Supplemental text

Supplemental Material

Carbohydrate utilization

The Haemophilus somnus 129Pt genome encoded genes involved in the utilization of glucose, ribose, xylose, fucose, galactitol, glucitol/sorbitol, mannose and mannitol (Table S5), while H. ducreyi 35000HP had genes involved in glucose and mannose utilization. The H. influenzae Rd genome contained genes involved in the utilization of glucose, ribose, xylose, fucose, glycerol, fructose and galactose (15, 27).

The most interesting aspect of carbohydrate utilization involved the genes for glucose and fructose transport. H. influenzae Rd has the enzyme I (ptsI, HI1712), Hpr (ptsH, HI1713) and glucose specific IIA (crr, HI1711) genes for the glucose phosphotransferase system (PTS), but not the gene encoding membrane bound glucose-specific enzyme II (ptsG) (Fleischmann et al, 1995). Both H. somnus 129Pt and H. ducreyi 350HP had the ptsI (HS_1097, HD0228), ptsH (HS_1098, HD0229) and crr (IIAGlc, HS_1096, HD0227) genes and, like H. influenzae, lacked ptsG.

We found that H. somnus 129Pt and H. ducreyi 35000HP, like H. influenzae Rd (15), lacked genes encoding the membrane-bound glucose specific enzyme IIBC (ptsG) of the glucose PTS, indicating that they did not have a functional glucose PTS for glucose transport. Therefore, instead of being involved in glucose utilization, the product of the crr gene may be involved in regulation of carbon metabolism and stress response in these Haemophilus strains, as has been proposed for S. coelicolor (22) and P. multocida (4). If this is true, how do these organisms take up glucose? A previous study of H. influenzae suggests that it takes up glucose via a non-PTS permease (26). We did find one gene in H. somnus 129Pt that encoded a sodium/glucose symporter (HS_0338). However, this gene was not present in H. influenzae Rd or H. ducreyi 35000HP. Glucose may enter these organisms via an unidentified permease, an ABC transport system or may be transported by a permease that is specific to another sugar (26). Another option for H. ducreyi 35000HP and H. somnus 129Pt is that they may be able to transport glucose via a mannose PTS, as has been described for P. multocida (4), since they both had genes encoding the mannose PTS. Further evidence in support of this option was our finding that H. ducreyi 35000HP had no genes with sequence similarity to known glucokinase genes. However, it is possible that H. ducreyi 35000HP contains a functional homolog with no sequence similarity to other glucokinase/hexokinase genes. As an alternative to glycolysis, P. multocida may use the pentose phosphate pathway to process glucose (4). This may also be the case for H. somnus 129Pt, H. ducreyi 35000HP and H. influenzae Rd, which all have the complete set of pentose phosphate pathway genes.

H. influenzae had the fruBKA operon (HI0446 – HI0448) encoding components of the fructose PTS, consisting of the fruB gene encoding the protein FPr, fruK (1-phosphofructokinase), and fruA, encoding the fructose-specific IIBC component. Neither H. somnus 129Pt nor H. ducreyi 35000HP had these genes. The closest relative to the fructose PTS is the mannitol PTS (35), which only H somnus 129Pt had. This is interesting in light of evidence that H. somnus can ferment fructose in culture (16). It is possible H. somnus takes up fructose via either a mannose or mannitol PTS, which it encodes, or by an unidentified permease. Both H somnus 129Pt and H. influenzae Rd had genes that were similar to the E. coli mak gene (HS_1253; HI1082), which encodes cryptic manno(fructo)kinase that converts fructose to fructose-6-phosphate, as well as pfkA (6-phosphofructokinase 1; HS_0485; HI0982) and fbaA (fructose bisphosphate aldolase class II; HS_0206; HI0524), all of which are part of the fructose degradation pathway. H. ducreyi 35000HP had homologs of pfkA (HD0465) and fbaA (HD0864) but not mak. H. somnus 129Pt also had a gene that encodes a possible fructose bisphosphate aldolase class I (HS_0055).

In H. influenzae, the galactose utilization genes galMKTR (HI0818 – HI0821) are located together, while galU (HI0812) and galE (HI0351) are not (28). Like H. influenzae Rd, H. somnus 129Pt had the genes galK, galM, galU and galE, but only galK (HS_0235) and galM (HS_0236) were located next to each other (Table S5). It did not have galT, which encodes galactose-1-phosphate uridylyltransferase or galR, the galactose operon repressor. H. ducreyi had only galU (HD1431) and galE (HD0829). GalE is required for the biosynthesis of extracellular polysaccharide materials such as lipopolysaccharide (LPS) and capsule (36). galU is an essential virulence gene that is critical in generating sugar precursors needed for polysaccharide formation and LOS outer core synthesis (49). The unlinked location of galE relative to the other gal operon genes in H. somnus 129Pt and H. influenzae Rd is also seen in Pasteurella (Mannheimia) haemolytica A1; the separation of galE from the rest of the gal operon may have resulted from a transposition event (36). P. haemolytica A1 is normally found inside its bovine host, and therefore may not need a catabolic gal operon, since it probably does not have to grow on galactose as its sole source of carbon (36). Since they are missing some or many of the gal operon genes, and can use other sugars, this may also be the case for H. somnus 129Pt.

Both H. somnus 129Pt and H. influenzae Rd had the ribose operon genes. The H. influenzae rbs genes were organized in an operon, in the same order as in E. coli (3). In contrast, the H. somnus 129Pt rbs operon genes were present in 4 locations around the genome: 1) rbsD, rbsK and rbsR (HS_0223 – HS_0227), HS_0224 and HS_0224a were possible transposases; 2) rbsACB (HS_0763-HS_0765); 3) and 4) H. somnus 129Pt had 1 extra copy of rbsA (HS_0768) and 2 extra copies of rbsC (HS_0769 and HS_1580).