Genome region comparison of actinomycete CM genes. (A) Genomic organization of aroQ genes in actinomycetes: M. tuberculosis (Mtb), M. bovis (Mbo), M. avium (Mav), C. diphtheriae (Cdi), M. leprae (Mle), N. farcinica (Nfa), and M. smegmatis (Msm). Black arrows indicate the conserved aroQ genes. White and gray arrows indicate genes homologous and nonhomologous to the M. tuberculosis pgi-Rv0950c region, respectively. Hatched arrows indicate pseudogenes. Gene names and numbers refer to the original ORF assignments or as defined by multiple sequence alignments. The numbering of genes is abbreviated, so 0947c represents Rv0947c, 0972c represents Mb0972c, 0890 represents MAP0890, 0833 represents DIP0833, 0149 represents ML0149, 49970 represents nfa49970, and 5539 represents MSMEG5539. (B) Genomic organization of *aroQ genes in mycobacteria. Black arrows indicate the conserved *aroQ genes. White and gray arrows indicate, respectively, genes homologous and nonhomologous to the M. tuberculosis cyp140-Rv1888c region. Gene names and numbers and the numbering of genes are as described for panel A.

SDS-PAGE analysis of expression of M. tuberculosis and M. smegmatis CMs in the soluble fraction of crude extracts from E. coli BL21(DE3) cells. All samples were grown in LB medium, and asterisks indicate the addition of 1 mM IPTG to cultures reaching an OD₆₀₀ of 0.4. E. coli transformed with the pET-23a(+) expression vector was used as negative control. (A) Expression of the AroQ_Mt protein (11.8 kDa) in E. coli after 2 h (lanes 2 to 5) and 4 h (lanes 6 to 9) of cell growth. Lanes: 1, molecular mass marker; 2, 4, 6, and 8, host cell extract (negative controls); 3, 5, 7, and 9, E. coli transformed with pET-23a(+)-Mtb aroQ. (B) Expression of the *AroQ_Mt protein (18.6 kDa) in E. coli after 3 h (lanes 1 to 4) and 6 h (lanes 6 to 9) of cell growth. Lanes: 1, 3, 6, and 8, host cell extract (negative controls); 5, molecular mass marker; 2, 4, 7, and 9, E. coli transformed with pET-23a(+)-Mtb *aroQ. (C) Expression of the AroQ_Ms protein (11.7 kDa) in E. coli after 3 h (lanes 2 to 5) and 6 h (lanes 6 to 9) of cell growth. Lanes: 1, molecular mass marker; 2, 4, 6, and 8, host cell extract (negative controls); 3, 5, 7, and 9, E. coli transformed with pET-23a(+)-Msm aroQ. (D) Expression of the *AroQ_Ms protein (17.8 kDa) in E. coli after 3 h (lanes 1 to 4) and 6 h (lanes 6 to 9) of cell growth. Lanes: 1, 3, 6, and 8, host cell extract (negative controls); 5, molecular mass marker; 2, 4, 7, and 9, E. coli transformed with pET-23a(+)-Msm *aroQ.

Multiple sequence alignments of bacterial CMs with the CM domain of the bifunctional protein CM-prephenate dehydratase from E. coli (EcCM). Sequence data were obtained as described in Materials and Methods, and the program CLUSTAL W (49) was used in both alignments. (A) The putative exported CMs for M. tuberculosis (Mtb), Mycobacterium bovis (Mbo), Mycobacterium avium (Mav), and M. smegmatis (Msm) were aligned with EcCM and the monofunctional *AroQ protein from Erwinia herbicola (Ehe), the first characterized periplasmic CM (54). The signal peptides of *AroQ proteins (as predicted by the program SignalP 3.0 [4]) are underlined. Identical residues are shown in white on a black background and are also indicated by asterisks below the alignment. Strongly similar and weakly similar residues are identified by colons and periods, respectively. The first residues of the C-terminal portion of *AroQ proteins (which displays no homology to known CMs [9]) are shown in light gray. Active-site residues of EcCM are marked with arrowheads, and each “h” represents a hydrophobic residue (29). Glu52 in EcCM has conserved residues in the M. smegmatis and E. herbicola sequences (in dark gray), and Ökvist et al. (34) propose that Glu106 in the M. tuberculosis sequence is equivalent (also in dark gray). We extrapolated their analysis to the M. bovis and M. avium sequences. (B) Multiple sequence alignment of actinomycete AroQ proteins with EcCM. Besides the previously mentioned species (except E. herbicola), Mycobacterium leprae (Mle), Nocardia farcinica (Nfa), and Corynebacterium diphtheriae (Cdi) were included in the alignment. Colors and symbols are the same as described for panel A. Black circles indicate substitutions that naturally occur in all analyzed sequences from actinomycetes (also shown in light gray) and that were demonstrated to maintain or increase CM catalytic efficiency when the corresponding E. coli residues (Leu7, Ala32, and Val35) are mutated in EcCM (28).

Determination of translational start site for ORFs MSMEG5513 and MSMEG2114 from M. smegmatis. The DNA sequences of the 5′ end and the region immediately upstream of the aroQ_Ms (MSMEG5513) and *aroQ_Ms (MSMEG2114) genes are shown (panels A and B, respectively). Potential translational start sites are indicated in boldface. Underlined start codons indicate the originally predicted start sites, and methionines in white on a black background correspond to the experimentally identified translational start sites. The introduced frameshift mutations are marked with arrows in each sequence. The PstI restriction site to which a lacZ fusion was made for each construction is underlined. (C) Quantitative analysis of β-Gal activity in M. smegmatis cells transformed with the wild-type (pTASS1) or mutant constructions (mut1, mut2, and mut3) for identifying the MSMEG5513 translational start site. (D) Quantitative analysis of β-Gal activity in M. smegmatis cells transformed with the wild-type (pTASS2) or mutant constructions (mut4 and mut5) for identifying the MSMEG2114 translational start site. The results are the means ± the standard deviations of three individual transformants, each assayed in duplicate. Units are given in nanomoles of o-nitrophenyl galactoside produced per minute per milligram of total protein.

Genetic organization of the aroQ gene and the fbpB operon (which contains the *aroQ gene) on the M. tuberculosis H37Rv chromosome. (A) The genome region containing the M. tuberculosis aroQ gene and its predicted promoter (C1) is shown. The potential −35/−10 consensus sequence and ribosomal binding site (rbs) are underlined, and the aroQ start codon is indicated in boldface. The C1 sequence comprises approximately 250 bp covering the 5′ end and the upstream sequence of aroQ and was cloned in pSM128 to make pCAN1. (B) The fbp operon from M. tuberculosis. The predicted fbpB (P1) and *aroQ (P2) promoters are shown. The potential −35/−10 consensus sequences and ribosomal binding sequences for both genes are underlined (rbs1 and rbs2 refer to an annotated and proposed ribosomal binding site ] for the *aroQ_Mt gene, respectively); start and stop codons are indicated in boldface. The fbpB transcriptional start site (+1) was determined by Harth et al. (25). The P1 (215 bp) and P2 (267 bp) sequences were PCR amplified and cloned into pSM128, yielding pPAN1 and pPAN2, respectively.