Complete Characterization of Polyacyltrehaloses from Mycobacterium tuberculosis H37Rv Biofilm Cultures by Multiple-Stage Linear Ion-Trap Mass Spectrometry Reveals a New Tetraacyltrehalose Family

Abstract

Polyacylated trehaloses in Mycobacterium tuberculosis play important roles in pathogenesis and structural roles in the cell envelope, promoting the intracellular survival of the bacterium, and are potential targets for drug development. Herein, we describe a linear ion-trap multiple-stage mass spectrometric approach (LIT MSⁿ) with high-resolution mass spectrometry to the structural characterization of a glycolipid family that includes a 2,3-diacyltrehalose, 2,3,6-triacyltrehalose, 2,3,6,2′,4′-petaacyltrehalose, and a novel 2,3,6,2′-tetraacyltrehalose (TetraAT) subfamily isolated from biofilm cultures of M. tuberculosis H37Rv. The LIT MSⁿ spectra (n = 2, 3, or 4) provide structural information to unveil the location of the palmitoyl/stearoyl and one to four multiple methyl-branched fatty acyl substituents attached to the trehalose backbone, leading to the identification of hundreds of glycolipid species with many isomeric structures. We identified a new TetraAT subfamily whose structure has not been previously defined. We also developed a strategy for defining the structures of the multiple methyl-branched fatty acid substituents, leading to the identification of mycosanoic acid, mycolipenic acid, mycolipodienoic acid, mycolipanolic acid, and a new cyclopropyl-containing acid. The observation of the new TetraAT family, and the realization of the structural similarity between the various subfamilies, may have significant implications in the biosynthetic pathways of this glycolipid family.

Graphical Abstract

The mycobacterial cell envelope contains a high level of lipid, including a structurally related glycolipid family that consists of a trehalose backbone to which two to five fatty acyl chains, including multiple methyl-branched fatty acids, are attached via an ester linkage.^1-4 Among them, the acylated trehaloses found in Mycobacterium tuberculosis^5-8 and Mycobacterium fortuitum^9,10 are located at the outermost compartment of the cell envelope¹¹ and consisted of several subfamilies.^8-10,12,13 The subfamily previously named glycolipid B is 2,3-di-O-acyltrehalose (DAT), which together with the antigenic 2,3,6-triacyl trehalose (TAT) subfamily was also identified in M. fortuitum,¹⁰ and in the clinical isolates and reference strains of M. tuberculosis.⁷ Another glycolipid subfamily previously characterized as 2,3,6,2′,4′-pentaacyl trehalose (PAT) was found in the outer mycomembrane of M. tuberculosis.^5,14

DAT, PAT, sulfolipid, and phthiocerol dimycocerosate share the same biosynthetic pathway. Biosynthesis of PAT requires MmpL10 to transport DAT to the M. tuberculosis cell surface. On the periplasmic face of the plasma membrane, the bacterium further elaborates the glycoconjugate to form PAT via transesterification reactions catalyzed by the acyltransferase Chp2 using polymethyl-branched fatty acyl substituents, including mycosanoyl, mycolipenoyl, and mycolipanolyl species released from DAT.^15,16

DAT is capable of modulating host immune responses¹⁷ and can inhibit the proliferation of murine T cells.¹⁸ DAT along with PAT also plays an important role in pathogenesis and a structural role in the cell envelope, promoting the intracellular survival of the bacterium.¹⁹ The DATs from M. tuberculosis and M. fortuitum are antigenic,^18,20,21 and their potential use in serodiagnosis has been postulated.^22,23

The structures of the glycolipids in the bacterial cell are complex, and mass spectrometry has played key roles in the determination of their structures. For example, fast atom bombardment mass spectrometry was previously used to determine the structures of DATs isolated from the cell envelope of M. fortuitum.¹² Besra and co-workers defined the structures of the acylated trehalose lipid family using gas chromatography—mass spectrometry, in conjunction with normal/reversed phase TLC, and one- and two-dimensional ¹H and ¹³C nuclear magnetic resonance spectroscopy analyses.¹³ We also applied linear ion-trap multiple-stage mass spectrometry (LIT MSⁿ) to characterize complex glycolipids, including 2,3-DAT,²⁴ sulfolipids,²⁵ glycopeptidolipids,²⁶ and phosphatidylinositol mannosides,^27,28 from mycobacteria. However, a simple mass spectrometric approach to a complete determination of the acylated trehalose lipid family has not been described. Here, we describe the development of a LIT MSⁿ approach combined with simple chemical reactions toward complete structural characterization of DAT, TAT, PAT, and a new tetraacyltrehlose (TetraAT), including the structures of the polymethylated fatty acyl chains of the molecules.

MATERIALS AND METHODS

Materials.

An AMP+ Mass Spectrometry Kit (50 test) containing AMPP derivatizing reagent, n-butanol (HOBt), 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide (EDC), and an acetonitrile/DMF [1:1 (v/v)] solution was purchased from Cayman Chemical Co. (Ann Arbor, MI). All other solvents (spectroscopic grade) and chemicals (ACS grade) were obtained from Sigma Chemical Co. (St. Louis, MO).

Sample Preparation.

M. tuberculosis strain H37Rv was grown planktonically in biofilm cultures, and total lipids were extracted and isolated as previously described.²⁹ Briefly, bacteria were cultured in polystyrene bottles (Corning) in Sauton’s medium containing 0.5 g L⁻¹ K₂HPO₄, 0.5 g L⁻¹ MgSO₄, 4.0 g L⁻¹ l-asparagine, 0.05 g L⁻¹ ferric ammonium citrate, 4.76% glycerol, and 1.0 mg L⁻¹ ZnSO₄, at a final pH of 7.0. Flasks were incubated at 37 °C with 5.5% CO₂ with tightened caps for 2 weeks. After 2 weeks, caps were loosened to promote biofilm formation.³⁰ To harvest total lipids, cultures were harvested by centrifugation, autoclaved for removal from biosafety level 3 (BSL3), and resuspended in a chloroform/methanol mixture [2:1 (v/v)]. Following extraction, lipids were dried under N₂ gas, resuspended in a chloroform/methanol mixture [2:1 (v/v)], and separated by a Supelco Ascentis C-8 column (100 mm × 2.1 mm, 2.7 μm particle size) at a flow rate of 250 μL/min with a previously described gradient system with slight modification.³¹ DAT (eluted at 20.9–22.9 min), TAT (eluted at 30.9–32.2 min), tetraAT (eluted at 34.3–36.5 min), and PAT (eluted at 37.5–39.6 min) (see Figure S1) fractions from three injections (~200 μg of total lipid/injection) were collected and pooled, dried under a stream of dry nitrogen, and stored at −20 °C until use.

Alkaline Hydrolysis and Preparation of a Fatty Acid-N-(4-aminomethylphenyl)pyridinium (FA-AMPP) Derivative.

To the dry lipid extract (~200 μg) were added 500 μL of methanol and 500 μL of tetrabutylammonium hydroxide (40 wt % solution in water). The solution was heated at 75 °C for 2 h and cooled to room temperature, and 2 mL of water and 2 mL of hexane were added, vortexed for 1 min, and centrifuged at 1200g for 2 min. The top layer containing fatty acids was transferred to a centrifuge tube and dried under a stream of nitrogen, and the AMPP derivative was made with the AMP+ Mass Spectrometry Kit, according to the manufacturer’s instructions. Briefly, the dried sample was resuspended in 20 μL of an ice-cold acetonitrile/DMF mixture [4:1 (v/v)], and 20 μL of ice-cold 1 M 3-[(dimethylamino)-propyl]ethylcarbodiimide hydrochloride (EDCI) in water was added. The vial was briefly mixed on a vortex mixer and placed on ice. To the vial were added 10 μL of a 5 mM N-hydroxybenzotriazole (HOAt) solution and 30 μL of a 15 mM AMPP solution (in distilled acetonitrile), vortexed for 30 seconds and heated at 65 °C for 30 min. After cooling to room temperature, 1 mL of water and 1 mL of n-butanol were added to the mixture. The final solution was vortexed for 1 min and centrifuged at 1200g for 3 min, and the organic layer was transferred to another vial.

Mass Spectrometry.

Both high-resolution (R = 100000 at m/z 400) higher-energy collision dissociation (HCD) and low-energy collision-induced dissociation (CID) tandem mass spectrometric experiments were conducted on a Thermo Scientific (San Jose, CA) LTQ Orbitrap Velos mass spectrometer with an Xcalibur operating system. Samples in methanol were loop injected into the ESI source, where the skimmer was set at ground potential, the electrospray needle was set at 4.0 kV, and the temperature of the heated capillary was 300 °C. The automatic gain control of the ion trap was set to 5 × 10⁴, with a maximum injection time of 100 ms. Helium was used as the buffer and collision gas at a pressure of 1 × 10⁻³ mbar (0.75 mTorr). The MSⁿ experiments were carried out with an optimized relative collision energy ranging from 25% to 40% with an activation q value at 0.25. The activation time was set for 10 ms to leave a minimal residual abundance of precursor ion (~20%). For the HCD experiments, the collision energy was set to 40–50% and mass scanned from m/z 100 to the upper m/z value that covers the precursor ions. The mass selection window for the precursor ions was set at 1 Da wide to admit the monoisotopic peak to the ion trap for CID for unit-resolution detection in the ion trap or high-resolution accurate mass detection in the Orbitrap mass analyzer. Mass spectra were recorded in the profile mode, typically for 3–10 min for MSⁿ spectra (n = 2, 3, or 4).

Nomenclature.

To facilitate the interpretation of data, the following abbreviations were adopted. The abbreviation of a long chain fatty acid such as palmitic acid attached to position 2 of trehalose is designated as 16:0. The multiple methyl-branched mycolipenic and mycolipanolic acids, for example, the 2,4,6-trimethyl-2-tetracosenoyl group attached to position 3 of the trehalose core, are designated as 27:1-acid to reflect the fact that the substituent contains a C₂₇ acyl chain with one double bond. Therefore, the DAT species consisting of 16:0- and 27:1-FA substituents at C2 and C3 of the trehalose backbone, respectively, is designated as 16:0/27:1-DAT. Likewise, the TAT molecule consisting of 18:0-, 27:1-, and 24:0-fatty acyl chains attached to C2, C3, and C6 of the trehalose backbone, respectively, is designated as 18:0/27:1/24:0-TAT, and a PAT molecule possessing 16:0-, 27:1-, 24:0-. 27:1-, and 27:1-fatty acyl chains at C2, C3, C6, C2′, and C4′, respectively, is designated as (16:0/27:1/24:0)(27:1/27:1)-PAT. To distinguish it from the TAT family, the novel tetraacyltrehalose family is abbreviated as TetraAT. Thus, for example, a TetraAT molecule with 16:0-, 27:1-, 24:0-, and 27:1-FA at C2, C3, C6, and C2′, respectively, is named (16:0/27:1/24:0)(27:1)-TetraAT.

RESULTS AND DISCUSSION

Glycolipids, including DAT, TAT, TetraAT, and PAT (structures shown in schemes), formed [M + Alk]⁺ ions (Alk = NH₄, Li, Na, etc.) in the positive ion mode. In the negative ion mode in the presence of X⁻, ions of the [M + X]⁻ (X = Cl, HCO₂, CH₃CO₂, or long chain fatty acid carboxylate anions such as palmitic and oleic acids) type were also formed.^24,32,33 The HPLC ESI-MS spectra of the various glycolipid families seen as the [M + NH₄]⁺ ions in the positive ion mode (Figure S1a-d, top panel) and the corresponding [M + HCO₂]⁻ ions in the negative ion mode (Figure S1a-d, bottom panel) were formed due to the presence of ammonium formate in the HPLC mobile phase. These mass spectra provide the information of the molecular species of these lipids.

Both [M + Na]⁺ and [M + HCO₂]⁻ adduct ions are useful for characterization of 2,3-diacyl DAT by LIT MSⁿ.²⁴ However, structural information from LIT MSⁿ on the [M + NH₄]⁺ ions is insufficient. Thus, we conducted studies focused on the [M + Na]⁺ ions, in combination with [M + HCO₂]⁻ ions. The LIT MSⁿ (n = 2, 3, or 4) approaches with high-resolution mass spectrometry for the characterization of these species are described below.

Characterization of DAT.

We previously defined the structures of 2,3-diacyl DAT lipids in M. tuberculosis H37Rv.²⁴ Interestingly, we found that the profile and structures of DAT in M. tuberculosis grown as biofilms (Figure 1a and Figure S1a) are drastically different in M. tuberculosis cells grown in 7H9 Middlebrook broth.²⁴ For example, the ion at m/z 981.7, representing an [M + Na]⁺ ion of 18:0/24:0-DAT, is the most abundant ion, and members of the 2-palmitoryl/stearoyl 3-mycolipanoloyl (β-hydroxy fatty acid) DAT subfamily such as 2,3-16:0/βh27:0-DAT and 2,3-18:0/βh27:0-DAT are also present (Table 1); in the previous study, these species are absent, and the ion at m/z 991.7, representing a [M + Na]⁺ ion of 16:1/27:1-DAT, is the base peak.²⁴ However, it is unclear if the differences are related to the biofilm cultures or growth medium.

Figure 1.

High-resolution ESI-MS spectra of the [M + Na]⁺ ions of (a) DAT, (b) TAT, (c) TetraAT, and (d) PAT.

Table 1.

DAT Species Found in a Cultured M. tuberculosis Biofilm by High-Resoltion LIT MSⁿ

m/z [M + Na]+	theoretical mass (Da)	relative intensity (%)	deviation (mDa)	composition	structuresa
939.6750	939.6743	4.49	0.71	C51H96O13Na	18:0/21:0, 16:0/23:0
951.5749	951.6743	3.7	0.6	C52H96O13Na	16:1/24:0, 16:0/24:1
953.6905	953.6900	17.89	0.52	C52H98O13Na	16:0/24:0
965.6906	965.6900	3.87	0.64	C53H98O13Na	16:0/25:1, 18:1/23:0, 17:l/24:0
967.7063	967.7056	8.62	0.66	C53H100O13Na	18:0/23:0, 17:0/24:0, 16:0/25:0
977.6906	977.6900	1.27	0.64	C54H98O13Na	18:1/24:1, 18:0/24:2, 16:l/26:1, 16:0/26:2, 15:1/27:1
979.7062	979.7056	15.69	0.54	C54H100O13Na	18:0/24:1, 18:1/24:0, 16:0/26:1, 16:1/26:0, 17:1/25:0
981.7218	931.7213	100	0.55	C54H102O13Na	18:0/24:0
991.7061	991.7056	3.4	0.46	C55H100O13Na	16:1/27:1, 18:1/25:1
993.7218	993.7213	10.83	0.5	C55H102O13Na	16:0/27:1, 18:0/25:1, 17:0/26:1
995.7375	995.7369	6.3	0.63	C55H104O13Na	18:0/25:0, 19:0/24:0, 16:0/27:0, 17:0/26:0
1005.7218	1005.7213	1.65	0.55	C56H102O13Na	b
1007.7375	1007.7369	6.06	0.58	C56H104O13Na	18:0/26:1, 17:0/27:1, 16:0/28:1
1009.7532	1009.7526	10.66	0.63	C56H106O13Na	18:0/26:0, 20:0/24:0
1019.7373	1019.7369	3.67	0.4	C57H104O13Na	18:1/27:1
1021.7532	1021.7526	12.07	0.65	C57H106O13Na	18:0/27:1
1035.7686	1035.7682	2.23	0.43	C58H108O13Na	19:0/27:1, 18:0/28:1
983.7011	983.7005	3.23	0.6	C53H100O14Na	18:0/βh23:0
997.7169	997.7162	14.07	0.74	C54H102O14Na	18:0/βh24:0
1009.7168	1009.7162	9.21	0.6	C55H102O14Na	b
1011.7324	1011.7318	26.77	0.6	C55H104O14Na	16:0/βh27:0, 18:0/βh25:0
1023.7324	1023.7318	2.58	0.59	C56H104O14Na	18:0/βh26:1, 17:0/βh27:1
1025.7480	1025.7475	8.18	0.55	C56H106O14Na	18:0/βh26:0, 17:0/βh27:0
1035.7322	1035.7318	1.3	0.39	C57H104O14Na	b
1037.7480	1037.7475	13.55	0.55	C57H106O14Na	18:1/βh27:0, 18:0/βh27:1
1039.7637	1039.7631	48.23	0.57	C57H108O14Na	18:0/βh27:0
1051.7639	1051.7631	1.31	0.77	C58H108O14Na	19:1/βh27:0, 18:0/βh28:1
1053.7794	1053.7788	3.43	0.61	C58H110O14Na	19:0/βh27:0, 18:0/βh28:0
1067.7951	1067.7944	1.29	0.65	C59H112O14Na	20:0/βh27:0

^aSpecies abundance in descending order.

^bStructure not defined.

The characterization of DATs in biofilm is exemplified by MS² on the [M + Na]⁺ ion at m/z 981 (Figure 2a), which yielded the predominated ion at m/z 819 (981 – 162), arising from the loss of a hexose residue, along with the ions at m/z 697 and 613, arising from losses of 18:0-FA (at C2) and 24:0-FA (at C3) substituents, respectively. The ion at m/z 697 is more abundant than the ion at m/z 613, indicating that loss of the FA substituent at C2 is a more facile process than the loss of the FA substituent at C3. Further dissociation of the ions at m/z 819 [981 → 819 (Figure 2b)] yielded ions at m/z 535 (819 – 284) and m/z 451 (831 – 368) arising from further losses of 18:0- and 24:0-fatty acid substituents, respectively. Again, the ion at m/z 535 is more abundant than the ion at m/z 451, consistent with the notion that loss of the FA substituent at C2 is a more facile than similar loss of the FA substituent at C3. These results also indicated that the 18:0- and 24:0-fatty acyl groups are attached to the same glucose (Glc 1). The assignment of the 18:0/24:0-DAT structure is further supported by the MS³ spectra of the ions at m/z 535 [981 → 535 (Figure 2c)] and m/z 451 [981 → 451 (Figure 2d)]. The former spectrum contained the prominent ion at m/z 517 (loss of H₂O) and ions at m/z 185 and 167 arising from further loss of the 24:0-fatty acid substituent as ketene and acid, respectively, along with an ion at m/z 391 representing a sodiated 24:0-FA. The latter spectrum contained ion at m/z 433 (loss of H₂O), and the prominent ions at m/z 185 and 167 arising from further loss of the 18:0-fatty acid substituent as ketene and acid, respectively, along with the ion at m/z 307 representing a sodiated 18:0-FA (Scheme 1).

Figure 2.

(a) MS² spectrum of the [M + Na]⁺ ion at m/z 981 and (b) its MS³ spectra of the ions at m/z 819 (981 → 819), (c) m/z 535 (981 → 535), and (d) m/z 451 (981 → 451) that define the 18:0/24:0-DAT structure.

Scheme 1. Major Fragmentation Pathways for the [M + Na]⁺ Ion of 18:0/24:0-DAT at m/z 981^a.

^aAll the ions shown in the scheme are sodiated ions, and “Na⁺” is omitted for simplicity.

In Figure 2c, ions at m/z 505, 475, and 445 arising from cleavages of the sugar ring are present. Ions from cleavage of the sugar ring are also seen at m/z 421, 391, and 361 in Figure 2d. The drastic differences between the two spectra (Figure 2c,d) may reflect the location of the fatty acid substituent on the sugar ring, leading to assignment of their position attached to the hexose (Scheme 1).

Similarly, the MS² spectrum of the [M + Na]⁺ ion at m/z 993 (Figure 3a) is dominated by the ion at m/z 831 (993 – 162) arising from loss of hexose, together with the ions at m/z 737 and 585, arising from losses of 16:0- and 27:1-fatty acid substituents, respectively. Further dissociation of the ion at m/z 831 [993 → 831 (Figure 3b)] yielded ions at m/z 575 (831 – 256) and m/z 423 (831 – 408), arising from further losses of 16:0- and 27:1-fatty acid substituents, respectively, indicating the presence of 16:0- and 27:1-fatty acyl groups on the same hexose (Glc 1). The assignment of the 16:0/27:1-DAT structure is further supported by the MS³ spectrum of the ion at m/z 737 [993 → 737 (Figure 3c)], which contained the prominent ions at m/z 329, arising from further loss of the 27:1-fatty acid substituent, and m/z 575, arising from loss of the glucose residue (Glc 2), together with the ion at m/z 431 representing a sodiated 27:1-FA (Scheme 1).

Figure 3.

(a) MS² spectrum of the [M + Na]⁺ ion at m/z 993 and its MS³ spectra of the ions at (b) m/z 831 (993 → 831), (c) m/z 737 (993 → 737), and (d) m/z 709 (993 → 709) (see the text for the assignment of the DAT structures).

In Figure 3a, ion pairs at m/z 709/613 and 723/599, arising from losses of 18:0/25:1 and 17:0/26:1 fatty acids substituents, respectively, are also present, consistent with the presence of the ion pairs at m/z 547/451 and 561/437 in Figure 3b. The results led to the assignment of the 18:0/25:1-DAT and 17:0/26:1-DAT isomers. The presence of these two minor isomers was further confirmed by the MS³ spectra of the ions at m/z 709 [993 → 709 (Figure 3d)] and m/z 723 [993 → 723 (not shown)]. The profile of the spectrum (Figure 3d) is similar to that of Figure 3c and contained the major ions at m/z 547 (709 – 162) and m/z 329 (709 – 380) arising from losses of glucose and 25:1-FA residues, respectively, together with the ion at m/z 403, corresponding to a sodiated 25:1-FA. The results support the presence of a 18:0/25:1-DAT isomer. Similarly, the MS³ spectrum of the ions at m/z 723 (not shown) contains the ions at m/z 561 (723 – 162) and m/z 343 (723 – 380), pointing to the presence of a minor 17:0/26:1-DAT isomer.

In addition to the DATs consisting of mycocerosic acid (e.g., 24:0) and mycolipenic acid (e.g., 27:1) substituents at C3 described above, members of the lipid subfamily consisting of mycolipanolic acid and mycolipodienoic acid substituents at C3 are also present (Table 1). For example, high-resolution mass measurements of the ion at m/z 1039.7637 gave an elemental composition of C₅₇H₁₀₈O₁₄Na (calculated m/z 1039.7631) (Table 1), indicating that the molecule contains an additional oxygen, likely present as a hydroxyl group on the fatty acid chain. This speculation is confirmed by the MS² spectrum of the ions at m/z 1039.9 ([M + Na]⁺) (Figure 4a), which contained a major ion at m/z 877 (1039 – 162), together with ions at m/z 755 and 613, arising from losses of 18:0- and βh27:0-fatty acid substituents, respectively. The location of the hydroxyl group on the fatty acid chain is further recognized by the presence of the ion at m/z 687, arising from cleavage of the C2─C3 bond of the βh27:0-fatty acid (3-hydroxy 2,4,6-trimethyl-tetraeicosanoic acid) as an aldehyde [CH₃(CH₂)₂₂CHO, 352 Da] to form a sodiated 2-stearyl, 3-propanoyl-DAT (18:0/3:0-DAT) (Scheme 2). Further dissociation of the ions at m/z 877 [1039 → 877 (Figure 4b)] gave rise to the prominent ions at m/z 593 (877 – 284) and m/z 451 (877 – 426), arising from similar losses of 18:0- and βh27:0-FA substituents, respectively, along with ions at m/z 525, representing a sodiated 18:0/3:0-Glc arising from similar loss of the aldehyde [CH₃(CH₂)₂₂CHO, 352 Da] residue. The MS⁴ spectrum of the ion at m/z 525 (1039 → 877 ↑ 525) (Figure 4c) contained the ion at m/z 451 arising from loss of the 3:0-FA (CH₃CH₂COOH, 74 Da) residue, further supporting the fragmentation process for formation of 18:0/3:0-Glc from the cleavage.

Figure 4.

(a) MS² spectrum of the [M + Na]⁺ ion of 18:0/βh27:0-DAT at m/z 1039.9, (b) its MS³ spectrum of the ion at m/z 877 (1039 → 877), and (c) MS⁴ spectrum of the ion at m/z 525 (1039 → 877 → 525) that define the 18:0/βh27:0-DAT structure. The presence of βh27:0 substituent is further supported by (d) the MS² spectrum of the corresponding [M + HCO₂]⁻ ion of 18:0/βh27:0-DAT at m/z 1061 and (e) its MS³ spectrum of the ion at m/z 663 (1061 → 663) and (f) its MS⁴ spectrum of the ion at m/z 589 (1061 → 663 → 589).

Scheme 2. Major Fragmentation Pathways for the [M + Na]⁺ Ion of 18:0/h27:0-DAT at m/z 1039^a.

^aAll the ions shown in the scheme are sodiated ions and “Na⁺” is omitted for simplicity.

To further confirm the presence of a β-OH group on the h27:0-FA chain, we also obtained the MS² spectrum of the corresponding [M + HCO₂]⁻ ion at m/z 1061 (Figure 4d), which gave rise to a prominent ion at m/z 663 arising from elimination of HCO₂H and CH₃(CH₂)₂₂CHO simultaneously, by a fragmentation process likely initiated by the HCO₂⁻ ion that renders nucleophilic attraction of the H⁺ on the OH, and released the CH₃(CH₂)₂₂CHO residue (Scheme 3). This fragmentation is further supported by the MS³ spectrum of the ion at m/z 663 (1061 → 663) (e) and the MS⁴ spectrum at m/z 589 (1061 → 663 → 589) (f) (Figure 4). The former spectrum (Figure 4e) contained ions at m/z 589 and 379 arising from loss of 3:0- and 18:0-FA substituents, respectively, consistent with the formation of 18:0/3:0-DAT by loss of CH₃(CH₂)₂₂CHO from ion at m/z 1061. The latter spectrum contained the ions at m/z 323 and 305 from loss of the 18:0-FA residue as ketene and acid, respectively, along with the ion at m/z 283 representing a 18:0-carboxylate anion, consistent with the fragmentation as shown in Scheme 3. By contrast, a similar fragmentation process initiated by HCO₂⁻ to eliminate the FA substituent as a ketene for DAT consisting of a non-hydroxy FA substituent is of low abundance.²⁴

Scheme 3.

Proposed Fragmentation Processes of the [M + HCO₂]⁻ Ions of 18:0/βh27:0-DAT at m/z 1061 under LIT MSⁿ in the Negative Ion Mode

The revelation of the βh27:0-fatty acid as a 3-hydroxy 2,4,6-trimethyl-tetracosanoic acid (mycolipanolic acids), a 24:0-FA substituent as 2,4-dimethyl docosanoic acid (mycocerosic), and a 27:1-FA substituent as 2,4,6-trimethyl-tetracosenoic acid (mycolipodienoic acid) was performed by the HCD product ion spectra of their AMPP derivative (described below).

Characterization of TAT.

The high-resolution ESI-MS spectrum of TAT (eluted at 28.96–29.91 min) desorbed as the [M + Na]⁺ ions (Figure 1b) has a profile similar to that seen as the [M + NH₄]⁺ (Figure S1b, top panel) and [M + HCO₂]⁻ (Figure S1b, bottom panel) ions. The recognition of the TAT structures (Table 2) is realized by the presence of an extra fatty acyl chain in the molecules as compared to their DAT analogues. For example, the ion at m/z 1372.1062 obtained by high-resolution mass measurements corresponds to an elemental composition of C₈₁H₁₅₂O₁₄Na (calculated m/z 1372.1074) (Table 2), indicating that the molecule possesses an additional C₂₅H₄₉CH═CO residue reflecting the presence of a 27:1-fatty acyl substituent, as compared to the 18:0/24:0-DAT ion at m/z 981.7212 (C₅₄H₁₀₂O₁₃Na).

Table 2.

TAT Species Found in a Cultured M. tuberculosis Biofilm by High-Resolution LIT MSⁿ

m/z [M + Na]+	theoretical mass (Da)	deviation (mDa)	relative intensity (%)	composition	major structuresa	minor structures
1316.0434	1316.0448	−1.39	9.28	C77H144O14Na	16:0/24:0/25:1	15:0/22:0/27:1, 18:0/22:0/25:1
1330.0590	1330.0605	−1.44	11.26	C78H146O14Na	18:0/24:0/24:1	16:0/23:0/27:1, 17:0/24:0/25:1
1342.0589	1342.0605	−1.59	10.5	C79H146O14Na	16:1/24:0/27:1	18:1/24:0/25:1, 18:0/24:0/25:2, 18:0/22:0/27:2
1344.0747	1344.0761	−1.39	40.95	C79H148O14Na	18:0/24:0/25:1	16:0/24:0/27:1
1356.0744	1356.0761	−1.74	6.88	C80H148O14Na	b
1358.0916	1358.0918	−0.2	15.31	C80H150O14Na	18:0/24:0/26:1	18:0/23:0/27:1, 18:0/25:0/25:1, 17:0/24:0/27:1, 16:0/24:0/28:1, 16:0/25:0/27:1, 19:0/24:0/25:1
1370.0908	1370.0918	−1.01	20.68	C81H152O14Na	18:0/24:0/27:2	18:0/24:1/27:1, 18:1/24:0/27:1
1372.1052	1372.1074	−1.22	100	C81H152O14Na	18:0/24:0/27:1
1384.1050	1384.1074	−1.48	11.46	C82H152O14Na	18 0/25 1/27 1	18:0/27:1/25:1, 16:0/27:1/27:1
1386.1232	1386.1231	0.09	11.88	C82H154O14Na	18:0/24:0/28:1	19:0/24:0/27:1, 18:0/25:0/27:1, 16:0/27:0/27:1
1388.1005	1388.1023	−1.89	10.9	C81H152O15Na	18:0/βh24:0/27:1
1398.1228	1398.1231	−0.25	8	C83H154O14Na	18:0/26:1/27:1	19:0/25:1/27:1, 17:0/27:1/27:1, 16:0/27:1/28:1
1400.1382	1400.1387	−0.55	9.63	C83H154O15Na	b
1402.1168	1402.1180	−1.21	23.79	C82H154O15Na	18:0/βh27:0/25:1	16:0/βh27:0/27:1, 18:0/β25:0/27:1
1412.1387	1412.1387	0.01	8.91	C84H156O14Na	b
1416.1334	1416.1336	−0.23	7.78	C83H156O15Na	b
1428.1336	1428.1336	−0.04	13.25	C84H156O15Na	18:0/βh27:1/27:1	18:1/βh27:0/27:1
1430.1495	1430.1493	0.23	24.92	C84H158O15Na	18:0/βh27:0/27:1
1444.1636	1444.1649	−1.35	5.51	C85H160O15Na	b

^aSpecies abundance in descending order.

^bStructure not defined.

MS² on the ion at m/z 1372.1 (Figure 5a) yielded a prominent ion at m/z 1210.1, arising from loss of glucose, suggesting that all of the fatty acyl substituents in the molecule are attached to the same sugar ring. The spectrum also contained ions at m/z 1087.8 and 1003.8 arising from losses of 18:0- and 24:0-FA substituents, respectively, along with ions at m/z 925.8 and 841.8, arising from further losses of 18:0- and 24:0-FA from the ion at m/z 1210.1, respectively. This further fragmentation process is supported by MS³ on the ion at m/z 1210.1 [1372 → 1210 (Figure 5b)], which gave rise to ions at m/z 925.8 and 841.8, and the ion at m/z 557, arising from the combined losses of the 18:0- and 24:0-FA moieties. The MS⁴ spectrum of the ion at m/z 557 [1372 → 1210 → 557 (Figure 5c)] contained an ion at m/z 431 representing a sodiated 27:1, a mycolipenic acid, along with ions at m/z 529, 489, 473, and 457 arising from the rupture of the sugar ring (Scheme 4), indicating the attachment of a 27:1-fatty acyl substituent to C6 of the glucose ring (Glc 1). These results point to the presence of a 2,3,6-triacyl trehalose (2-18:0, 3-24:0, 6-27:1 TAT) structure. In Figure 5a, the ion at m/z 968 (1372 – 408) arising from loss of the 27:1-FA substituent is absent. The ion at m/z 801 (1210 – 408.5) arising from similar loss of 27:1-FA is also absent in Figure 5b, indicating that loss of the 27:1-FA substituent at position 6 is a less favorable fragmentation process. This discrimination in the FA loss due to its location on the Glc ring may be useful for assignment of the FA substituent location.

Figure 5.

(a) MS² spectrum of the [M + Na]⁺ ion of 2-stearyl, 3-2,4-dimethyldimethyltetracosanoyl, 6-2,4,6-trimethyltetracos-2-enoyl trehalose (18:0/24:0/27:1-TAT) at m/z 1372.1, (b) its MS³ spectrum of the ion at m/z 1210.1 (1372 → 1210), and (c) MS⁴ spectrum of the ion at m/z 557 (1372 → 1210 → 557).

Scheme 4. Major Fragmentation Pathways for the [M + Na]⁺ Ions of 18:0/24:0/27:1-TAT at m/z 1372^a.

^aAll the ions are sodiated ions.

Similarly, the MS² spectrum of the [M + Na]⁺ ion at m/z 1430.4 (Figure 6a) contained the prominent ion at m/z 1268 arising from loss of glucose, together with ions at m/z 1146 and 1003.9 arising from elimination of the 18:0- and h27:0-fatty acid substituents at positions 2 and 3, respectively, and the ion at m/z 1022 arising from loss of the 27:1-FA at position 6 is absent. The notion that the h27:0-FA substituent represents a β-OH fatty acid substituent is supported by the presence of the ion at m/z 1078, arising from loss of a CH₃(CH₂)₂₂CHO residue by cleavage of the C2─C3(OH) bond of the 3-hydroxy fatty acid substituent as seen above (Scheme 2).

Figure 6.

(a) MS² spectrum of the [M + Na]⁺ ion at m/z 1430, (b) its MS³ spectrum of the ion at m/z 1268 (1430 → 1268), and (c) MS⁴ spectrum of the ion at m/z 983 (1430 → 1268 → 983), which led to the assignment of the 18:0/βh27:0/27:1-TAT structure.

Further dissociation of the ion at m/z 1268.4 [1430 → 1268 (Figure 6b)] gave rise to ions at m/z 984.1 and 841.9, arising from elimination of 18:0- and h27:0-fatty acid substituents at positions C2 and C3, respectively, along with the ion at m/z 916.0, arising from elimination of a CH₃(CH₂)₂₂CHO residue by cleavage of the C2─C3(OH) bond, consistent with the presence of the βh27:0-FA substituent at position 3 of Glc 1. The ions at m/z 631 and 557 (Figure 6b) arose from further losses of CH₃(CH₂)₂₂CHO and h27:0-FA, respectively. These ions are seen in the MS⁴ spectrum of the ion at m/z 984.1 [1430 → 1268 → 984.1 (Figure 6c)], supporting the fragmentation processes. The MS⁴ spectrum of the ion at m/z 557 [1430 → 1268 → 557 (data not shown)] is identical to that shown in Figure 5c, consistent with the notion that the 27:1-FA substituent is located at position 6. The results presented above led to the assignment of the 18:0/βh27:0/27:1-TAT structure.

Characterization of PAT.

The structure of PAT was previously defined as 2,3,6,2′,4′-pentaacyl trehalose.^5,14 Figure 1d shows the high-resolution ESI-MS spectrum of the PAT eluted at 38.82–43.92 min seen as the [M + Na]⁺ ions, which are similar to the profiles seen as the [M + NH₄]⁺ (Figure S1d, top panel) and [M + HCO₂]⁻ ions (Figure S1d, bottom panel). Table 3 summarizes the structures of the major PAT species determined by LIT MSⁿ on the [M + Na]⁺ ions. For example, the MS² spectrum of the major ion at m/z 2152.9 (Figure 7a) contained the ions at m/z 1868 and 1784 arising from losses of 18:0- and 24:0-FA acid substituents, respectively, along with ions at m/z 1500 arising from the combined losses of the 18:0- and 24:0-FA substituents. The spectrum also contained the ions at m/z 1210.1, representing a sodiated 2-stearoyl (18:0), 3-(2,4,6)-trimethyldocosanoyl (24:0), 6-(2,4,6)-trimethyltetracosenoyl (27:1) glucose (18:0/24:0/27:1-triacylglucose) similar to that seen for TAT (Figure 5a). This structural information is further supported by the MS³ spectrum of the ion at m/z 1210 (2152.9 → 1210.1) (Figure S2), which is identical to that shown in Figure 5b, indicating that the PAT molecule possesses the same (18:0/24:0/27:1)-triacylglucose moiety as TAT. The cleavage of the glycosidic bond to eliminate a (18:0/24:0/27:1)-triacylglucose also led to the ion at m/z 983, representing a sodiated 2′,4′-di(2,4,6)-trimethyltetracosenoyl-Glc (27:1/27:1-Glc) (Scheme 5). This structural information is further supported by the MS³ spectrum of the ion at m/z 983 [2152.9 → 983.6 (Figure 7c)], which contained ions at m/z 575 by loss of 27:1-FA and ions at m/z 533, 503, and 473 arising from cleavages of the glucose ring. The results are consistent with the presence of 27:1-FA substituents at positions 2′ and 4′ of Glc2 (Scheme 5). The structural information presented above led to the assignment of the (18:0/24:0/27:1)(27:1/27:1)-PAT structure.

Table 3.

PAT Species Found in a Cultured M. tuberculosis Biofilrn by High-Resoltion LIT MSⁿ

m/z [M + Na]+	theoretical mass (Da)	deviation (mDa)	relative intensity (%)	composition	major structure	minor structuresa
2040.7568	2040.7546	2.25	14.58	C127H236O16Na	(16:0/24:0/25:1)(25:1/25:1)	b
2054.7704	2054.7702	0.15	19.14	C128H238O16Na	(18:0/24:0/24:1)(25:1/25:1)	b
2066.7712	2066.7702	1	18.94	C129H238O16Na	(18:1/24:0/25:1)(25:1/25:1)	b
2068.7851	2068.7859	−0.72	39.89	C129H240O16Na	(18:0/24:0/25:1)(25:1/25:1)	b
2078.7719	2078 7702	1.71	6.27	C130H238O16Na	(18:0/24:2/26:1)(25:1/25:1)	b
2080.7852	2080.7859	−0.65	24.67	C130H240O16Na	(18:1/24:0/26:1)(25:1/25:1)	b
2082.7996	2082.8015	−1.96	43.28	C130H242O16Na	(18:0/24:0/26:1)(25:1/25:1)	b
2094.8015	2094.8015	−0.04	45.19	C131H242O16Na	(18:0/24:0/25:2)(27:1/25:1)	(18:1/24:0/27:1)(25:1/25:1), (16:1/24:0/27:1)(25:1/27:1)
2096.8146	2096.8172	−2.53	75.72	C131H244O16Na	(18:0/24:0/25:1)(27:1/25:1)	(18:0/24:0/27:1)(25:1/25:1), (16:0/24:0/27:1)(25:1/27:1), (16:0/24:0/25:1)(27:1/27:1)
2106.8010	2106.8015	−0.53	14.53	C132H242O16Na	(18:0/25:2/25:1)(27:1/25:1)	b
2108.8169	2108.8172	−0.25	39	C132H244O16Na	(18:0/23:1/27:1)(27:1/25:1)	b
2110.8287	2110.8328	−4.13	50.48	C132H246O16Na	(18:0/24:0/26:1)(27:1/25:1)	(16:0/24:0/26:1)(27:1/27:1), (18:0/24:0/24:1)(27:1/27:1), (15:0/27:0/28:1)(25:1/25:1)
2120.8166	2120.8172	−0.54	23.31	C133H244O16Na	(18:0/24:1/27:2)(27:1/25:1)	b
2122.8320	2122.8328	−0.78	58.94	C133H246O16Na	(18:0/24:0/27:2)(27:1/25:1)	(18:0/24:0/25:2)(27:1/27:1)
2124.8453	2124.8465	−3.15	87.85	C133H248O16Na	(18:0/24:0/27:1)(27:1/25:1)	(18:0/24:0/25:1)(27:1/27:1), (16:0/24:0/27:1)(27:1/27:1)
2136.8470	2136.8485	−1.42	34.55	C134H248O16Na	(18:0/25:1/27:1)(27:1/25:1)	b
2140.8478	2140.8434	4.44	18.71	C133H248O17Na	(18:0/βh27:1/24:0/)(27:1/25:1)	b
2150.8617	2150.8641	−2.41	40.94	C135H250O16Na	(18:0/24:0/27:2)(27:1/27:1)	(16:0/26:0/27:2)(27:1/27:1)
2152.8762	2152.8798	−3.52	58.29	C135H252O16Na	(18:0/24:0/27:1)(27:1/27:1)	(16:0/26:0/27:1)(27:1/27:1), (18:0/26:0/27:1)(25:1/27:1), (18:0/26:0/25:1)(27:1/27:1)
2154.8608	2154.859	1.76	20.44	C134H250O17Na	(18:0/βh27:1/25:0)(27:1/25:1)	b
2164.8758	2164.8798	−3.98	25.71	C136H252O16Na	(18:0/25:1/27:1)(27:1/27:1)	b
2168.8728	2168.8747	−1.83	15.65	C135H252O17Na	(18:0/βh27:1/24:0)(27:1/27:1)	b
2168.9094	2168.9111	−1.71	8.39	C136H256O16Na	(18:0/27:0/25:0)(27:1/27:1)	b
2178.8939	2178.8954	−1.54	11.23	C137H254O16Na	(18:0/26:0/27:2)(27:1/27:1)	b
2180.9119	2180.9111	0.87	9.07	C137H256O16Na	(18:0/26:0/27:1)(27:1/27:1)	b
2182.8867	2182.8903	−3.62	26.15	C136H254O17Na	(18:0/βh27:0/27:1)(27:1/25:1)	(18:0/βh27:0/25:1)(27:1/27:1)
2192.9088	2192.9111	−2.25	10.31	C138H256O16Na	(18:0/27:1/27:1)(27:1/27:1)	b
2194.9207	2194.9267	−6.04	7.18	C138H258O16Na	(18:0/27:0/27:1)(27:1/27:1)	b
2196.9018	2196.906	−4.19	9.32	C137H256O17Na	(18:0/βh27:1/26:0)(27:1/27:1)	(18:0/βh27:0/26:1)(27:1/27:1)
2206.8915	2206.9169	1.12	3.55	C138H256O16Na	(18:0/27:1/28:1)(27:1/27:1)	b
2208.9044	2208.906	−1.58	10.46	C138H256O17Na	(18:0/βh27:1/27:1)(27:1/27:1)	b
2210.9167	2210.9216	−4.89	15.01	C138H258O17Na	(18:0/βh27:0/27:1)(27:1/27:1)	b
2220.9522	2220.9424	9.83	1.48	C140H260O16Na	(20:0/27: 1/27: 1)(27: 1/27: 1)	(18:0/29:1/27:1)(27:1/27:1)
2222.9240	2222.9216	2.34	1.76	C139H258O17Na	(18:0/βh27:1/28:1)(27:1/27:1)	b
2234.9538	2234.958	−4.27	0.58	C141H262O16Na	b	b
2248.9758	2248.9736	2.21	0.5	C142H264O16Na	b	b

^aSpecies abundance in descending order.

^bStructure not defined.

Figure 7.

(a) MS² spectrum of the [M + Na]⁺ ion of the major (18:0/24:0/27:1)(27:1/27:1)-PAT at m/z 2152.9 and its MS³ spectra of the ions at (b) m/z 983.6 (2152.9 → 983.6) and (c) m/z 955.6 (2152.9 → 955.6).

Scheme 5. Major Fragmentation Pathways for the [M + Na]⁺ Ions of (18:0/24:0/27:1)(27:1/27:1)-PAT at m/z 2152^a.

^aAll the ions shown in the scheme are sodiated species, and “Na⁺” is omitted for simplicity.

In Figure 7a, minor ions at m/z 1896 and 1756 arising from loss of 16:0- and 26:0-FA substituents, respectively, were also present. These ions paired with ions at m/z 1500 (losses of 16:0 and 26:0) and m/z 1091 (losses of 16:0, 26:0, and 27:1), signifying the presence of a 16:0/26:0/27:1-triacylglucose moiety, which is paired with the ion at m/z 983 representing the sodiated 2′,4′-di-27:1-Glc. The results led to the assignment of a (16:0/26:0/27:1)(27:1/27:1)-PAT minor isomer. The spectrum also contained minor ions at m/z 1471.9 (losses of 18:0 and 26:0) and m/z 1063 (losses of 18:0, 26:0, and 27:1) along with the ion at m/z 1238, representing a sodiated (18:0/26:0/27:1)-triacylglucose. These ions are paired with the ions at m/z 955 representing a sodiated 27:1/25:1-diacyl glucose, indicating that a minor (18:0/26:0/27:1)(27:1/25:1)-PAT isomer is also present. The presence of 27:1/25:1-diacyl (or 25:1/27:1-diacyl) glucose is further confirmed by the MS³ spectrum of the ion at m/z 955 (Figure 7c), which contained ions at m/z 529 and 547 arising from loss of 27:1- and 25:1-FA substituents, respectively, along with ions at m/z 505/473 and 533/445 arising from cleavages of the Glc ring similar to those seen in Figure 7b.

Characterization of the Novel TetraAT Family.

A tetraacyltrehalose (TetraAT) family was reported by Touchette and co-workers, but the structure was not identified.¹⁶ Our positive ion ESI LC/MS (eluted at 32.23–35.25 min) mass spectrum contained the [M + NH₄]⁺ ion series (Figure S1c, top panel), which is nearly identical to that observed as the [M + HCO₂]⁻ ions (Figure S1c, bottom panel) in the negative ion mode. The presence of this TetraAT family is further supported by the observation of the [M + Na]⁺ ions by HRMS (Figure 1c and Table 4). The characterization of this lipid subfamily is exemplified by the LIT MS² spectrum of the ion at m/z 1762 (Figure 8a), which contained ions at m/z 1478 (loss of 18:0-FA), m/z 1394 (loss of 24:0-FA), and m/z 1110 (losses of 18:0-FA and 24:0-FA), along with ions at m/z 1210, and 593, arising from cleavage of the glycosidic bond to form the sodiated 18:0/24:0/27:1-triacyl Glc and 27:1-Glc residues, respectively (Scheme 3). The MS³ spectrum of the ion at m/z 1209.9 (data not shown) is identical to the spectrum in Figure 5b, indicating that the molecule possesses the same (18:0/24:0/27:1)-triacyl Glc moiety as seen for TAT and PAT earlier (i.e., an identical triacyl Glc substituent). The spectrum (Figure 8a) also contained the ion at m/z 1354 arising from loss of a 27:1-FA substituent, likely at position 2′ of the trehalose core. This speculation is based on the findings that the MS³ spectrum of the ion at m/z 1354 [1762 → 1354 (Figure 8b)] contained ions at m/z 1192 and 1210, arising from further loss of Glc2, along with ions at m/z 1252, 1264, and 1324, likely arising from cleavages of the Glc ring (Scheme 6). This assignment of the 27:1-FA substituent at position 2′ of the trehalose is also supported by the MS³ spectrum of the ion at m/z 593 [1762 → 593 (Figure 8c)] and the MS⁴ spectrum of the ion at m/z 575 (1762 → 593 → 575) (Figure S3). The former spectrum contained ions at m/z 431, representing a sodiated 27:1-FA, along with ions at m/z 473, 503, and 533 arising from cleavages of the Glc ring (see the inset for the fragmentations). The latter spectrum also contained the ion at m/z 431, together with ions at m/z 545, 515, 503, 473, and 443, arising from cleavages across the Glc ring (see the inset for the fragmentations). Taken together, this structural information defines the 2-stearyl, 3-tetracosanoyl, 6-mycolipenyl, 2′-mycolipenyl trehalose [(18:0/24:0/27:1)(27:1)-TetraAT] structure, rather than the 2,3,6,4′-tetraacyl trehalose structure previously proposed.¹⁶

Table 4.

TetraAT Species Found in a Cultured M. tuberculosis Biofilm by High-Resolution LIT MSⁿ

m/z [M + Na]+	theoretical mass (Da)	deviation (mDa)	relative intensity (%)	composition	major structure	other isomersa
1610.3399	1610.3371	2.83	3.66	C97H182O15Na	b
1624.3559	1624.3527	3.12	2.26	C98H134O15Na	b
1636.3551	1636.3527	2.37	5.62	C99H184O15Na	(16:0/18:0/26:1)(27:1)
1638.3703	1638.3684	1.89	5.74	C99H186O15Na	b
1650.3697	1650.3684	1.34	12.66	C100H186O15Na	(16:0/18:0/27:1)(27:1)
1662.3698	1662.3684	1.38	3.14	C101H186O15Na	b
1664.3855	1664.384	1.48	11.27	C101H188O15Na	(18:0/24:0/22:1)(25:1)	(16:0/24:0/22:1)(27:1)
1676.4011	1678.3997	1.37	18.76	C102H188O15Na	(18:0/24:0/23:1)(25:1)	(16:0/24:0/23:1)(27:1)
1690.4010	1690.3997	1.28	11	C103H190O15Na	b
1692.4173	1692.41 53	1.96	19.6	C103H192O15Na	(18:0/24:0/24:1)(25:1)	(16:0/24:0/24:1)(27:1)
1702.4002	1702.39 97	0.51	3.97	C104H190O15Na	b
1704.4166	1704.4153	1.23	21.04	C104H192O15Na	(18:0/24:0/25:2)(25:1)	(16:0/24:0/25:2)(27:1)
1706.4327	1706.431	1.66	40.96	C104H194O15Na	(18:0/24:0/25:1)(25:1)	(16:0/24:0/25:1)(27:1)
1716.4165	1716.4153	1.12	6.37	C105H192O15Na	b
1718.4326	1718.431	1.56	21.7	C105H194O15Na	(18:0/23:1/25:1)(27:1)	(16:0/25:1/25:1)(27:1)
1720.4474	1720.4466	0.76	32	C105H196O15Na	(18:0/24:0/24:1)(27:1)	(16:0/24:1/26 1)(27:1), (18:0/24:0/25:1)(25:1), (17:0/23:0/26:1)(27:1)
1730.4323	1730.431	1.26	11.86	C106H194O15Na	(18:1/25:1/27:1)(25:1)	(16:1/25:1/27:1)(27:1), (18:1/24:1/26:1)(27:1)
1732.4477	1732.4466	1.09	43.61	C106H196O15Na	(18:0/24:1/25:1)(27:1)	(16:1/24:0/27:1)(27:1), (18:0/24:0/25:1)(27:2), (18:0/24:0/27:1)(25:2), (16:0/24:1/27:1)(27:1), (18:1/24:0/27:1)(25:1)
1734.4632	1734.4623	0.89	83.02	C106H198O15Na	(18:0/24:0/27:1)(25:1)	(16:0/26:0/27:1)(25:1)
1744.4478	1744.4466	1.16	11.69	C107H196O15Na	b
1746.4629	1746.4623	0.59	29.78	C107H198O15Na	(18:0/25:1/25:1)(27:1)	(16:0/25:1/27:1)(27:1), (18:0/25:1/27:1)(25:1)
1748.4776	1748.4779	−0.38	34.69	C107H200O15Na	b
1756.4472	1756.4466	0.58	4.45	C108H196O15Na	b
1758.4631	1758.4623	0.84	14.81	C108H198O15Na	(18:1/24:1/27:1)(27:1)	(18:0/24:0/27:2)(27:2), (18:1/24:0/27:1)(27:2)
1760.4777	1760.4779	−0.23	52.27	C108H200O15Na	(18:0/24:0/27:2)(27:1)	(16:0/26:0/272)(27:1), (18:1/24:0/27:1)(27:1), (18:0/24:1/27:1)(27:1), (18:1/24:0/27:1)(27:1), (16:0/26:1/27:1)(27:1)
1762.4935	1762.4936	−0.11	88.25	C108H202O15Na	(18:0/24:0/27:1)(27:1)	(16:0/26:0/27:1)(27:1)
1772.4778	1772.4779	−0.16	15.14	C108H200O15Na	(18:0/25:1/27:2)(27:1)
1774.4934	1774.4936	−0.24	33.98	C109H202O15Na	(18:0/25:1/27:1)(27:1)
1778.4885	1778.4885	−0.02	22.05	C108H202O16Na	(18:0/βh24:0/27:1)(27:1)
1786.4924	1786.4936	−1.22	7.3	C110H202O15Na	b
1788.5100	1788.5092	0.74	15.39	C110H204O15Na	(18:0/26:1/27:1)(27:1)
1790.4893	1790.4885	0.76	18.51	C109H202O16Na	(18:0/βh27:0/25:2)(27:1)
1792.5030	1792.5042	−1.17	31.32	C109H204O16Na	(18:0/βh27:0/25:1)(27:1)	(16:0/βh27:0/27:1)(27:1)
1802.5243	1802.5249	−0.6	17.41	C111H206O15Na	(18:0/27:1/27:1)(27:1)
1814.5239	1814.5249	−1.02	3.01	C112H206O15Na	b
1818.5194	1818.5198	−0.44	14.96	C111H206O16Na	(18:0/βh27:1/27:1)(27:1)	(18:1/βh27:0/27:1)(27:1)
1820.5355	1820.5355	−1.91	22.03	C111H208O16Na	(18:0/βh27:0/27:1)(27:1)
1830.5196	1830.5198	−0.26	2.14	C112H208O16Na	b
1832.5363	1832.5355	0.83	4.48	C112H208O16Na	b
1834.5503	1834.5511	−0.83	5.53	C112H210O16Na	b
1844.5724	1844.5718	0.56	2.21	C114H212O15Na	b
1848.5616	1848.5668	−5.17	2.05	C113H212O16Na	b

^aSpecies abundance in descending order.

^bStructure not defined.

Figure 8.

(a) LIT MS² spectrum of the [M + Na]⁺ ions of the main (18:0/24:0/27:1)(27:1)-TetraAT at m/z 1762 and its MS³ spectra of the ions at (b) m/z 1354 (1762 → 1354) and (c) m/z 593 (1762 → 593). The proposed fragmentation pathway of the ion at m/z 593 is shown in the inset of panel c.

Scheme 6. Major Fragmentation Pathways for the [M + Na]⁺ Ions of (18:0/24:0/27:1)(27:1)-TetraAT at m/z 1762.

^aAll the ions shown in the scheme are sodiated species, and “Na⁺” is omitted for simplicity.

In Figure 8a, another set of ions at m/z 1506 (loss of 16:0) and m/z 1366 (loss of 26:0) are also present, consistent with the observation of the ions at m/z 1098 (loss of 16:0-FA) and m/z 958 (loss of 26:0-FA) in Figure 7b, signifying the presence of a (16:0/26:0/27:1)-Glc moiety. The presence of this moiety combined with the presence of the ion at m/z 593, representing a sodiated 27:1-Glc, readily led to the assignment of the (16:0/26:0/27:1)(27:1)-TetraAT minor isomer. The structures of the major species in this TetraAT family identified by the LIT MSⁿ approach are listed in Table 4.

Characterization of Mycolipanolic, Mycolipenic, Mycolipodienoic, and Mycoserosic Acid Substituents as AMPP Derivatives.

To determine the structure of the fatty acid substituents attached to the trehalose backbone, the FA released from hydrolysis was further derivatized to FA-AMPP derivatives, which were further subjected to high-resolution CID/HCD tandem mass spectrometry using an Orbitrap. Table 5 shows the released FAs, their FA-AMPP species, and the assigned structures. For example, the HCD MS² spectrum of the ion at m/z 575 (Figure 9a) gave rise to ions at m/z 169, 183, and 211 that are typical ions for FA-AMPP derivatives,^34,35 along with ions at m/z 251, 279, 293, and 321 that readily revealed the methyl side chain, and the double bond at position C2 (Δ²27:1) (see the inset scheme for the fragmentation processes). The results indicate that the molecule represents a M⁺ ion of 2,4,6-trimethyltetracos-2-enoic acid-AMPP, a 27:1-mycolipenic acid structure.

Table 5.

Fatty Acid Substituents on the Polyacyltrehalose Determined by LIT MSⁿ with High-Resolution Mass Spectrometry

FA substituents seen as [M − H]− ions				FA-AMPP derivative seen as M+ ions					Structure	Subclass	Compound name
m/z	Theo. Mass	Delta (mmu)	Composition	measured m/z	Theo. Mass	Delta (mmu)	Rel. Intensity	Composition
[M - H]-	Da	mDa		M+	Da	mDa	%
311.2955	311.2956	−0.08	C2O H39 O2	479.3996	479.3996	0.05	12.67	C32 H51 O N2	20:0-FA		eicosanoic acid
				481.3789	481.3789	0.05	2.08	C31 H49 O2 N2	h19:0-FA	#
				491.3998	491.3996	0.18	1.26	C33 H51 O N2	21:1-FA	#
				493.3789	493.3789	0.06	1.52	C32 H49 O2 N2	h20:1-FA	#
				493.4154	493.4152	0.16	7.32	C33 H53 O N2	21:0-FA	b	2-methyleicosanoic acid
				495.3946	495.3945	0.13	13.53	C32 H51 O2 N2	h20:0-FA		3-hydroxy-docosanoic acid
				505.4154	505.4152	0.16	1.34	C34 H53 O N2	22:1-FA	#
339.3267	339 3269	−0.2	C22 H43 O2	507.4311	507.4309	0.21	4.46	C34 H55 O N2	22:0-FA	b	2,4-dimethyleicosanoic acid
				509.4104	509.4102	0.24	1.75	C33 H53 O2 N2	h21:0-FA	#
				521.4104	521.4102	0.25	3.94	C34 H53 O2 N2	h22:1-FA	#
				521.4468	521.4465	0.3	9.48	C35 H57 O N2	23:0-FA		tricosanoic acid
				523.4261	523.4258	0.25	10.63	C34 H55 O2 N2	h22:0-FA	#
				533.4469	533.4465	0.37	11.00	C36 H57 O N2	24:1-FA	a	2,4-dimethyldocos-2-enoic acid
367.3580	367.3582	−0.13	C24 H47 O2	535.4625	535.4622	0.35	77.11	C36 H59 O N2	24:0-FA	b	2,4-dimethyldocosanoic acid
				537.4417	537.4415	0.28	5.61	C35 H57 O2 N2	h23:0-FA	#
				545.4470	545.4465	0.43	2.38	C37 H57 O N2	25:2-FA	d	2,4,6-trimethyldocos-2,13-dienoic acid
379.3580	379.3582	−0.15	C25 H47 O2	547.4626	547.4622	0.44	69.75	C37 H59 O N2	25:1-FA	a	2,4,6-trimethyldocos-2-enoic acid
381.3736	381.3738	−0.21	C25 H49 O2	549.4783	549.4778	0.43	9.22	C37 H61 O N2	25:0-FA	b	2,4-dimethyltricosanoic acid
383.3529	383.3531	−0.2	C24 H47 O3	551.4575	551.4571	0.43	6.21	C36 H59 O2 N2	h24:0-FA	c	3-hydroxy-2,4,6-trimethylheneicosanoic acid
				559.4627	559.4622	0.49	1.37	C38 H59 O N2	26:2-FA	#
393.3736	393.3738	−0.2	C26 H49 O2	561.4778	561.4778	0.01	13.12	C38 H61 O N2	26:1-FA	a	2,4,6-trimethyltricos-2-enoic acid
395.3893	395.3895	−0.12	C26 H51 O2	563.4935	563.4935	0.05	12.08	C38 H63 O N2	26:0-FA	b	2,4-dimethyltetracosanoic acid
397.3686	397.3687	−0.11	C25 H49 O3	565.4728	565.4728	0.06	7.32	C37 H61 O2 N2	h25:0-FA	c	3-hydroxy-2,4,6-trimethyldocosanoic acid
405.3737	405.3738	−0.13	C27 H49 O2	573.4779	573.4778	0.09	18.54	C39 H61 O N2	27:2-FA	d	2,4,6-trimethyltetracos-2,15-dienoic acid
407.3893	407.3895	−0.15	C27 H51 O2	575.4936	575.4935	0.06	100.00	C39 H63 O N2	27:1-FA	a	2,4,6-trimethyltetracos-2-enoic acid
409.4048	409.4051	−0.27	C27 H53 O2	577.5091	577.5091	−0.01	3.38	C39 H65 O N2	27:0-FA	b
411.3841	411.3844	−0.22	C26 H51 O3	579.4885	579.4884	0.08	3.72	C38 H63 O2 N2	h26:0-FA	c	3-hydroxy-2,4,6-trimethyltricosanoic acid
				587.4936	587.4935	0.16	1.97	C40 H63 O N2	28:2-FA	a	2,4,6-trimethyl 15,16-methylenepentacos-2-enoic acid
421.4049	421.4051	−0.17	C28 H53 O2	589.5093	589.5091	0.14	16.46	C40 H65 O N2	28:1-FA	a	2,4,6,16-tetramethyltetracos-2-enoic acid
423.3842	423.3844	−0.14	C27 H51 O3	591.4886	591.4884	0.21	3.20	C39 H63 O2 N2	h27:1-FA	#
423.4205	423.4208	−0.23	C28 H55 O2	591.5249	591.5248	0.11	1.56	C40 H67 O N2	28:0-FA	#
425.3998	425.4000	−0.18	C27 H53 O3	593.5043	593.5041	0.23	28.95	C39H6S02N2	h27:0-FA	c	3-hydroxy-2,4,6-trimethyltetracosanoic acid
439.4155	439.4157	−0.19	C28 H55 O3	607.5200	607.5197	0.25	6.50	C40 H67 O2 N2	h28:0-FA	c	3-hydroxy-2,4,6-trimethylpentacosanoic acid

^#Structuure not defined.

^aMycolipenoic acid.

^bMycoserosic acid.

^cMycolipanoic acid.

^dMycolipodienoic acid.

Figure 9.

HCD MS² spectra of the M⁺ ions of (a) 2,4,6-trimethyltetracos-2-enoic acid-AMPP at m/z 575, (b) 2,4,6-trimethyltetracos-2,15-dienoic acid-AMPP at m/z 573, (c) 2,4-dimethyltetracosanoic acid-AMPP (mycoserosic acid) at m/z 535, (d) 3-hydroxy-2,4,6-trimethyltetracosanoic acid-AMPP (27:0-mycolipanolic acid-AMPP) at m/z 593 [and its MS³ spectrum of the ion at m/z 241 (593 → 241) (panel d, inset)], (e) 2,4,6,15-tetramethyltetracos-2-enoic acid-AMPP at m/z 589, and (f) 2,4,6-tetramethyl 15,16-methylenetetracos-2-enoic acid-AMPP at m/z 587. The fragmentation pathways leading to the characterization of these LCFA with polymethyl side chains are shown in the inset in each panel. Note that ions marked with asterisks (panel f) are not structurally related and most likely arise from isobaric precursors.

In contrast, the ion at m/z 573 (Table 2) is 2 Da (H₂) lighter than the ion at m/z 575, indicating that the molecule may possess an additional double bond. The HCD MS² spectrum of the ion at m/z 573 (Figure 9b) contained ions at m/z 169, 183, 211, 251, 279, 293, and 321 identical to those seen in Figure 9a, indicating that the molecule consisted of the same 2,4,6-trimethyltetracos-2-enoic acid structure, but the spectrum also contained ions at m/z 419 and 475, arising from allylic cleavages of the double bond at C15 (see the inset of Figure 9b for the fragmentation scheme), leading to assignment of 2,4,6-trimethyltetracos-2,15-dienoic acid, a 27:2-mycolipodienoic acid, as previously reported.⁶

A similar approach was applied to define the structures of mycolipanolic acid and mycoserosic acid. As shown in Figure 9c, the HCD MS² spectrum of the ion at m/z 535 contained the AMPP compound feature ions at m/z 169, 183, and 211, along with ions at m/z 239, 253, 281, 309, 321, 323, etc., arising from cleavages of the C─C bonds along the mycoserosic acid chain. The 28 Da gap between the ions at m/z 211 and 239 and between the ions at m/z 253 and 281 clearly placed the methyl side chain at positions C2 and C4, respectively, of the long fatty acid chain (see the inset of Figure 9c for the fragmentation scheme), pointing to the 2,4-dimethyltetracosanoic acid structure.

The elemental composition deduced from the high-resolution mass measurement of the ion at m/z 593.5038 (calculated m/z 593.5040 Da) gave a formula of C₃₉H₆₅O₂N₂, which is one H₂O heavier than the ion at m/z 575. The HCD MS² spectrum of the M⁺ ion at m/z 593 (Figure 9d) is dominated by the ion at m/z 241, arising from loss of a C₁₈H₃₇CH(CH₃)CH₂CH(CH₃)CHO residue to form a M+ ion equivalent to a propanoic acid-AMPP via the facile β-cleavages of the C2─C3(OH) bond.³⁴ This fragmentation process is further supported by the MS³ spectrum of the ion at m/z 241 (Figure 9d, inset), which is dominated by ions at m/z 185 and 169 that are signal ions of FA-AMPP. The results indicated that a propanoic acid-AMPP is formed (see the inset of Figure 9d for fragmentation pathways), following loss of an aldehyde residue. Taken together, the structural information readily locates the β-OH group and defines the 3-hydroxy-2,4,6-trimethyltetracosanoic acid structure, a 27:0-mycolipanolic acid.

In Table 5, an ion at m/z 589.5093 (calculated C₄₀H₆₅ON₂ m/z 589.5091) was also observed. MS² of the ion at m/z 589 (Figure 9e) gave rise to the ions at m/z 251, 279, 293, and 321 similar to those seen in panels a and b of Figure 9, indicating the presence of a 2,4,6-trimethyltetracos-2-enoic acid structure. The spectrum also contained the ion series m/z 335, 349, …, 391, 405, 419, 433, 447, 475, 489, etc., arising from cleavages of the C─C bonds along the FA chain. However, a 28 Da (C₂H₂) gap was seen between m/z 447 and 475, indicating the attachment of a methyl side chain at C15, leading to the identification of a 2,4,6,15-tetramethyltetracos-2-enoic acid structure (see the inset of Figure 9e for fragmentation pathways).

High-resolution mass measurement of the minor ion at m/z 587.4936 (calculated C₄₀H₆₃ON₂ m/z 587.4935) (Table 5) indicates that the ion is H₂ lighter than the ion at m/z 589 and probably possesses an additional unsaturated bond. High-resolution MS² on the ion at m/z 587 contained ions at m/z 251, 279, 293, and 321 similar to those seen in Figure 9e (also Figure 9a,b), suggesting that the molecule consists of the same 2,4,6-trimethyltetracos-2-enoic acid structure. The spectrum also contained the ion series m/z 335, 349, …, 391, 405, 419, 433, 473, 475, and 489, with a 42 Da gap (C₃H₄) between the ions at m/z 433 and 475, a pattern that identifies the cyclopropyl group chain.³⁴ The results provided the assignment of a 2,4,6-trimethyl 15,16-methylenetetracos-2-enoic acid structure. The observation of this molecule along with the presence of 2,4,6-trimethyltetracos-2,15-dienoic acid [m/z 573 (Figure 9b)] and 2,4,6,15-tetramethyltetracos-2-enoic acid [m/z 589 (Figure 9e)] is consistent with the classic synthetic pathways that utilize S-adenosylmethionine (SAM) for formation of the cyclopropyl group and methyl side chain from a double bond.^36,37

CONCLUSION

Employing LIT MSⁿ with high-resolution mass spectrometry in combination with HPLC separation and chemical reactions, we are able to completely define the structures of DAT, TAT, PAT, and a new TetraAT subfamily found in M. tuberculosis cultured as biofilms, revealing hundreds of molecular species, including many isomers. The TetraAT subfamily is defined as 2,3,6,2′-tetraacyl trehalose, which also contained FA substituents similar to those of the other subfamilies.

Both Belardinelli et al. and Touchette et al. reported that biosynthesis of PAT requires MmpL10 to transport DAT to the cell surface, and polymethyl-branched fatty acyl substituents, including mycosanoyl, mycolipenoyl, and/or mycolipanolyl substituents, released by DAT are used to form PAT by transesterification reactions catalyzed by the acyltransferase Chp2 on the periplasmic face of the plasma membrane.^15,16 From our HPLC/MS analysis (Figure S1), we estimated that the PAT/TetraAT/TAT/DAT abundance ratio was close to 1.25/1/1.3/1, and DAT, TAT, TetraAT, and PAT all contain similar 2-palmitoyl(stearoyl) and polymethyl-branched fatty acyl substituents at positions 2′, 3, 4′, and/or 6. Our findings are consistent with the synthetic pathways of PAT. However, it is also reasonable to speculate that biosynthesis of TAT and TetraAT may also share the same processes, and formation of PAT can also arise from further transesterification of TAT and TetraAT. We analyzed the surface glycolipids extracted with hexane, the same protocol used by Touchette and co-workers. We found that both DAT and TAT are absent in the extracts, similar to their findings.¹⁶ The lack of DAT and TAT can likely be attributed to the fact that DAT and TAT are more polar than TetraAT and PAT and cannot be efficiently extracted with hexane.

To convert the polymethyl-branched fatty acyl substituents to their AMPP derivatives, we are able to precisely determine the structures of mycolipanolic, mycolipenic, mycolipodienoic, mycoserosic acids, and a new cyclopropl-mycolipenic acid, using ESI LIT MS² with high-resolution mass spectrometry. The approach is simple, and more importantly, the method is sensitive; thus, structures of the minor species can be identified. For example, minor FA species such as 2-methyleicosanoic acid (m/z 493.4154) and 2,4,6-trimethyl 15,16-methylenepentacos-2-enoic acid (m/z 587.4936) are readily detectable as FA-AMPP M⁺ ions. However, the corresponding [M − H]⁻ ions are not detectable in the lipid acid hydrolysate (Table 5). The high-resolution MS² spectrum of the FA-AMPP also allows extraction of the accurate elemental composition of the authentic fragment ions, distinguishing them from the disingenuous fragment ions likely arising from isobaric precursors (in Figure 9f, ions marked with asterisks) (the authentic fragment ions from FAAMPP have to bear a unique −COC₁₂H₁₁N₂ tail; i.e., the formula must contain two nitrogens and one or two oxygens and have a double bond equivalent ≥8.5) (see Tables S1-S4), leading to accurate structural assignments. Therefore, the combination of semipreparative HPLC, chemical reactions, and high-resolution tandem mass spectrometry as described here affords a nearly complete identification of a complex microbial lipid family whose structures would be very difficult to define using other analytical methods.