Running title: Negative MALDI-MS/MS of neutral saccharides

Supplemental Dataas noted in the text

Structural characterization of neutral saccharides by negative ion MALDI mass spectrometry using a superbasic proton sponge as deprotonating matrix

C.D. Calvano*1,2, T.R.I. Cataldi1,2, J. F. Kögel3, A. Monopoli1, F. Palmisano1,2, J. Sundermeyer3

1Dipartimento di Chimica and 2Centro di Ricerca Interdipartimentale SMART -Università degli Studi di Bari Aldo Moro, via Orabona 4, 70126 Bari (Italy)

3Fachbereich Chemie, Philipps-Universitat Marburg, Hans-Meerwein-Straße, 35032 Marburg (Germany)

Number of Supplementary Figures: S1-S10

Number of SupplementaryTables: S1

Keywords:novel matrices, neutral saccharides, negative ion, superbasic proton sponge, tandem MALDI-ToF MS, fragmentation.

Figure S1. Negative-ion MALDI-ToF/ToF mass spectra of precursor ions at m/z 101 (A) and m/z 71 (B) observed in the full mass spectra of monosaccharides (see Figure 1). TPPN was used as matrix.

Figure S2. Negative-ion MALDI-ToF/ToF mass spectra of precursor ions at m/z 281 (A) and m/z 251 (B) observed in the full mass spectra of monosaccharides (see Figure 1).TPPN was used as matrix.

Figure S3.Negative ion MALDI-ToF/ToF mass spectra of precursor ions at m/z179.06 galactopyranose(A) and mannopyranose(B) by using TPPN as a matrix.

Figure S4. Negative-ion MALDI-ToF/ToF mass spectra at m/z 214of (A) galactose, (B) fructose (C) glucose, and (D) mannose by using nor-harman as a matrix.

Figure S5. Positive-ion MALDI-ToF/ToF mass spectra at m/z 203of (A) galactose, (B) fructose (C) glucose, and (D) mannose by using THAP as a matrix.

The feasibility of MALDI-MS/MS usingTPPN as a matrixfor a rapid screening of sugar alditols in food products is presented in Figure S6 where the outcome of a sugar-free chewing-gum sample are depicted. Based on the MS/MS spectra acquired for each ion (data not shown), it was possible to suggest the following assignments (Table S1):

Table S1.

Experimental Value (m/z) / Proposed Chemical Formula / Theor. Value (m/z) / Proposed Assignment
151.060 / [C5H12O5–H]– / 151.061 / Xylitol
181.075 / [C6H14O6–H]– / 181.072 / Sorbitol/Mannitol
223.086 / [C8H16O7–H]– / 223.082 / Xylitol + 3-oxoprop-1-en-2-olate
253.092 / [C9H18O8–H]– / 253.093 / Sorbitol + 3-oxoprop-1-en-2-olate
341.111 / [C12H22O11–H]– / 341.109 / Sucrose
343.126 / [C12H24O11–H]– / 343.125 / Maltitol
363.153 / [C12H28O12–H]– / 363.151 / Dimeric Sorbitol/Mannitol

Where xylitol, sorbitol/mannitol and maltitolcorrespond to the main ingredients reported in the chewing-gum label. As for monosaccharides, some detected peaks in the spectrum can be considered as a combination of alditols with fragments generated under basic pH by retro-aldolization. For instance, the peak at m/z 223.086 can be assigned as the deprotonated molecule resulting from the gas-phase adduct formation of xylitol and 3-oxoprop-1-en-2-olate; the peak at m/z 253.102 may be due to an adduct between sorbitol or mannitol with 3-oxoprop-1-en-2-olate; the peak at m/z 363.167 is a dimer of sorbitol or mannitol.

Figure S6. Negative-ion MALDI-ToF mass spectrum of an aqueous solution of a commercial chewing-gum sugar-free sample by using TPPN as a matrix.

Figure S7. Negative-ion MALDI-ToF/ToF mass spectrum of sucrose at m/z 341 occurring in a soy drink (A) and lactose at m/z 340 in bovine milk (B) by using TPPN as a matrix.

Figure S8. Sucrose intensity response vssucrose amount on the plate. Sucrose at different concentrations (from 100 pmol/µL down to 100 fmol/µL) was mixed with matrix (5 mM) and spotted. The absolute intensities of the peak at m/z 341 were plotted against sucrose amount showing a linear relationship. The regression line was described by the following equation:y=(-992.0±851.1)+(565.4±15.4)x; R2= 0.997. A signal-to-noise ratio ≥10 was obtained for the [M-H]− monoisotopic peak at 1 pmol of sucrose on plate, so the limit of quantitation can be estimated around 1 pmol/spot or 2 nmol/L.

Figure S9. Negative-ion MALDI-ToF/ToF mass spectra of chloride adduct at m/z 377 for (A) sucrose, (B) lactose and (C) maltose by using nor-harman as a matrix.

Figure S10. Positive-ion MALDI-ToF/ToF mass spectra of sodiated adducts at m/z= 365 for (A) sucrose, (B) lactose and (C) maltose by using THAP as a matrix.

Cyclodextrins

For the interpretation of Figure 7, indicating as R the C4H8O4 unit (120 Da), G the glucose unit (162 Da) and AH the anhydro hexose unit (144 Da) it is possible to explain all product ions observed in the spectrum from glycosidic bond and cross-ring cleavages as follows: [αCD–2R–H]– (731.24), [αCD–R–AH–H]– (707.24), [αCD–R–G–H]– (689.23), [αCD–2R–AH–H]– (587.21), [αCD–2R–G–H]– (569.20), [αCD–R–G–AH–H]– (545.19), [αCD–R–2G–H]– (527.19), [αCD–3G–H]– (485.18), [αCD–2R–2AH–H]– (443.16), [αCD–2R–G–AH–H]– (425.16), [αCD–2R–2G–H]– (407.13), [αCD–R–2G–AH–H]– (383.14), [αCD–R–3G–H]– (365.12), [αCD–3G–AH–H]– (341.12), [αCD–4G–H]– (323.12), [αCD–2R–G–2AH–H]– (281.10), [αCD–2R–2G–AH–H]– (263.08), [αCD–5G–H]– or [G–H]– (161.06) and [AH–H]– (143.05). The tandem mass spectrum of [βCD–H]– (m/z 1133), shown in plot B, reveals the same fragmentation pathway as [αCD–H]–.

1