Metabolic Analysis of Moraxella catarrhalis and the Effect of Selected In Vitro Growth Conditions on Global Gene Expression

Abstract

The nucleotide sequence from the genome of Moraxella catarrhalis ATCC 43617 was annotated and used both to assess the metabolic capabilities and limitations of this bacterium and to design probes for a DNA microarray. An absence of gene products for utilization of exogenous carbohydrates was noteworthy and could be correlated with published phenotypic data. Gene products necessary for aerobic energy generation were present, as were a few gene products generally ascribed to anaerobic systems. Enzymes for synthesis of all amino acids except proline and arginine were present. M. catarrhalis DNA microarrays containing 70-mer oligonucleotide probes were designed from the genome-derived nucleotide sequence data. Analysis of total RNA extracted from M. catarrhalis ATCC 43617 cells grown under iron-replete and iron-restricted conditions was used to establish the utility of these DNA microarrays. These DNA microarrays were then used to analyze total RNA from M. catarrhalis cells grown in a continuous-flow biofilm system and in the planktonic state. The genes whose expression was most dramatically increased by growth in the biofilm state included those encoding a nitrate reductase, a nitrite reductase, and a nitric oxide reductase. Real-time reverse transcriptase PCR analysis was used to validate these DNA microarray results. These results indicate that growth of M. catarrhalis in a biofilm results in increased expression of gene products which can function not only in energy generation but also in resisting certain elements of the innate immune response.

Moraxella catarrhalis is a gram-negative, unencapsulated bacterium that can colonize the mucosal surface of the human nasopharynx, most frequently in infants and very young children (22). When this organism traverses the eustachian tube in these very young individuals, it can cause otitis media (7). Alternatively, in colonized adults, M. catarrhalis gains access to the bronchi and there causes exacerbations of chronic obstructive pulmonary disease (48). This organism can also infrequently cause other types of infections (for reviews, see references 39 and 73).

Information about the virulence mechanisms employed by this organism in the production of disease is still very limited, although a number of putative virulence factors have been identified, including proteins located in or attached to the outer membrane (1, 6, 9, 23, 24, 26, 32-34, 40, 44, 49, 51, 54, 57, 65, 71), as well as lipooligosaccharide (42, 59). Validation of the actual involvement of these different gene products in disease processes has been severely hindered by the lack of an appropriate animal model for M. catarrhalis disease (39). Similarly, little is known about how M. catarrhalis colonizes the nasopharynx, and while several M. catarrhalis adhesins which function in vitro have been identified (33, 40, 61, 71), the relative importance of these macromolecules in the colonization process in vivo remains to be determined. Recent studies aimed at addressing this issue have included the use of reverse transcriptase PCR (RT-PCR) to detect mRNA species expressed by M. catarrhalis in the nasopharynx of young children (46). Finally, little is known about the regulation of gene expression in M. catarrhalis, although recent studies indicate that both phase variation (41, 47, 57, 64) and temperature (30) can affect expression of some M. catarrhalis proteins.

Nasopharyngeal colonization by other pathogens that cause otitis media, especially Streptococcus pneumoniae and Haemophilus influenzae, has been studied in considerable detail, and bacterial gene products likely involved in this process have been identified through the use of relevant animal models (17, 63, 76). In the nasopharynx, it is reasonable to assume that M. catarrhalis, similar to other bacteria that colonize the nasopharynx (37, 52), exists in a biofilm, and it was recently reported that M. catarrhalis, like both H. influenzae and S. pneumoniae, can form a biofilm on the middle ear mucosa of children with otitis media (29). At least for H. influenzae, there have been extensive studies that have identified H. influenzae gene products involved in or affected by growth in the biofilm state in vitro (27, 50, 52, 68, 77). Whether there may be preferential expression of certain gene products in the biofilm state remains to be determined for M. catarrhalis, but recent studies with Neisseria meningitidis, another gram-negative pathogen that often colonizes the nasopharynx and can form biofilms in vitro (78), indicates that certain genes are expressed or up-regulated in response to contact with human cells in vitro (28). To date, however, there have been only a few reports describing biofilm formation by M. catarrhalis (13, 58), and only one of these involved a continuous-flow biofilm system (13).

In this study, we used the available nucleotide sequence from the genome of M. catarrhalis ATCC 43617 to facilitate investigation of different aspects of this pathogen. First, some of the basic metabolic capabilities and limitations of M. catarrhalis were inferred from the predicted gene products. Second, we developed and validated an M. catarrhalis DNA microarray for measuring global gene expression by this organism and then used this microarray to identify changes in gene expression and potential metabolic changes resulting from growth of M. catarrhalis in a continuous-flow biofilm system in vitro.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

M. catarrhalis strains ATCC 43617 and ATCC 25238 were obtained from the American Type Culture Collection, Manassas, VA. M. catarrhalis strain ETSU-9 was obtained from Steven Berk, Quillen-Disher College of Medicine, East Tennessee State University, Johnson City. The base medium employed in this study was brain heart infusion (BHI) (Difco, Detroit, MI) medium, and broth cultures were incubated at 37°C with aeration. BHI medium was supplemented with kanamycin (15 μg/ml) or spectinomycin (15 μg/ml) when appropriate. All BHI agar plates were incubated at 37°C in an atmosphere containing 95% air and 5% CO₂.

Growth of bacteria under iron-limiting conditions.

Iron limitation was achieved by adding the chelator deferoxamine mesylate (Desferal; Novartis, East Hanover, NJ) to BHI medium (3). Planktonic growth for iron limitation studies was obtained by inoculating bacterial growth from a BHI agar plate (grown overnight) into 20 ml of BHI broth (with or without Desferal) in a 500-ml flask. The cultures were grown at 37°C with shaking (200 rpm). Cell growth was measured by use of a Klett-Summerson photoelectric colorimeter (VWR International, West Chester, PA). For Western blot analysis, whole-cell lysates were prepared from bacteria grown in broth for 6 h. Briefly, the cells were pelleted by centrifugation and then resuspended in phosphate-buffered saline to a density of 300 Klett units. A 5-ml portion of the suspension was then subjected to centrifugation, and the resultant cell pellet was used to prepare a whole-cell lysate as described previously (56). Western blot analysis was performed using the M. catarrhalis CopB protein-specific monoclonal antibody 10F3 (31).

Biofilm growth system.

The Sorbarod cellulose filter-based continuous-flow system (13) for growing M. catarrhalis biofilms was used essentially as described previously (57). Bacterial cells from a stationary-phase BHI broth culture were used as the inoculum for the Sorbarod system. Briefly, a 3-ml portion of this culture was added to the cellulose filter, and then the flow of medium into the system was initiated. After 3 days of growth, the system was disassembled and the layer of bacterial growth that extended upward along the silicone tubing away from the cellulose filter was harvested for RNA extraction. Bacterial growth within or adherent to the Sorbarod filter itself was not used in the experiments. To obtain planktonic cells for RNA extraction, cells grown to the mid-logarithmic phase in BHI broth were harvested.

Development of an M. catarrhalis DNA microarray.

M. catarrhalis strain ATCC 43617 was originally derived from a patient with chronic bronchitis. A total of 1.9 Mb of nucleotide sequence from the genome of M. catarrhalis ATCC 43617 (World patent WO0078968) was obtained from the European Network of Patent Databases (http://ep.espacenet.com) as 41 different contigs. The same contigs can also be found at NCBI (GenBank accession numbers AX067426 to AX067466). The genome of M. catarrhalis strain ATCC 25238 was estimated to contain 1.75 to 1.94 Mb (25, 53), and therefore, the available nucleotide sequence from M. catarrhalis ATCC 43617 likely represents most if not all of this strain's genome. GLIMMER 2.02 software, obtained from The Institute for Genomic Research, was used to predict the open reading frames (ORFs) in the contigs (18). The 41 contigs were concatenated into one contiguous sequence (see Table S1 in the supplemental material) for use with the GLIMMER program. To minimize the possibility that GLIMMER would erroneously predict false ORFs spanning the contig junctions, a nucleotide sequence (5′-TTAACTAACTAG-3′) containing translation termination codons in all possible reading frames was placed between each pair of contigs. It should be noted that this assembly did not reflect the actual order of the DNA fragments in the M. catarrhalis chromosome and was done only to permit annotation. GLIMMER identified 1,761 putative ORFs (encoding proteins with at least 50 amino acids) in this contiguous sequence; these ORFs were translated using the standard genetic code and annotated by aligning the amino acid sequences against the nonredundant protein database at NCBI using BLASTP (see Table S2 in the supplemental material). The 1,761 annotated ORFs were used to design 70-mer oligonucleotides that would specifically anneal to each ORF. Because genome-directed primers (10, 19) that annealed within the 3′ 30% of each ORF were used to amplify the mRNA, each 70-mer probe was designed to anneal to a region within the 5′ 70% of each ORF. Following synthesis of the 70-mers (QIAGEN, Valencia, CA), the probes were spotted in triplicate on Corning UltraGAP II slides by Microarrays, Inc. (Nashville, TN). Each slide also contained three irrelevant 70-mers as negative controls.

RNA isolation.

RNA samples were isolated from planktonic or biofilm-derived cells of M. catarrhalis ATCC 43617 by using a QIAGEN RNeasy midi kit (QIAGEN) and following the manufacturer's protocol. For use in DNA microarray analysis, RNA samples were treated with QIAGEN RNase-free DNase. For real-time RT-PCR analysis, RNA samples were further treated with a MessageClean kit (GenHunter Corp., Nashville, TN) by following the manufacturer's protocol. For the DNA microarray experiments involving iron restriction, four independent experiments were performed to obtain four sets of RNA samples. For the biofilm-related DNA microarray experiments, five independent experiments were performed to obtain five sets of RNA samples.

Preparation of cDNA for DNA microarray hybridization.

A CyScribe postlabeling kit (GE Healthcare, Piscataway, NJ) was used according to the manufacturer's protocol, except that M. catarrhalis genome-directed primers (10, 19) were used to synthesize cDNA in the presence of amino allyl-dUTP. Twenty micrograms of total RNA and 3 μg of genome-directed primers were used for cDNA synthesis. Amino allyl-labeled cDNA was purified by using a QIAGEN QIAquick gel extraction kit. CyDye-labeled cDNA samples were purified by using Microcon YM30 columns (Millipore Corp., Bedford, MA). “Dye swap” experiments were done three times with cDNAs from the iron-related experiments and twice with cDNAs from the biofilm-related experiments. A NanoDrop spectrophotometer (NanoDrop Technologies, Wilmington, DE) was used to determine the concentration of nucleic acids in RNA, cDNA, and CyDye-labeled cDNA samples.

DNA microarray hybridization.

Cy3- and Cy5-labeled cDNA samples were mixed, lyophilized, and then dissolved in 14 μl of distilled H₂O (dH₂O). Following denaturation at 94°C for 3 min, these samples were kept on ice until they were used for DNA hybridization. A 40.5-μl mixture containing 10 μl of 4× hybridization buffer (GE Healthcare), 16 μl of formamide, 0.5 μl of yeast tRNA (10 mg/ml), and 14 μl of the mixture of Cy3- and Cy5-labeled cDNA was added to a DNA microarray slide and incubated at 50°C overnight (∼16 h). The slides were then washed twice in 6× SSPE containing 0.01% Tween 20 at 50°C for 5 min, twice in 0.8× SSPE containing 0.001% Tween 20 at 50°C, and twice in 0.8× SSPE at room temperature (1× SSPE is 0.18 M NaCl, 10 mM NaH₂PO₄, and 1 mM EDTA [pH 7.7]). The slides were dried by centrifugation at 900 × g in loosely capped 50-ml conical tubes for 3 min. Microarray slides were scanned by using a GenePix 4100A microarray reader (Molecular Devices, Sunnyvale, CA) and analyzed with GenePix and Acuity 4.0 software (Molecular Devices).

Data were processed using the Acuity 4.0 software package as follows. Raw data were first normalized using a ratio-based normalization method to equalize the means and medians of the features to 1 and exclude ratios that were less than 0.1 or greater than 10. Additionally, the features that were flagged by the software as “bad,” “absent,” or “not found” were also excluded from further analysis. The array provided results for 86.8 and 85.7% of the ORFs represented on the array for iron- and biofilm-related experiments, respectively. This indicates that the majority of the genome was being interrogated by the array. To identify the genes that exhibited an altered expression profile, we utilized a threshold of twofold-increased or -decreased expression over the selected baseline. In the case of the iron-limiting growth conditions, the iron-replete growth was used as the baseline, and in the biofilm experiments, the planktonic growth was used as the baseline. We then applied two additional constraints on the data: (i) the gene had to demonstrate the altered expression profile in at least five of the six DNA microarrays and (ii) a P value of <0.05 as calculated by the one-sample t test. The data were utilized to guide the selection of targets for real-time RT-PCR analysis, and in all cases real-time RT-PCR confirmed the trends observed in the microarray data, suggesting that the metrics used for analysis were valid.

Real-time qRT-PCR.

Oligonucleotide primer pairs for quantitative RT-PCR (qRT-PCR) (Table 1) were designed using PrimerExpress software (Applied Biosystems, Foster City, CA). Each 25-μl qRT-PCR mixture contained 12.5 μl of SYBR green PCR Master Mix (Applied Biosystems), 0.5 μl each of two oligonucleotide primers (from 2.5 μM stock solutions), 6.375 μl of dH₂O, 0.125 μl of MultiScribe reverse transcriptase (Applied Biosystems), and 5 μl of total RNA (5 ng/μl). Controls lacking reverse transcriptase or RNA template contained the appropriate amount of dH₂O in place of the enzyme or template. This one-step qRT-PCR method, including both the controls lacking reverse transcriptase and the no-template controls, was used to test freshly isolated RNA preparations for DNA contamination and the presence of “primer dimers,” respectively, using a 7500 real-time PCR system (Applied Biosystems) with the dissociation step. After RNA samples were tested for DNA contamination, one-step relative qRT-PCR (ΔΔC_T) was performed using the same machine. All relative qRT-PCR experiments involved the use of two independently isolated RNA preparations. The results were analyzed by using Relative Quantification Study software (Applied Biosystems).

TABLE 1.

Oligonucleotide primers used in this study

Primer	Oligonucleotide sequence (5′-3′)
MCORF1209-Fw	CTTTCTTGCCTCCGCCTTTA
MCORF1209-Rv	TGGTGGGAGAGTCATGATTGC
MCORF1234-Fw	TGGCAACAACAACTGGGTGA
MCORF1234-Rv	TGATTACCAGAAAGCTGAAGCACA
MCORF1377-Fw	CAGGCGGTATTGACAGTGCA
MCORF1377-Rv	CGGCATACACCCTATCAGAACC
MCORF139-Fw	GGGTGTTGACCGTTTGACAGT
MCORF139-Rv	CCCGTTGGTGAGATGATTTGT
MCORF141-Fw	TCGCGTTCTGGCTCAAGC
MCORF141-Rv	TGGTAAACGACGATAGATAGGCATT
MCORF1822-Fw	GCAGCCAATCAACGGAAGTAA
MCORF1822-Rv	GCTTAGCACGGCTGATGGTAA
MCORF1866-Fw	ATTCATCAGCGCGTCTGGA
MCORF1866-Rv	TTGTTGGCGTAATGACCTGC
MCORF231-Fw	CAGGCTTACAGCTTAAGGCCA
MCORF231-Rv	TGCACGCTTAGCCAAATCAC
MCORF292-Fw	CTACCTGAACGCCAAGGCA
MCORF292-Rv	CCGACAATTAAGAGAATTGGGAAT
MCORF328-Fw	GCTTTTGACCATCGAATCTGC
MCORF328-Rv	ACCGAGCGACCCTGTAAAGA
MCORF337-Fw	GCGTGTTGTCGGTCAGCTT
MCORF337-Rv	GTGCCAACTTTTGGGTTTGG
MCORF406-Fw	GGGCTTTAATACCATGCGTTATG
MCORF406-Rv	GGGCAGCCTCAAAGGACAC
MCORF660-Fw	GCATGGAACTCATCGCTCGT
MCORF660-Rv	AAGTGCCTGTACGCCTGTCG
MCORF679-Fw	GGTGAAGGCGATAAGAACCT
MCORF679-Rv	GCCAGTATTCCATGATGGCA
MCORF681-Fw	GCCAAAAAGCAGCACGGTA
MCORF681-Rv	GTTCTTGTGAGGCAGCATTATCTG
MCORF849-Fw	TCGCTGTGCTTGTGATTGC
MCORF849-Rv	CATAATCGCGGCAGGTGTT
MCORF923-Fw	CAATGTGCATGAACTGATTCCAT
MCORF923-Rv	TGCATTTGTTCGCCAAAGG
MCORF924-Fw	GATGGCATCGTGCTGATTGA
MCORF924-Rv	GGCAGCCTGAGATGCACAT
MCORF925-Fw	CTTGTCATGCAAAGCCGTCTT
MCORF925-Rv	GCCATCTGGTCAATTAAGGTGG
MCORF926-Fw	CGCCTTGACAGTGTGCATCT
MCORF926-Rv	GCCTGCACGCCATGTGTAT
WW191	CGGGATCCTGACTATTTTCCAAGA
WW192	CCTCTCGGCTGGTCCCGGGCGAGTGCGAGGATAGTC
WW193	GGACCAGCCGAGAGGGCATGGA
WW194	GATGGAGCTCATTGGCACGCCGTCTTGTA
WW195	GGGCTGCAGGCTCCGTCGATACTATGTTA
WW196	GACATCTAAATCTAGGTACTA

Construction of a narGH mutant.

An approximately 980-bp DNA fragment containing some upstream DNA and the extreme 5′ end of the M. catarrhalis narG gene (MCORF923) and an approximately 870-bp DNA fragment containing the majority of the narH gene (MCORF924) were PCR amplified by using the WW191-WW192 and WW193-WW194 oligonucleotide primer pairs (Table 1), respectively, with M. catarrhalis ATCC 43617 genomic DNA as the template. The two PCR amplicons were used in overlapping extension PCR (36) with oligonucleotide primers WW191 and WW194 (Table 1) to obtain a 1.8-kb PCR product that was subsequently digested with both BamHI and SacI. The DNA fragment was ligated into the M. catarrhalis cloning vector pWW115 (74), which itself had been digested with the same two restriction enzymes. This ligation reaction was used to electroporate M. catarrhalis ATCC 25238, and a plasmid purified from a spectinomycin-resistant recombinant was designated pWW119. The kan gene from pAC7 (75) was amplified by PCR using the WW195-WW196 primer pair (Table 1) and was ligated into the SmaI site within the cloned amplicon in pWW119. The resultant plasmid, pWW120, with the kan cartridge inserted in the same direction as the narG gene, was used as the template for a PCR with the WW191-WW194 primer pair to obtain a 2.9-kb PCR product containing the plasmid insert. This PCR product was used to electroporate M. catarrhalis ETSU-9, and transformants (i.e., putative narGH deletion mutants) were selected with kanamycin.

Evaluation of the biofilm formation ability of the ETSU-9 narGH mutant.

A competitive index approach was used to determine whether the narGH mutant had a reduced ability to form a biofilm in the Sorbarod continuous-flow biofilm system. Briefly, the kanamycin-resistant narGH mutant (3 × 10⁷ CFU) was mixed with a streptomycin-resistant mutant of the ETSU-9 parent strain (3 × 10⁷ CFU) and used to inoculate the Sorbarod system. After 3 days of growth, the biofilm was harvested as described above, and the resultant cell suspension was serially diluted and plated onto both BHI agar plates containing kanamycin and BHI agar plates containing streptomycin to determine the relative numbers of viable wild-type and mutant organisms present in the biofilm.

Microarray data accession numbers.

The raw data from the DNA microarray experiments were deposited in the NCBI Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/) under accession numbers GSE4346 (for the iron-limited growth studies) and GSE4348 (for the biofilm growth studies).

RESULTS

We deduced the metabolic capabilities of M. catarrhalis ATCC 43617 from the genomic sequence data for two reasons: to validate the completeness of our M. catarrhalis DNA microarrays for measuring global gene expression in this organism and to determine likely changes in metabolism, if any, that might result from growth in the biofilm state. To assess the completeness of the DNA microarray, we determined the probable functions of the ORF-encoded protein products, deduced the metabolic capabilities of the organism from these proposed functions, and compared these capabilities to known metabolic characteristics of M. catarrhalis (Table 2).

TABLE 2.

Predicted metabolic capabilities and limitations of M. catarrhalis ATCC 43617

Function, pathway, or proteins	Functional capability or status
Energy and carbohydrate metabolism
Glycolytic pathways	Incomplete Embden-Meyerhof-Parnas pathway (missing phosphofructokinase and pyruvate kinase); no components of Entner-Dourdoroff pathway; incomplete pentose cycle (missing both dehydrogenases, lactonase, and transaldolase)
Carbohydrate catabolic pathways	None
Carbohydrate transport systems	None (e.g., missing all components of phosphotransferase sugar transport systems)
Citric acid cycle	Missing only succinyl coenzyme A synthetase
Glyoxylate cycle	Complete
Anapleurotic pathways	Missing (e.g., no gene for phosphoenolpyruvate carboxylase)
Gluconeogenesis	Complete
Aerobic electron transport and ATP synthase	Apparently complete
Anaerobic functions	Nitrate, nitrite, and nitric oxide reductases (anaerobic respiration) and acetate kinase-phosphotransacetylase (substrate level phosphorylation)
Nitrogen metabolism
High-affinity ammonia assimilation	Missing high-affinity ammonia transport (AmtB) and all components of the Ntr system (PII, NtrB, NtrC, and σ54) for high-level glutamine synthetase production, but appears to have a ferrodoxin-dependent glutamate synthase
Low-affinity ammonia assimilation	Only one gene for a glutamate dehydrogenase homolog is present, and its function is uncertain given the apparent inability to assimilate ammonia
Amino acid synthesis	All pathways intact except for proline (missing ProA and ProB) and arginine (missing ArgC and ArgE); arginine is a growth requirement, but proline is not, possibly because of a potential pathway from arginine (via ornithine and glutamic-5-semialdehyde) catalyzed by a putative arginase, ornithine transaminase, and ProC; the ability to synthesize glutamate is uncertain in the absence of ammonia assimilation
Purine and pyrimidine synthesis	Both pathways appear to be intact; there is only one pathway of dTMP synthesis from dCTP (via dUTP and dUMP); no recognizable enzyme converts (deoxy)nucleoside diphosphates to (deoxy)nucleoside triphosphates (missing nucleoside diphosphokinase, pyruvate kinase, and succinyl coenzyme A synthetase), but such activity must be present for viability
Polyamines	Enzymes present for synthesis of putrescine but not for synthesis of spermidine; instead, the enzymes for norspermidine synthesis are present
Other biosynthetic pathways	Pathways for fatty acids, isoprenoids, three different phospholipids, riboflavin, flavin adenine mononucleotide, flavin adenine dinucleotide, biotin, NAD, and heme appear to be intact; pathways for ubiquinone, pyridoxal phosphate, and pantothenate are missing one or two enzymes; the latter pathways are probably intact, and the missing enzymes are probably not recognizable
Catabolic pathways	Limited catabolic capacity; there are recognizable pathways for the degradation of fatty acids, some amino acids (alanine, arginine, asparagine, glycine, histidine, proline, serine, threonine, and perhaps tyrosine and phenylalanine), γ-aminobutyrate, and lactic acid; there is no recognizable pathway for the catabolism of any carbohydrate
Sigma factors	Only σD and σH
Two-component systems	Seven response regulators, five sensor kinases, and only four apparent response regulator-sensor kinase pairs
Other transcriptional regulators	27 apparent regulators (excluding response regulators)

An unusual feature of the deduced metabolism of M. catarrhalis is its apparent inability to utilize any exogenous carbohydrate. This bacterium appears to lack intact glycolytic pathways, all carbohydrate catabolic enzymes, all carbohydrate transport systems, and all anapleurotic reactions (Table 2). This is consistent with data in Bergey's Manual of Determinative Bacteriology (35) which indicate that M. catarrhalis does not produce acid from glucose and does not utilize any carbohydrate tested (for a review, see reference 16). The inability to utilize exogenous carbohydrates suggests that gluconeogenesis is essential, and the apparent presence of all the gluconeogenic enzymes is consistent with this conclusion.

M. catarrhalis appears to possess all of the enzymes required for aerobic energy metabolism. Both subunits of succinyl coenzyme A synthetase are missing, but otherwise this bacterium has an intact citric acid cycle, glyoxylate cycle, electron transport system, and ATP synthase. The utilization of acetate by M. catarrhalis (35) is consistent with a functional glyoxylate cycle. Enzymes associated with anaerobic metabolism (nitrate reductase, nitrite reductase, nitric oxide reductase, and acetate kinase-phosphotransacetylase) are also present and perhaps could be used in a low-oxygen or microaerophilic environment. The reported utilization of nitrate by M. catarrhalis (16) is consistent with the presence of nitrate reductase.

The genome sequence also suggests several unusual aspects of nitrogen metabolism. First, M. catarrhalis appears to have both pathways for ammonia assimilation: a glutamate dehydrogenase pathway and a glutamate synthase-glutamine synthetase pathway. However, despite the presence of genes for these enzymes, it is not apparent that M. catarrhalis can assimilate ammonia. It lacks all regulators that are associated with nitrogen limitation in organisms such as Escherichia coli: RpoN, P_II, NtrB, and NtrC. Furthermore, the observation that an ammonium salt did not stimulate M. catarrhalis growth in a defined medium (38) implies that M. catarrhalis cannot assimilate even high ammonia concentrations. Although the genes for enzymes of ammonia assimilation are present, it is not clear that they function in this capacity. Their sole function may be in glutamate and glutamine synthesis, and they may be insufficiently expressed to participate in ammonia assimilation. Second, M. catarrhalis has intact pathways for the synthesis of all amino acids except proline and arginine. The latter finding is consistent with the report that M. catarrhalis is an arginine auxotroph (38). Proline auxotrophy has not been reported for M. catarrhalis. It is possible that proline can be synthesized from arginine (via ornithine), and all the enzymes required for this unconventional pathway appear to be present.

The deduced metabolic capabilities of M. catarrhalis suggest an unusual and unpredicted pattern of metabolism. Nonetheless, all of the conclusions described above are consistent with published observations. Most importantly, every pathway predicted to be complete is intact. For example, M. catarrhalis is a histidine prototroph, and all of the numerous different enzymes required for histidine synthesis are encoded by the genome. The same is true for processes such as protein synthesis, since we were able to identify all of the tRNA synthetases. Furthermore, the few pathways predicted to have deficiencies (e.g., arginine synthesis) are missing only one or two enzymes. Therefore, the oligonucleotide probes used in the M. catarrhalis DNA microarray described in this study account for the vast majority, if not all, of the genes encoding metabolic enzymes of M. catarrhalis and, by extrapolation, perhaps all of the genes of this pathogen.

Validation of the M. catarrhalis DNA microarray for use in transcriptome analysis.

To establish the utility of the DNA microarray developed by using the genome information from M. catarrhalis ATCC 43617, we first performed studies using this DNA microarray to analyze the transcriptomes from M. catarrhalis cells grown under iron-replete and iron-limiting conditions. Previous studies with other M. catarrhalis strains had shown that the chelator Desferal could be used to effectively limit iron availability in vitro (3), and the identities of several different M. catarrhalis proteins whose expression is affected by the availability of iron in the growth environment have been well established (3, 11, 14, 43, 44). To determine the concentration of Desferal necessary to effect iron-limited growth of M. catarrhalis strain ATCC 43617, this strain was grown in BHI broth in the presence of increasing concentrations of this chelator (Fig. 1A). It was found that 30 μM Desferal effectively limited growth of this strain (Fig. 1A). When protein expression by the cells was examined, it was found that the cells grown in the presence of both 30 and 50 μM Desferal exhibited increased expression of the CopB protein (Fig. 1B). Expression of CopB in other M. catarrhalis strains has previously been shown to be regulated by the availability of iron (3, 15). These results indicated that growth of M. catarrhalis ATCC 43617 in the presence of 30 μM Desferal was iron limited.

FIG. 1.

Effect of iron limitation on growth and protein expression by M. catarrhalis ATCC 43617. (A) Growth of strain ATCC 43617 in BHI broth containing various amounts (10 to 50 μM) of Desferal. (B) Western blot analysis of CopB protein expression by this strain grown in BHI medium in the presence of no Desferal (lane 1), 10 μM Desferal (lane 2), 30 μM Desferal (lane 3), and 50 μM Desferal (lane 4). Proteins in whole-cell lysates were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to a polyvinylidene difluoride membrane, and probed with the CopB-specific monoclonal antibody 10F3 (2). The positions of molecular weight markers are indicated on the left.

Having established iron-limiting growth conditions for this strain, we proceeded to extract total RNA from M. catarrhalis ATCC 43617 cells grown into logarithmic phase in BHI broth with and without 30 μM Desferal. DNA microarray analysis revealed that the expression of over 100 different genes was affected at least twofold by iron limitation (42 genes with increased expression and 65 genes with decreased expression [Table 3]) . The genes whose expression was most highly up-regulated by iron limitation, as determined by DNA microarray analysis, included lbpB, lbpA, tbpB, copB, and several other genes predicted to encode proteins involved in iron uptake (Table 3). The first four genes listed above have previously been shown to be up-regulated by iron limitation (3, 12, 14, 43). The majority of the genes whose expression was most adversely affected by iron limitation encoded ribosomal proteins (Table 3); this is likely due to the growth inhibition caused by iron limitation (Fig. 1A). These DNA microarray data demonstrating up-regulation of M. catarrhalis genes known to be affected by iron limitation indicated that this DNA microarray can be used to assess global gene expression by this pathogen.

TABLE 3.

Genes in M. catarrhalis ATCC 43617 whose expression was maximally affected by iron availability

ORF	Description	Median log2 DF30/DF0 ratioa	SD
Genes whose expression was increased by iron limitation
MCORF323	TonB-like protein	4.8335	1.0285
MCORF645	LbpB	4.1935	0.5723
MCORF1767	YfeB	3.7050	0.4739
MCORF1766	YfeA	3.5605	0.5591
MCORF647	LbpA	3.5130	0.9453
MCORF1768	AfeC	3.4285	0.4218
MCORF1769	AfeD	3.3130	0.6454
MCORF648	Unknown	3.0925	0.7858
MCORF324	MotA/TolQ/ExbB proton channel	3.0760	0.7270
MCORF73	tRNA-i(6)A37 modification enzyme MiaB	3.0610	0.8975
MCORF325	ExbD/TolR	3.0525	0.9187
MCORF1696	Malate synthase	2.8445	1.6233
MCORF830	Fumarase C	2.7550	0.8955
MCORF755	TbpB	2.7435	0.9448
MCORF1916	Predicted periplasmic lipoprotein involved in iron transport	2.5790	1.2184
MCORF903	Transcriptional regulator, BadM/Rrf2 family	2.4505	0.1916
MCORF1917	Putative iron-dependent peroxidase	2.4000	0.3890
MCORF168	ABC transporter, iron-binding protein	2.2040	0.7090
MCORF1528	Hypothetical protein	1.9660	0.2547
MCORF1068	Peptidoglycan-binding LysM	1.9625	0.5003
MCORF906	IscA	1.8990	0.4637
MCORF167	Binding protein-dependent transport system component	1.8265	0.3555
MCORF1063	Hypothetical protein	1.7755	0.2633
MCORF1658	Peptidase M16-like	1.7480	0.7444
MCORF904	Cysteine desulfurase IscS	1.7325	0.2654
MCORF643	Hypothetical protein	1.6485	0.3732
MCORF905	FeS cluster assembly scaffold IscU	1.6480	0.3590
MCORF1164	Hypothetical protein	1.5500	0.3127
MCORF166	ABC transporter ATP-binding protein	1.4860	0.7503
MCORF345	Carboxyl-terminal protease	1.4305	0.2837
MCORF186	Hypothetical protein	1.4250	0.3438
MCORF1725	PepSY-associated transmembrane helix	1.3790	0.6800
MCORF1657	Probable insulinase-like peptidase	1.3770	0.5831
MCORF1261	CopB major outer membrane protein	1.3565	0.8131
MCORF763	LolA	1.3465	0.2705
MCORF1715	Acetoin reductase	1.2830	0.4801
MCORF1915	Hypothetical protein	1.2630	0.3662
MCORF1572	Nitroreductase	1.2555	0.4565
MCORF389	Phospholipid/glycerol acyltransferase	1.2550	0.1812
MCORF1452	DNA topoisomerase IV subunit B	1.2325	0.3288
MCORF46	Lytic transglycosylase	1.1720	0.3471
MCORF1348	Polymerase and histidinol phosphatase-like	1.1510	0.5350
Genes whose expression was decreased by iron limitation
MCORF114	Cytochrome c, class I	−3.9170	0.8080
MCORF519	Phosphate transport system permease protein 1	−3.1305	0.6570
MCORF1569	Fumarate hydratase	−2.9610	1.4808
MCORF517	Phosphate ABC transporter permease PstC	−2.9260	0.8717
MCORF3	Fumarate hydratase	−2.8485	1.3297
MCORF382	Hypothetical protein	−2.5205	0.5313
MCORF1139	Catalase, KatA	−2.4630	2.0642
MCORF12	Bacterioferritin	−2.3860	0.9915
MCORF735	Putative electron transfer flavoprotein	−2.3750	0.4571
MCORF1229	4Fe-4S ferredoxin, iron-sulfur binding	−2.3110	0.7078
MCORF1862	Succinate dehydrogenase cytochrome b566 subunit	−2.2700	0.7239
MCORF1865	Succinate dehydrogenase iron-sulfur protein	−2.1875	0.3394
MCORF1198	Cytochrome b/b6-like	−2.1450	0.1966
MCORF1390	GTP diphosphokinase	−2.0810	0.6494
MCORF1232	Cytochrome c oxidase ccb3 type, subunit I	−2.0315	0.2444
MCORF192	Cytochrome c, class II	−2.0230	0.5614
MCORF1199	Ubiquinol-cytochrome c reductase, iron-sulfur subunit	−1.9810	0.1912
MCORF1597	Outer membrane protein E	−1.9320	0.4192
MCORF1864	Succinate dehydrogenase	−1.9030	0.3475
MCORF516	Periplasmic phosphate-binding protein	−1.8940	0.6549
MCORF381	Hypothetical protein	−1.8840	1.1530
MCORF1863	Succinate dehydrogenase, hydrophobic subunit	−1.8655	0.3632
MCORF338	Ribosomal protein L10	−1.8320	0.3041
MCORF1233	Cytochrome c oxidase, monoheme subunit	−1.8110	0.6055
MCORF679	Nitric oxide reductase	−1.7770	0.5404
MCORF1234	Cytochrome c oxidase cbb3 subtype, subunit III	−1.7405	0.2262
MCORF1672	Protein of unknown function	−1.7260	0.3958
MCORF765	NADH dehydrogenase beta subunit	−1.7240	0.3852
MCORF874	Metal ion (Mn2+/Fe2+) transporter, Nramp family	−1.6600	0.5657
MCORF1874	Outer membrane protein P2 homolog	−1.6550	1.0270
MCORF746	GTP-binding protein, HSR1 related	−1.6540	0.7837
MCORF1161	Probable cytochrome c	−1.6480	0.2617
MCORF990	Oligopeptide permease	−1.6425	0.2604
MCORF337	50S ribosomal protein L1	−1.5560	0.5726
MCORF766	NADH dehydrogenase I chain C,D	−1.5520	0.5323
MCORF985	Oligopeptide permease	−1.5330	0.2785
MCORF1049	Outer membrane porin M35	−1.5300	0.4881
MCORF1722	Hypothetical protein	−1.5255	0.3334
MCORF1022	Ribosomal protein L9	−1.4830	0.3460
MCORF336	50S ribosomal protein L11	−1.4775	0.2523
MCORF137	Ribosomal protein S8	−1.4670	0.4295
MCORF180	GTP cyclohydrolase I	−1.4550	0.3349
MCORF975	MchA1	−1.4550	0.2103
MCORF1435	Aconitate hydratase	−1.4525	0.5169
MCORF1207	Hypothetical protein	−1.4290	0.3590
MCORF132	30S ribosomal protein S17	−1.4250	0.2534
MCORF880	Superoxide dismutase	−1.4145	0.8468
MCORF94	Sodium/sulfate symporter	−1.3790	0.4101
MCORF1432	50S ribosomal protein L19	−1.3680	0.1125
MCORF356	Ribosomal protein S2	−1.3660	0.2583
MCORF175	Cytochrome c, class I	−1.3570	0.4286
MCORF1031	Probable glutaredoxin	−1.3550	0.2498
MCORF131	Ribosomal protein L16	−1.3510	0.1795
MCORF1143	NADH:ubiquinone oxidoreductase, Na+ translocating, D subunit	−1.3500	0.2102
MCORF145	Ribosomal protein S4	−1.3450	0.1847
MCORF984	OppB	−1.3165	0.1086
MCORF146	DNA-directed RNA polymerase alpha subunit	−1.3000	0.2361
MCORF1147	Na-translocating NADH-quinone reductase subunit A	−1.2930	0.2667
MCORF129	50S ribosomal protein L22	−1.2410	0.2266
MCORF130	30S ribosomal protein S3	−1.2230	0.2849
MCORF1845	Dihydrodipicolinate reductase	−1.2000	0.2834
MCORF1506	Hypothetical protein	−1.1740	0.3589
MCORF136	30S ribosomal protein S14	−1.1380	0.3511
MCORF980	MhaB1	−1.1170	0.3049
MCORF147	Ribosomal large subunit protein L17	−1.1110	0.2632

^aMedian log₂ ratio of expression with 30 μM Desferal (DF30) to expression with no Desferal (DF0) from six experiments (P < 0.02).

DNA microarray analysis of genes affected by growth in a biofilm.

M. catarrhalis ATCC 43617 was grown in BHI broth into the mid-logarithmic phase of growth or in the Sorbarod biofilm system. Total RNA extracted from both the broth-grown and biofilm-grown cells was subjected to DNA microarray analysis. For 83 genes, expression was affected at least twofold by growth in the biofilm state relative to growth in the planktonic state (54 genes with increased expression and 29 genes with decreased expression [Table 4]) . The genes whose expression was increased the most by growth in a biofilm were predicted to encode components of the nitrate respiratory chain (Table 4). These genes included genes encoding enzymes involved in nitrate reduction, a nitrite reductase, and a nitric oxide reductase. Other up-regulated genes included genes predicted to encode the Lon protease and DnaK (Table 4). The genes whose expression was most reduced by growth in a biofilm were mainly the genes predicted to encode ribosomal proteins (Table 4).

TABLE 4.

M. catarrhalis ATCC 43617 genes whose expression was maximally affected by growth in a biofilm

ORF	Description	Median log2 biofilm/planktonic ratioa	SD
Genes whose expression was increased by growth in a biofilm
MCORF924b	Respiratory nitrate reductase, beta subunit; NarH	6.1430	1.9179
MCORF925b	Respiratory nitrate reductase, delta subunit; NarJ	5.1485	1.4230
MCORF926b	Respiratory nitrate reductase, gamma subunit; NarI	4.8830	1.1844
MCORF923b	Respiratory nitrate reductase, alpha subunit; NarG	4.4220	1.8378
MCORF942	Nitrate/proton symporter	4.3815	0.7894
MCORF1192	ATPase AAA-2	3.9065	1.0013
MCORF943	Nitrate transporter	3.1795	0.7592
MCORF529	N5,N10-Methylene tetrahydromethanopterin reductase	3.1630	0.9232
MCORF682	Hypothetical protein	3.0900	1.0005
MCORF849b	Heavy metal-translocating P-type ATPase	3.0825	0.4627
MCORF681b	Probable nitrite reductase	3.0225	0.6060
MCORF660b	Acyl coenzyme A dehydrogenase-like	3.0225	0.3881
MCORF550	No annotation	2.3560	1.1989
MCORF662	Acyl coenzyme A dehydrogenase-like	2.2750	0.6095
MCORF930	Molybdate transport permease protein	2.2150	1.0878
MCORF712	Excinuclease ABC subunit A	2.1670	0.8658
MCORF1209b	Exonuclease SbcD	2.1565	0.4844
MCORF1822b	DnaK	2.1550	0.7948
MCORF1116	Chaperonin Cpn60/TCP-1	2.1290	0.7713
MCORF783	Putative peroxiredoxin/glutaredoxin family protein	2.1165	0.5776
MCORF2	Probable organic hydroperoxide resistance protein	2.1020	0.2478
MCORF679b	Nitric oxide reductase	2.0430	0.3355
MCORF1437	Urocanate hydratase	2.0240	0.6063
MCORF1511	HSP33 family protein	1.9650	0.6120
MCORF1887	Acetyl coenzyme A acetyltransferase	1.9635	0.8506
MCORF717	Glycosyl transferase	1.9510	0.4526
MCORF1877	Oligopeptidase A	1.9450	0.7186
MCORF406b	3-Isopropylmalate dehydrogenase	1.9250	0.4506
MCORF1440	Urocanate hydratase	1.8805	0.9007
MCORF292b	Peptidase M41, FtsH	1.8795	0.4735
MCORF910	Hypothetical protein	1.8440	0.9557
MCORF328b	Lon protease	1.7705	0.6108
MCORF387	Heat shock protein 90	1.7240	0.7001
MCORF932	Hypothetical protein	1.6955	0.8780
MCORF1434	Aconitate hydratase	1.6775	0.4724
MCORF1086	Helicase-like	1.6640	0.2485
MCORF163	Putative alkyl hydroperoxide reductase, subunit C	1.6110	0.6280
MCORF294	Peptidase M41, FtsH	1.6040	0.5143
MCORF1318	Amino acid carrier protein	1.5425	0.6691
MCORF1031	Probable glutaredoxin	1.4995	0.4265
MCORF1210	Hypothetical protein	1.4950	0.2024
MCORF510	Thioredoxin	1.4865	0.3328
MCORF720	CcmE/CycJ protein	1.4570	0.7565
MCORF1783	AAA ATPase, central region	1.4065	0.4982
MCORF293	Peptidase M41, FtsH	1.3825	0.7000
MCORF34	Modification methylase HgiDII	1.3750	0.3551
MCORF1030	Protein export protein SecB	1.3580	0.1776
MCORF1712	HesB/YadR/YfhF	1.3015	0.4287
MCORF692	GTP pyrophosphokinase	1.2690	0.1693
MCORF1322	Protein of unknown function	1.2530	0.1690
MCORF114	Cytochrome c, class I	1.2015	0.2578
MCORF1441	Histidine ammonia lyase	1.1315	0.3983
MCORF688	Cell division protein-like	1.0955	0.5216
MCORF1784	3,4-Dihydroxy-2-butanone 4-phosphate synthase	1.0905	0.5389
Genes whose expression was decreased by growth in a biofilm
MCORF137	Ribosomal protein S8	−1.8495	0.4193
MCORF93	Hypothetical protein	−1.6800	0.5717
MCORF753	Hypothetical protein	−1.6700	0.5817
MCORF1722	Hypothetical protein	−1.6250	0.2655
MCORF125	Ribosomal protein L4	−1.5870	0.2885
MCORF141b	Ribosomal protein L15	−1.5300	0.2985
MCORF126	Ribosomal protein L23	−1.5190	0.5287
MCORF519	Phosphate transport system permease protein 1	−1.5170	0.3775
MCORF132	Ribosomal protein S17	−1.4620	0.4765
MCORF146	DNA-directed RNA polymerase alpha subunit	−1.4115	0.3918
MCORF138	Ribosomal protein L6	−1.4050	0.2727
MCORF1219	Phosphate transporter	−1.4000	0.2164
MCORF127	Ribosomal protein L2	−1.3870	0.3832
MCORF337b	Ribosomal protein L1	−1.3710	0.5238
MCORF1054	Aspartate semialdehyde dehydrogenase	−1.3495	0.7263
MCORF139b	Ribosomal protein L18	−1.3340	0.2324
MCORF985	Oligopeptide permease	−1.3150	0.6237
MCORF129	Ribosomal protein L22	−1.2895	0.3433
MCORF1127	Phosphate acetyltransferase	−1.2860	0.3139
MCORF145	Ribosomal protein S4	−1.2360	0.3168
MCORF1201	Ribosomal protein L13	−1.2285	0.5406
MCORF417	SecF	−1.2160	0.4570
MCORF336	Ribosomal protein L11	−1.2095	0.4956
MCORF130	Ribosomal protein S3	−1.1645	0.5656
MCORF1882	ATP-dependent Clp protease, ClpX	−1.1615	0.3734
MCORF1016	Aminomethyltransferase	−1.1615	0.5637
MCORF133	Ribosomal protein L14	−1.1230	0.3314
MCORF131	Ribosomal protein L16	−1.1065	0.3645
MCORF1110	RNase BN	−1.1000	0.3265
Genes whose expression was unaffected by growth in a biofilmc
MCORF231b	Glyceraldehye-3-phosphate dehydrogenase	0.1775	0.1248
MCORF1234b	Cytochrome c oxidase cbb3 type, subunit III	0.0850	0.0607
MCORF1866b	2-Oxoglutarate dehydrogenase, E1 component	−0.0700	0.1864
MCORF1377b	NAD synthetase	−0.1180	0.1308

^aMedian log₂ ratio of expression in a biofilm to expression in a planktonic state from six experiments (P < 0.02).

^bORF whose expression was also measured by real-time RT-PCR.

^cRepresentative genes from this study (P > 0.05).

Validation of DNA microarray data.

Real-time qRT-PCR analysis was used to confirm the validity of the DNA microarray results (Table 4). Twenty different M. catarrhalis ATCC 43617 genes were selected for this analysis. These genes included 13 genes whose expression appeared to be up-regulated by growth in a biofilm, as measured by DNA microarray analysis, 3 genes that appeared to be down-regulated, and 4 genes representative of the genes that were apparently unaffected (Table 4). The values obtained by qRT-PCR correlated well with those obtained from DNA microarray analysis (Fig. 2).

FIG. 2.

Use of real-time RT-PCR to verify DNA microarray results. Expression of 20 selected genes in M. catarrhalis ATCC 43617 cells grown in a biofilm and in the planktonic state was measured by real-time RT-PCR analysis as described in Materials and Methods. The logarithm of the difference in gene expression between biofilm and planktonic cells as determined by real-time RT-PCR is plotted adjacent to the results obtained in the DNA microarray analysis.

Construction of an M. catarrhalis narGH mutant.

The M. catarrhalis genes that were proportionally up-regulated the most by growth in the Sorbarod biofilm system included narG, narH, narI, and narJ (Table 4). In other organisms, including E. coli (72) and Mycobacterium tuberculosis (66), these four genes form an operon or are grouped together. In M. catarrhalis ATCC 43617, these genes are similarly located together, and RT-PCR indicated that these four ORFs are cotranscribed in M. catarrhalis to form a single mRNA transcript (data not shown). To determine whether lack of expression of the narGHIJ operon affected the ability of M. catarrhalis to form a biofilm, the narG and narH genes were inactivated in M. catarrhalis strain ETSU-9. The ETSU-9 strain was used for mutant construction because we have been unsuccessful to date in constructing mutants of any type in M. catarrhalis ATCC 43617 (data not shown). A narGH mutant was constructed by deleting the majority of the narG ORF and the 5′ end of the narH gene and replacing this DNA with a kan cartridge. When tested for its ability to form a biofilm in the Sorbarod continuous-flow biofilm system, however, the mutant did not exhibit any apparent deficiency in biofilm formation (data not shown).

DISCUSSION

The deduced metabolism of M. catarrhalis was used to assess the completeness of the available nucleotide sequence from the genome of M. catarrhalis ATCC 43617 and to provide a context for interpreting transcriptome profiles. In addition, a comparison of the metabolism of M. catarrhalis with that of two other gram-negative colonizers of the nasopharynx, H. influenzae and N. meningitidis, has the potential to provide some insight into metabolic approaches used by these three organisms to survive in the environment of the nasopharynx. Table 5 presents some key metabolic features of these three pathogens, with an emphasis on significant differences.

TABLE 5.

Comparison of some of the metabolic properties of M. catarrhalis, H. influenzae, and N. meningitidis^a

Species	Genome size (Mb)	No. of ORFs	Glycolytic pathways	Phosphotransferase components	Glycogen synthesis	Citric acid cycle	Glyoxylate cycle	Anaerobic metabolism	Anaerobic metabolism involving nitrogen compounds	High-affinity ammonia assimilation	Low-affinity ammonia assimilation	Primary triamine	Glutathione	Growth requirements	Sigma factors	No. of response regulators	No. of sensor kinases
M. catarrhalis	∼1.9	1,761	None	None	No glg genes	Missing sucCD	Yes	Acetate kinase-phosphotrans acetylase	Nitrate reductase, nitrite reductase, nitric oxide reductase	No, missing Ntr components	No, but has gdhA homologd	Norspermidinee	?	Arginine, glutamate?	rpoD, rpoH	7	5
H. influenzae	1.8	1,709	Embden-Meyerhof-Parnas, Entner-Doudoroff, pentose cycle	ptsH, ptsI, ptsN, crr, fruA, fruB	Yes, has glgA, glgB, glgC, glgP, and glgX	Missing gltA, icd, acnA, acnB, sdhABCD, fumABc	No	Acetate kinase- phosphotrans acetylase, formate dehydrogenase, fumarate reductase, others	Nitrate reductase, nitrite reductase	No, missing Ntr componentsc	Unknown, has gdhA homolog	Norspermidinee	Noc	Arginine, uracil, NAD, hemec	rpoD, rpoH	6f	4f
N. meningitidisb	2.2	2,121 and 2,158g	Embden-Meyerhof-Parnas, Entner-Doudoroff, pentose cycle	ptsH, ptsI, ptsN, single IIAB component	No glg genes	Intact	No	Acetate kinase-phosphotrans acetylase	Nitrite reductase, nitric oxide reductase	No, missing Ntr components	Unknown, has gdhA homolog	Uncertain, could not find spermidine synthase	Yes	Asparagine, methionine, folic acid, pantothenate, pyridoxal	rpoD, rpoH, rpoN	2	5

^aAlmost all the information was obtained from http://mbgd.genome.ad.jp/, a website which facilitates such comparisons. Other footnotes and the references cited in other footnotes either confirm the information or explain how the conclusion was reached.

^bData compiled from analyses of the genomes of N. meningitidis MC58 (70) and Z2491 (55).

^cSee reference 69.

^dSee reference 38.

^eNorspermidine was deduced from finding all enzymes of the pathway and failing to find spermidine synthase.

^fSee reference 60.

^gValues apply to strain Z2491 (55) and strain MC58 (70), respectively.

At the outset, it is readily apparent that these organisms utilize different metabolic strategies for survival in the nasopharynx. All three organisms grow readily under aerobic conditions, while S. pneumoniae is an aerotolerant anaerobe. H. influenzae can grow anaerobically in the apparent absence of alternative respiratory substrates (e.g., nitrate) (20, 21), and both growth experiments (21) and analysis of encoded protein products of the H. influenzae Rd genome (69) indicated that this organism might prefer growth under reducing conditions. N. meningitidis cannot grow under strictly anaerobic conditions but has effective systems for using both nitrite and nitric oxide as respiratory substrates (5, 62). Similar to the meningococcus, M. catarrhalis cannot grow anaerobically (4, 35). However, in addition to containing the genes encoding predicted nitrite and nitric oxide reductases (Table 4), M. catarrhalis also possesses the ability to reduce nitrate to nitrite (16, 35). Interestingly, during growth in a biofilm, these genes encoding the enzymatic machinery necessary to reduce nitrate to the level of nitrous oxide were among those that were most highly expressed (Table 4).

M. catarrhalis differs substantially from the other two pathogens with respect to central metabolic pathways. M. catarrhalis is unable to utilize exogenous carbohydrates (16), apparently lacking both glycolytic pathways and sugar transport systems (Table 5). In contrast, both H. influenzae and N. meningitidis can utilize a limited number of carbohydrates (16, 35), which is consistent with the presence of intact glycolytic pathways and complete phosphotransferase sugar transport systems in these two organisms. In aggregate, these organisms can utilize only a limited variety of carbohydrates, which suggests that the nasopharynx may be restricted in carbohydrate diversity and perhaps availability. Only N. meningitidis appears to have a complete citric acid cycle. M. catarrhalis appears to lack both subunits of succinyl coenzyme A synthetase, whereas H. influenzae is missing several genes encoding citric acid cycle enzymes. Finally, M. catarrhalis has a glyoxylate cycle, whereas the other two organisms do not (Table 5).

Each of these organisms has certain nutritional deficiencies, which is not surprising considering their relatively small genomes. All three bacteria are missing crucial components for high-affinity ammonia assimilation but appear to possess low-affinity ammonia assimilation capability via the activity of glutamate dehydrogenase. However, M. catarrhalis has been reported to be unable to assimilate ammonia (38), which implies that M. catarrhalis is effectively a glutamate auxotroph. The absence of high-affinity ammonia assimilation in N. meningitidis and H. influenzae also implies that there is glutamate auxotrophy in the nasopharynx, unless a high concentration of ammonia is present. Given that M. catarrhalis requires at least arginine for growth and is unable to utilize carbohydrates, it seems likely that amino acids are available in the nasopharynx for the amino acid auxotrophies and energy, although the source of amino acids in the nasopharynx is not readily apparent. This conclusion is reinforced by the similar properties of N. meningitidis and H. influenzae: a limited ability to utilize carbohydrates, at least one requirement for an amino acid, and a potential glutamate auxotrophy in the absence of sufficient ammonia.

When total RNA extracted from M. catarrhalis cells grown under both iron-replete and iron-limiting conditions was subjected to DNA microarray analysis using probes derived from M. catarrhalis ATCC 43617, the genes that were markedly up-regulated included those previously shown by protein expression measurements (3, 12, 14, 43) to be affected by the availability of iron in the growth environment. The DNA microarrays were then used to identify M. catarrhalis genes whose expression was affected by growth in a biofilm. To date, there are only very limited data available about biofilm development by M. catarrhalis (13, 58). Similarly, there is limited information about biofilm formation by N. meningitidis (78), although genes induced or up-regulated by contact of this pathogen with human cells in an in vitro system have been identified by DNA microarray analysis (28). Studies of biofilm formation by H. influenzae are more extensive, and several gene products of this pathogen which are involved in or affected by biofilm growth have been identified (27, 50, 52, 68, 77). One of the H. influenzae gene products that is up-regulated by growth in a biofilm is the thiol-dependent peroxidase peroxiredoxin-glutaredoxin, and isogenic H. influenzae mutants unable to express this protein were shown to be deficient in biofilm formation in vitro (52). A similar peroxiredoxin was also up-regulated during growth of M. catarrhalis in a biofilm (MCORF783) (Table 4), although mutant analysis of this gene product was not performed in the present study.

In a preliminary effort to extend our findings with the DNA microarray-derived data and determine whether genes maximally up-regulated during biofilm growth were essential for this mode of growth, we inactivated two of the genes (narG and narH) which were among those most highly up-regulated by growth in a biofilm (Table 4). However, when tested in a competitive index experiment in the Sorbarod continuous-flow biofilm system, the narGH mutant did not appear to have a deficiency in the ability to form biofilms (data not shown). This result suggests that nitrate reductase activity, while substantially up-regulated during biofilm growth, is not essential for biofilm development in this model system as used in this study. The up-regulation of these particular genes may instead reflect some type of sensing of reduced oxygen tension in the biofilm state.

We also noted that expression of the ORFs predicted to encode nitrite reductase and nitric oxide reductase was highly up-regulated in the biofilm (Table 4). The ability of N. meningitidis to survive in nasopharyngeal tissue has been shown to be enhanced by nitric oxide detoxification systems (67), and in Pseudomonas aeruginosa nitric oxide is involved in signaling biofilm dispersal (8). The predicted ability of M. catarrhalis to reduce nitrate to the level of nitrous oxide may provide an alternative means for energy generation by this organism under oxygen-limited conditions. In addition, the ability to reduce nitric oxide may also provide M. catarrhalis with some level of protection against macrophage-generated nitric oxide. In this context, it is interesting to note that there has been one report describing the selection of M. catarrhalis variants or mutants that were more resistant to nitric oxide than their wild-type parent strain (45), but the identity of the relevant gene product(s) was not determined.

In summary, this study used nucleotide sequence data from the genome of M. catarrhalis ATCC 43617 to provide a preliminary analysis of M. catarrhalis metabolism and to construct DNA microarrays that were used to evaluate global gene expression under defined conditions in vitro. The resultant data indicate that growth of M. catarrhalis ATCC 43617 in a biofilm in vitro differentially affected the expression of genes in only a relatively few categories. The genes whose expression was most highly up-regulated were associated with energy generation involving the reduction of nitrate, nitrite, and nitric oxide. Expression of ribosomal genes was down-regulated, which is consistent with slower growth, while some heat shock genes had increased expression, which would be consistent with stress. More extensive genetic analyses are required to determine which M. catarrhalis genes are specifically required for biofilm formation in the continuous-flow system.