Analysis of hair components by nanoparticle-assisted laser desorption/ionization mass spectrometry imaging

2.1 Sample preparation

Bleach-treated and untreated black hair of females approximately 20 years old was used in the test. Collected hairs were set in a Bio-slicer (KATANA, YAC DAStech, Inc., Saitama, Japan) and sliced lengthwise. To identify the region between bleached and growing (black) hair, hair containing a borderline was sliced according to visual judgement. After preparation of hair sections, we confirmed the presence/absence of a borderline by optical microscopy. Cryofilm (type2C(9), SECTION-LAB, Hiroshima, Japan) was placed on ITO-coated glass slides (Bruker Daltonik GmbH, Germany) using double-sided tape (Norinopod Plus Co., Tokyo, Japan). The sliced hair was placed on the adhesive side with the cross section facing up. Optical images of the tissues were obtained using a microscope scanner (Nanozoomer, Hamamatsu Photonics, Shizuoka, Japan) before analysis by MALDI- or Nano-PALDI MS imaging (MSI).

2.2 Preparation of nanoparticles (NPs)

NPs were prepared by adding an aqueous solution of FeCl₂·4H₂O (0.1 M; 20 mL) (FUJIFILM Wako Chemicals, Tokyo, Japan) to 20 mL of γ-aminopropyltriethoxysilane (Shin-Etsu Chemicals, Niigata, Japan). After stirring for 1 h at room temperature, the solution was centrifuged at 22,300 g (CF15RXII, Hitachi, Tokyo, Japan) at 4 °C for 1 h. The resulting precipitate was washed three times with ultrapure water. After washing, the NP precipitate was suspended in methanol.

2.3 MS and MSI by MALDI- and Nano-PALDI method

Two ionization methods, MALDI and Nano-PALDI, were used to detect and image target molecules. Target MS plates were coated with 10 mg mL⁻¹ of α-cyano-4-hydroxycinnamic acid (CHCA, Nacalai Tesque Inc., Kyoto, Japan) for MALDI or 10 mg mL⁻¹ solution of the matrix NPs for Nano-PALDI. A 1.0-µL aliquot of the standard, 10 pmol µL⁻¹ of cystine (Tokyo Chemical Industry, Tokyo, Japan) or 18-MEA (FUJIFILM Wako Chemicals, Tokyo, Japan), in 80 % methanol (residue-analysis grade, FUJIFILM Wako Chemicals) was added dropwise onto a coated plate. To confirm ionization of the standards, we used MALDI-TOF (rapifleX, Bruker Daltonik GmbH) for MS and MSI. To obtained tandem MS spectra, we used a collision-induced dissociation method.

For MALDI MSI, CHCA was sprayed on the hair tissue sections on ITO-coated glass slides using an automated pneumatic sprayer (TM-Sprayer, HTX Technologies, LLC). Ten spraying passes was made under the following conditions: flow rate, 120 μL/min; air flow, 10 psi; nozzle speed, 1,100 mm/min; temperature, 75 °C. For Nano-PALDI MSI, NPs were sprayed onto the hair tissue sections on the ITO-coated glass slides using an air brush (nozzle caliber, 0.2 mm).

After spraying the sections with an ionization-assisting reagent (CHCA or NPs), ionization in the laser spot areas was detected by scanning the sections. Mass spectra were acquired using a rapifleX (MALDI-TOF-MS) instrument and a 10-kHz pulse laser, and the laser spot areas (50 shots) were analyzed with a spot-to-spot center distance of 5 μm in each direction on the section. Signals between m/z 80–980 were collected. The tissue surface was irradiated with YAG laser shots in the positive ion detection mode. The laser power was optimized to minimize in-source decay of the targets. The obtained MS spectra were reconstructed to MS images with a mass bin width of m/z ± 0.1 from the exact mass using FlexImaging 5.1 software (Bruker Daltonik GmbH).

3.1 Characterization of nanoparticle

The XRD pattern of the nanoparticles indicated that the very sharp peaks at 2θ = 35 and 63°were confirmed to be the (1 3 0), (2 0 0) reflections and the (3 3 0), (0 6 0) reflections of macaulayite which was composed by iron oxide core. The FT-IR analysis (a) indicated the presence of O—H (3600–3100 cm⁻¹), C—H (2820–2970 cm⁻¹), C—N (1500–1480 cm⁻¹), and Si-O (1000–800 cm⁻¹) bonding as chemical bonds on the surface of the nanoparticles, even after several washing of the samples to remove physical adsorption of organic molecules. The diameter of the nanoparticle was measured to be 3.6 ± 0.06 nm (mean ± SEM; n = 300) (Supporting information) (Taira et al., 2008).

3.2 Detection of hair components using Nano-PALDI-TOF-MS and MALDI-TOF-MS

Both Nano-PALDI and MALDI MS analysis of standard cystine and 18-MEA detected at m/z 241.3 and 327.4, respectively, in the positive ion mode. The coefficients of determination (R²s) of semilog plots of the signal intensity versus amount were relatively no difference between Nano-PALDI and MALDI indicating that both ionization method has linearity. In Nano-PALDI, the limits of detection (LODs) of cystine and 18-MEA indicated 10 and 4.5 fmol, while those for MALDI were higher (Table 1). These results indicate that the sensitivity of Nano-PALDI is higher than that of MALDI.

Cystine and 18-MEA on hair sections were confirmed by tandem Nano-PALDI MS using the above-detected ions as precursor ions. As expected based on the chemical structure, a cystine fragment ion was observed at m/z 152.0 from standard sample (Fig. 1a) and on section (Fig. 1b). 18-MEA fragment ions were observed at m/z 310.3 and 254.3 from standard sample (Fig. 1c) and on section (Fig. 1d). We could confirmed exist of cystine and 18-MEA on section by Nano-PALDI MS. Because the m/z values of these fragmented ions on section were in good agreement with those of the respective standard samples (Fig. 1c and d). Detection of cystine may involve fragmentation from free cystine originally present in the hair or from proteins produced by laser irradiation of MS. Either way, cystine and 18-MEA are representative components to detect because it is an amino acid and lipid present in large amount in hair, reproducibly.

Close

Fig. 1

Tandem mass spectra of standard (a) and on section (b) of cystine and standard (c) and on section (d) of 18-MEA.

Export to PPT

3.3 Comparison of Nano-PALDI and MALDI-MSI

Fig. 2 shows the results of Nano-PALDI and MALDI-MSI analyses of cystine detected on samples of non-treated black hair fiber. Hair sections were sliced longitudinally without embedding. Optical images provide information regarding the hair region. The cuticle is localized as most outer layer. The cortex and medulla were confirmed as inner region and center of the hair fiber, respectively (Fig. 2a and d). To compare the spatial resolution, cystine was imaged by selecting m/z 241.3 as the protonated cystine ion (Fig. 2b, e). For Nano-PALDI-MSI, we observed that cystine (m/z 241.3) localized in the cuticle and cortex (Fig. 2b). MALDI-MSI also showed cystine localized in the cuticle and cortex, but the obtained images were blurry and exhibited background noise (Fig. 2e). We analyzed this difference by line profile analysis. Result in Nano-PALDI (Fig. 2g), line pattern indicated the lowest gray value at 0–20 and 80–100 pixels. On the other hand, two strong value areas observed at both end at hair region (20–70 pixels). The center of hair region as cortex showed broad value. In the case of MALDI (Fig. 2h), multiple peaks observed from all measured reason (0–100 pixels), especially strongest peak was detected at background region (0–20 and 70–100 pixels), indicating that matrix signal was higher than target signal and strongly involved high background signal. Notably, Nano-PALDI (Fig. 2b and g) MSI showed extreme lower background noise than that of MALDI (Fig. 2e and h). This difference in image quality was ascribed to the different characteristics of the ionization-assisting reagents. Our data suggest that only the target sample was ionized and the NPs themselves are not ionized in Nano-PALDI-MSI; therefore, no NP-derived signals are detected, resulting in a low-noise image. However, in the presence of organic matrix, both the matrix itself and the target sample are ionized (Shiono and Taira 2020). A common concern is that MALDI-based MSI is impractical for the imaging of low-molecular-weight targets due to the high background interference derived from organic matrices. In addition, MALDI, it is assumed that co-crystals formed by an organic matrix uniformly applied on the target surface are irradiated by the laser and ionized. Thus, for high-resolution images (<20 µm), the construction of small size of the matrix crystals required by an automated pneumatic sprayer. On the other hand, in Nano-PALDI, we easily applied nanoparticle on section using manual air brush (nozzle caliber, 0.2 mm) due to no need formation of co-crystals.

Close

Fig. 2

MS imaging of a hair sample by Nano-PALDI (a-e) and MALDI (f-j). Optical image of hair fiber (a and d), MS image of protonated cystine (m/z 241.3) by Nano-PALDI (b) and MALDI (e), merged images (c and f), line profile of MSI data by Nano-PALDI (g) and MALDI (h), and mass spectrum of hair section by Nano-PALDI (i) and MALDI (j).

Export to PPT

Hair section mass spectra are shown in Fig. 2i and j. The signal-to-noise ratio of Nano-PALDI (S/N 150) was 12.5 times higher than that of MALDI (S/N 12). Because MS spectrum in organic matrix showed several noises signal due to ionization of matrix itself. but NPs do not ionize itself. From line profile of images, Nano-PALDI showed lower background than that of MALDI. Nano-PALDI MSI is therefore suitable for imaging low molecular weight target molecules and small-sized samples.

3.4 Nano-PALDI-MSI of bleach-treated and untreated hair sections

We used a single hair fiber sample that could be divided into bleached and growing hair (untreated) regions at the center orange line. The term ‘untreated hair region’ refers to an area of new growth after bleaching. Fig. 3 shows Nano-PALDI-MSI data for cystine and 18-MEA in the bleach-treated and untreated regions. The optical images provide information on the bleached and untreated hair regions. The left side of the hair from the center is the bleached region, and the right side is the untreated region (Fig. 3a). Cystine was imaged by selecting m/z 241.3 and found to be localized in the cuticle and cortex of the untreated hair region; toward the right side, there was more localization of cystine from the cuticle to the cortex (Fig, 3b and c). This could have been due to the fact that cystine, which is present in abundance within untreated hair, leaked out due to the damage-associated loss of hydrophobicity of the hair surface (Korte et al., 2014). From line profile analysis, in the bleached region (Fig. 3b-i-iii), cystine imaging intensity was only one-third that of the untreated region. The closer to the root, the intensity of cystine increased at cuticle and cortex region (Fig. 3b-iv-vi).

Close

Fig. 3

MS imaging evaluation of sample borderline by Nano-PALDI. Optical image of hair fiber (a), MS image of protonated cystine (m/z 241.3) (b), merged image (c), line profile of cystine, MS image of protonated 18-MEA (m/z 327.4) (e), merged image (f), line profile of 18-MEA (g).

Export to PPT

18-MEA was imaged by selecting m/z 327.4 and found to be localized in the cuticle of the untreated hair region (Fig. 3d-x-xii and e). This result was in good agreement with a previous report that confirmed the localization of 18-MEA by atomic force microscopy (Tokunaga et al., 2019). In the bleach-treated region, 18-MEA was not imaged and not detected (Fig. 3d-vii-ix), although cystine was marginally imaged. We hypothesis that the difference was attributed to the easier loss of 18-MEA than cystine from the hair surface, as occurs with chemical treatments such as coloring and bleaching due to that 18-MEA which is more hydrophobic than cystine affected chemical treatment. Nano-PALDI MSI gave us clear identifiable image of the boundary region.