC-3′
Plac1	5′-ATGTATTTGAATCTCCAATGAACATTGC-3′
Plac2	5′-AACCGTTTAAATCTTGGATATGG-3′

^aHighlighted portions of sequence (by boldface type) are designed restriction sites in primers as follows: 5′-GAATTC-3′, EcoRI; 5′-AAGCTT-3′, HindIII; and 5′-GCGGCCGC-3′, NotI. 

^bThe primer Phgpnot1 contains both a HindIII and a NotI site. 

The primers Phgpfus1 and Phgpnot1 additionally engineered an EcoRI and a HindIII restriction site at each end of the PCR product. The 1.3-kbp EcoRI/HindIII fragment of pFusNotII was directionally subcloned into EcoRI/HindIII-digested pLKC480 (16) to yield pRepLac1. Plasmid pLKC480 contains a lacZY gene fusion cassette with an aminoglycoside resistance marker, and the subcloning resulted in a fusion with active β-galactosidase. To promote recombination of the gene fusion into the H. influenzae chromosome, a length of H. influenzae DNA was cloned downstream of the aminoglycoside resistance marker in pRepLac1. One kilobase pair of H. influenzae DNA, including the HI0592 putative coding region, was amplified by using the primer pair Plac1 and Plac2 (Table 2) and 100 ng of H. influenzae Rd KW20 chromosomal DNA as the template (annealing at 62°C). The PCR product was ligated into the unique NruI site of pRepLac1 to yield pFusion-LacI6 (Fig. 1) and was also ligated into pCR-Blunt (Invitrogen) to yield pDown1. NotI digestion, followed by religation of pFusion-LacI6, resulted in the non-CCAA-containing fusion construct pFusion-LacII for use as a control.

Plasmids pFusion-LacI6 and pFusion-LacII were transformed into competent H. influenzae Rd KW20 (14), and transformants were selected for by growth on brain heart infusion (BHI) agar supplemented with 10 μg of heme and 10 μg of β-NAD per ml (supplemented BHI [sBHI]) and containing ribostamycin (15 μg/ml). One ribostamycin-resistant Rd KW20 colony from each transformation was selected for further investigation. Appropriate chromosomal rearrangement was confirmed by Southern analysis (Fig. 2B and C), using DNA probes labeled by using the enhanced chemiluminescence (ECL) random prime labeling kit (Amersham Pharmacia Biotech, Piscataway, N.J.). Hybridization was detected by using the ECL nucleic acid detection reagents. The labeled 1.0-kbp downstream flanking region, excised from pDown1, hybridized to an approximately 1.9-kbp BamHI/EcoRI fragment in wild-type Rd KW20 (Fig. 2B, lane 2) and to an approximately 8.9-kbp BamHI/EcoRI fragment in the Rd KW20 hgpA-lacZ recombinant strain containing the CCAA repeats (Fig. 2B, lane 3). The labeled 6.3-kbp SmaI fragment containing the lacZ-aminoglycoside resistance cassette from pLKC480 hybridized to the 8.9-kbp BamHI/EcoRI fragment in the Rd KW20 hgpA-lacZ strain (Fig. 2C, lane 3) and did not hybridize to wild-type Rd KW20 (Fig. 2C, lane 2). Similar data were obtained for the non-CCAA-containing recombinant Rd KW20 strain (data not shown). The recombinant strain containing the CCAA repeats was designated HI1718, and the recombinant strain containing the fusion without CCAA repeats was designated HI1719.

Determination of variable expression of the hgpA-lacZ fusion.

To examine expression of the hgpA-lacZ fusion, a single H. influenzae colony was selected and incubated in sBHI at 37°C for 16 h. One hundred microliters of the culture was serially diluted and spread on sBHI agar to obtain 200 to 500 colonies per plate. After overnight incubation at 37°C, the plates were flooded with 1 ml of a 5-mg/ml 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) solution. Color differences among the colonies were evident within 15 min.

When a single white colony of HI1718, i.e., not expressing lacZ, was subcultured and plated, several blue colonies were seen on a plate flooded with X-Gal (Fig. 3). Similarly, when a single blue colony of HI1718 was subcultured and plated, several white colonies were noted (data not shown). In contrast, HI1719, the hgpA-lacZ fusion strain lacking the CCAA repeats, always expressed lacZ (i.e., colonies were always blue) (data not shown). These data demonstrated that variable expression of the hgpA-lacZ fusion occurred in H. influenzae and that the variable expression is dependent on the presence of the CCAA repeat region.

FIG. 3.

Phase variation between a non-LacZ-expressing phenotype and a LacZ-expressing phenotype in the hgpA-lacZ chromosomal fusion strain H. influenzae HI1718. A single white colony of HI1718 (not expressing LacZ) was grown overnight, and the culture was plated on sBHI plates. After the cultures were allowed to grow overnight, plates were flooded with X-Gal (5 mg/ml) and LacZ-expressing colonies were enumerated. The non-LacZ-expressing colonies appear gray, while the LacZ-expressing colonies appear black.

Frequency of variable expression of the hgpA-lacZ translational gene fusion in H. influenzae.

The frequency of alteration in expression of the hgpA-lacZ fusion was determined in H. influenzae HI1718. Because hemoglobin and hemoglobin-haptoglobin binding is heme repressible (4), the frequency of variation in expression of the hgpA-lacZ gene fusion in HI1718 under heme-replete and -depleted growth were determined (Table 3). Heme-depleted growth was performed in BHI supplemented with 10 μg of β-NAD per ml (heme-depleted BHI). Additional experiments were initiated with bacteria which had been passaged through hemoglobin-haptoglobin as previously described (11, 13). No significant difference (P > 0.05 by the Student t test) was found in the frequency of variation under different heme conditions for either blue-to-white or white-to-blue transitions. The overall frequency of variants, calculated based on results from all growth conditions, in H. influenzae HI1718 was 1.07% ± 0.09% for blue-to-white transitions and 0.60% ± 0.05% for white-to-blue transitions. A significant difference in rates between blue-to-white and white-to-blue transitions was detected (P < 0.03 by the Student t test): the frequency of blue-to-white transitions was approximately twice that of white-to-blue transitions. These findings are consistent with the slipped-strand hypothesis, since an in-frame gene may switch to either of two out-of-frame sequences, while an out-of-frame gene may switch to only one in-frame sequence.

TABLE 3.

Frequency of variable expression of the hgpA-lacZ translational gene fusion in H. influenzae HI1718

Growth conditions, transition, expt, and parameter	No. of colonies and frequency of phenotypic transition
	CFU with phenotype change	Total CFU	Frequency (%)
Heme-replete
Blue→White
Expt 1	76	5,976	1.27
Expt 2	58	6,780	0.86
Expt 3	194	11,096	1.75
Expt 4	116	11,308	1.03
Average (±SEM)			1.23 (±0.19)
White→blue
Expt 1	99	16,316	0.61
Expt 2	87	8,140	1.07
Expt 3	63	7,612	0.83
Expt 4	26	9,308	0.28
Average (±SEM)			0.70 (±0.17)
Heme-depleted
Blue→white
Expt 1	57	5,329	1.07
Expt 2	33	4,176	0.79
Expt 3	48	4,736	1.01
Expt 4	52	3,777	1.38
Average (±SEM)			1.06 (±0.12)
White→blue
Expt 1	49	5,680	0.86
Expt 2	17	4,364	0.39
Expt 3	42	10,108	0.42
Average (±SEM)			0.56 (±0.15)
Hemoglobin-haptoglobin-passaged
Blue→white
Expt 1	103	10,010	1.03
Expt 2	63	8,459	0.74
Expt 3	74	7,756	0.95
Average (±SEM)			0.91 (±0.09)
White→blue
Expt 1	54	8,246	0.65
Expt 2	160	21,489	0.74
Expt 3	65	27,201	0.24
Average (±SEM)			0.54 (±0.15)

Regulation of gene expression by slipped-strand mispairing: direct analysis of chromosomal DNA.

The NotI restriction sites engineered into the hgpA-lacZ fusion were used to directly determine the length of DNA in the CCAA region. Chromosomal DNA was digested with NotI and separated on a QuickPoint sequencing gel (6% [wt/vol] polyacrylamide–7 M urea) (NOVEX, San Diego, Calif.), and following transfer to a membrane, the blot was probed with a (CCAA)₆ oligonucleotide labeled by using the ECL 3′-oligonucleotide labeling kit (Amersham Pharmacia Biotech). Following development of the blot, hybridizing bands were sized by comparison to a radioactive sequencing reaction run on the same sequencing gel. The length of the CCAA repeat region was followed over several generations of phenotypic switching. A single band from each variant hybridized to the probe, and variable numbers of CCAA repeats were noted in successive generations (Fig. 4). All of the changes in the lengths of the fragments resulted from elongation or contraction by 4 or 8 bp, corresponding to an increase or reduction in the length of the CCAA repeats of one or two tetranucleotide repeats (Fig. 4). In each case, the length of the DNA fragment in a variant expressing β-galactosidase activity was consistent with an in-frame gene, and the CCAA length in a variant lacking β-galactosidase activity was consistent with an out-of-frame gene. These data demonstrate that changes in the length of the CCAA region are associated with phase variation of the hgpA-lacZ fusion.

FIG. 4.

Southern analysis of the CCAA repeats of the hgpA-lacZ translational gene fusion in serial generations, and association of CCAA repeat length with phenotypic changes. Chromosomal DNA from 12 variants of the hgpA-lacZ fusion-containing H. influenzae HI1718 was digested with NotI to excise the CCAA repeat-containing region of the hgpA-lacZ fusion. The first NotI site is 34 bp upstream of the first CCAA repeat; the second NotI site is 9 bp downstream of the last CCAA repeat. The DNA fragments were probed with the labeled (CCAA)₆ oligonucleotide. The numbers and the known sequence indicate the sizes of the fragments (in base pairs) as determined from the DNA sequencing reaction shown to the left of the blot. The generation number, phenotype, and number of CCAA repeats corresponding to the fragment lengths are indicated at the bottom of the figure. Phenotype abbreviations: B, blue colony (expressing LacZ); W, white colony (not expressing LacZ).

To further investigate the relationship between the CCAA region and blue-white phase variation of the fusion construction, the CCAA repeat regions of 33 variants of H. influenzae HI1718 that had undergone phase variation were amplified by PCR, using primers Pprof and Placr (Table 2). The amplified DNA was purified by using the QIAquick PCR purification kit (Qiagen, Valencia, Calif.) and directly sequenced (ABI model 373A; Recombinant DNA/Protein Resource Facility, Oklahoma State University, Stillwater). In each case, phase transitions between blue and white colony types were accompanied by changes in the number of CCAA repeats. Of the 28 switches involving a single CCAA repeat, 17 were subtractions and 11 were additions. Two switches were accompanied by subtraction of two CCAA units, and one switch was accompanied by addition of two CCAA units. No other sequence alterations were detected (data not shown).

The data presented here demonstrate that the number of CCAA repeats associated with hgpA changes in H. influenzae and that the alteration in the CCAA repeat length is associated with variable expression of the encoded protein.

An analogous mechanism has been demonstrated to mediate phase variation of H. influenzae lipooligosaccharide (LOS). Strand slippage across a CAAT repeat motif places potential initiation codons in or out of frame with the remainder of the ORF, leading to variable expression of the gene from different start codons (12, 15, 18). Unlike the expression of LOS, expression of HgpA occurs in only one frame, so that the alteration in the CCAA repeat region would lead to a direct on-and-off switch rather than a modulation in expression levels as seen in the case of LOS. High et al. demonstrated that the CAAT repeat motif is required for phase variation but not for biosynthesis of LOS (5). Similarly, the hgpA structural gene lacking the CCAA repeats was expressed in E. coli, and the recombinant E. coli bound both hemoglobin and the hemoglobin-haptoglobin complex (6).

In addition to the CCAA-mediated phase variation, hemoglobin binding activity is also heme repressible (4). It is unclear what advantage might be gained by the regulation of hemoglobin binding proteins by both phase variation and heme levels. It has been proposed that phase variation of a hemoglobin binding protein in gonococci might enable efficient utilization of menstrual hemoglobin (2). Such a mechanism may modulate expression of proteins in different host sites, depending on the prevalent heme source, enabling the infecting organism to adapt to specific host microenvironments. Alternatively, strand slippage may provide a mechanism to avoid the immunological response of the host. Through phase variation of multiple epitopes, the bacterium can express a diverse, but limited, number of different surface structures. Phase variants expressing specific epitopes have enhanced virulence in an animal model (9) and are selected for or induced in the course of systemic infections of humans (18). Thus, heme repression and phase variation of hemoglobin binding may conserve metabolic energy, enhance evasion of host immune response, or allow adaption to heme sources.

In conclusion, we have used an hgpA-lacZ translational gene fusion to investigate the association of CCAA repeats with phase variation of the hemoglobin- and hemoglobin-haptoglobin binding proteins in H. influenzae. Phase-variable expression of an hgpA-lacZ fusion was observed for individual colonies and was associated with changes in the number of CCAA repeats. The characteristics of the phase variation process were consistent with a model in which the expression of HgpA and other CCAA repeat-containing genes is regulated by slipped-strand mispairing during DNA replication, resulting in the introduction of frameshift mutations.