Wang et al. 10.1073/pnas.0611606104. |
Fig. 6. Mass spectrometry of purified Francisella novicida lipid A in the negative ion mode. (A) High-resolution electrospray ionization mass spectrum of the predominant lipid A component (A1) present in wild-type F. novicida. The major single-charged peak at m/z 1,665.153 (monoisotopic) is interpreted as [M-H]- of a lipid A molecule acylated with three R-3-hydroxy-C18 chains and one secondary C16 group (Inset); the predicted m/z of the monoisotopic peak of the [M-H]- ion is 1,665.173 (1). Minor isobaric molecular species are present but less abundant. (B) High-resolution electrospray ionization mass spectrum of the major lipid A component (A4) found in the lpxF mutant of F. novicida. The predominant double-charged peak at m/z 999.156 (monoisotopic) is interpreted as [M-2H]2- of a lipid A molecule that contains two phosphate residues, one hexosamine moiety, three primary R-3-hydroxy-C18 chains, one primary R-3-hydroxy-C16 chain, and one secondary C16 group (Inset). The predicted m/z for the monoisoptoic peak of this [M-2H]2- ion is 999.178. Other isobaric combinations arising from fatty acid chain length heterogeneity are present. The single-charged peak at 2,000.343 is interpreted as [M-H]- of the same compound as is seen at m/z 999.156, but containing one 13C atom. The double-charged peak at m/z 985.112 arises from lipid A molecules with two fewer methylene units than the species at m/z 999.156, again the result of fatty acyl chain length microheterogeneity in F. novicida lipid A. A complete tandem MS analysis of all of the molecular ions (data not shown) confirmed the acyl chain assignments and the presence of the hexosamine group, previously shown to be galactosamine (1).
Fig. 7. Cumulative growth yield of F. novicida U112 versus lpxF mutant XWK4. Cells were grown as described in Fig. 1, except that the cultures were diluted 10-fold whenever the A600 reached 0.4. The y axis represents the cumulative growth yield, which takes the dilution factor into account.
Fig. 8. Lipid A of wild-type and lpxF mutant F. novicida is inactive against human TLR2 and TLR4. HEK293 cell lines stably expressing fluorescently tagged TLR2 or TLR4/MD2 were described previously (2). Activation of HEK293 cells expressing TLR2YFP, TLR4YFP/MD2, or empty vector was evaluated with the NF-kB luciferase reporter assay (2). Lipid A of wild-type and lpxF mutant F. novicida was also inactive in stimulating RAW 264.1 macrophage tumor cells to produce TNFa (not shown).
Table 1. 1H and 13C NMR assignments of A1 and A4 from F. novicida
Compound | A1* | A4* | A1* | A4* |
Residue | dH, [mult, J(Hz)] | dH, [mult, J(Hz)] | dC, [mult] | dC, [mult] |
10 | 5.71 [dd,3.5,5.2] | 5.73 [dd,3.6,5.8] | 93.6 [dd] | 93.0 [dd] |
20 | 3.48 [dt,3.5,10.7] † | 3.50 [dt,3.3,10.7] † | 52.1 [dd] | 51.9 [dd] |
30 | 4.05 [m] | 4.07 [m] | 67.2 [d] | 68.0 [d] |
40 | 3.93 [m] | 3.99 [m] | 69.1 [d] | 69.1 [d] |
50 | 4.06 [m] | 4.08 [m] | 73.2 [d] | 73.0 [d] |
6a0 | 3.77 [m] | 3.89 [m] | 61.9 [t] | 61.7 [t] |
6b0 | 3.69 [m] | 3.85 [m] | ||
1 | 5.45 [dd,3.5,7.9] | 5.48 [dd,3.1,7.9] | 94.5 [dd] | 94.0 [dd] |
2 | 4.14 [dt,3.5,10.7] † | 4.13 [dt,3.5,10.7] | 52.5 [dd] | 52.5 [dd] |
3 | 5.17 [m] | 5.16 [m] | 72.9 [d] | 73.4 [d] |
4 | 3.42 [dd] | 3.44 [dd] | 69.2 [d] | 69.1 [d] |
5 | 4.21 [m] | 4.15 [m] | 73.2 [d] | 72.5 [d] |
6a | 4.02 [m] | 4.06 [m] | 71.3 [t] | 71.5 [t] |
6b | 3.80 [dd,7.3,12.0] | 3.84 [dd,7.3,12.0] | ||
1¢ | 4.45 [d,8.6] | 4.58 [d,8.6] | 103.9 [d] | 102.5 [d] |
2¢ | 3.70 [m] | 4.02 [m] | 56.4 [d] | 54.0 [d] |
3¢ | 3.45 [m] | 5.15 [m] | 74.9 [d] | 74.0 [d] |
4¢ | 3.34 [m] | 4.18 [m] | 71.2 [d] | 71.8 [d] |
5¢ | 3.33 [m] | 3.51 [m] | 77.1 [d] | 75.6 [d] |
6¢a | 3.87 [m] | 3.79 [m] | 61.6 [t] | 62.5 [t] |
6¢b | 3.70 [m] | 3.71 [m] | ||
a2a | 2.32 [m] | 2.29 [m]] | 44.2 [t] | 44.1 [t] |
a2b | 2.27 [m] | 2.25 [m] | ||
b2 | 3.94 [m] | 3.93 [m] | 68.9 [d] | 69.0 [d] |
g2 | 1.43 [m] | 1.45 [m] | 37.4 | |
a3a | 2.48 [m] | 2.46 [m] | 42.6 [t] | 44.2 [t] |
a3b | 2.38 [dd,9.9,15.2] | 2.37 [m] | ||
b3 | 3.97 [m] | 3.96 [m] | 68.7 [d] | 69.3 [d] |
g3 | 1.46 [m] | 1.50 [m] | 37.4 | |
a2¢a | 2.54 [dd,3.1,14.2] | 2.50 [m] | 41.9 [t] | 44.2 [t] |
a2¢b | 2.46 [m] | 2.34 [m] | ||
b2¢ | 5.19 [m] | 5.09 [m] | 72.0 [d] | 69.3 [d] |
g2¢ | 1.59 [m] | 1.61 [m] | 37.4 | |
a3¢a | 2.46 [m] | 41.6 [t] | ||
a3¢b | 2.46 [m] | |||
b3¢ | 4.00 [m] | 71.8 [d] | ||
g3¢ | 1.45 [m] | |||
a CH2 | 2.30 [m] | 2.28 [m] | 35.2 [t] | 35.1 [t] |
bCH2 | 1.60 [m] | 1.61 [m] | 25.9 | |
g CH2 | 1.3 | 1.3 | ||
(CH2)n | 1.24 [m] | 1.24 [m] | 30.4 [m] | 30.4 [m] |
w-2 CH2 | 1.26 [m] | 1.26 [m] | 32.7 [t] | 32.7 [t] |
w-1 CH2 | 1.26 [m] | 1.26 [m] | 23.40 [t] | 23.40 [t] |
w CH3 | 0.86 [t, ~7] | 0.88 [t, ~7] | 14.4 [q] | 14.3 [q] |
*1H and 13C chemical shifts at 25°C in CDCl3/CD3OD/D2O [4:4:1 (v/v) for A1; 6:5:1 (v/v) for A4] are relative to an internal tetramethylsilane standard and are assigned based on two-dimensional NMR experiments (1). The numbering scheme is the same as described by Wang et al. (1). Published chemical shifts for compound A1 (1) are provided for comparative purposes. The significant downfield shifts of H-3¢ and H-4¢ in A4 versus A1 support the proposed acylation of the 3¢-OH group and the phosphorylation of the 4¢ -OH group in A4, as shown in Fig. 1.
†
Measured from the resolved H-2 double-doublet in the 31P-decoupled 1H NMR difference spectrum (Fig. 3).