Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2023 Nov 13;95(47):17220–17227. doi: 10.1021/acs.analchem.3c02637

Direct 3D Sampling of the Embryonic Mouse Head: Layer-wise Nanosecond Infrared Laser (NIRL) Ablation from Scalp to Cortex for Spatially Resolved Proteomics

Jelena Navolić †, Manuela Moritz ‡, Hannah Voß ‡, Simon Schlumbohm §, Yannis Schumann §, Hartmut Schlüter ‡, Julia E Neumann †,∥,*, Jan Hahn ‡,*
PMCID: PMC10688223  PMID: 37956982

Abstract

graphic file with name ac3c02637_0005.jpg

Common workflows in bottom-up proteomics require homogenization of tissue samples to gain access to the biomolecules within the cells. The homogenized tissue samples often contain many different cell types, thereby representing an average of the natural proteome composition, and rare cell types are not sufficiently represented. To overcome this problem, small-volume sampling and spatial resolution are needed to maintain a better representation of the sample composition and their proteome signatures. Using nanosecond infrared laser ablation, the region of interest can be targeted in a three-dimensional (3D) fashion, whereby the spatial information is maintained during the simultaneous process of sampling and homogenization. In this study, we ablated 40 μm thick consecutive layers directly from the scalp through the cortex of embryonic mouse heads and analyzed them by subsequent bottom-up proteomics. Extra- and intracranial ablated layers showed distinct proteome profiles comprising expected cell-specific proteins. Additionally, known cortex markers like SOX2, KI67, NESTIN, and MAP2 showed a layer-specific spatial protein abundance distribution. We propose potential new marker proteins for cortex layers, such as MTA1 and NMRAL1. The obtained data confirm that the new 3D tissue sampling and homogenization method is well suited for investigating the spatial proteome signature of tissue samples in a layerwise manner. Characterization of the proteome composition of embryonic skin and bone structures, meninges, and cortex lamination in situ enables a better understanding of molecular mechanisms of development during embryogenesis and disease pathogenesis.

Introduction

In the past decade, there have been increasing efforts in mass spectrometry-based proteomics to analyze even smaller sample quantities down to single cells. Based on fluorescence-activated cell sorting (FACS), different cell types of a tissue sample could be analyzed with bottom-up proteomics.1,2 However, the spatial information and important components from the tissue microenvironment such as the extracellular matrix (ECM) are lost.3 Thus, many research groups have focused on methods for spatial proteomics. For example, different macroscopic brain regions (cerebellum, hippocampus, thalamus, striatum, olfactory bulb, prefrontal cortex) were dissected prior to subsequent bottom-up proteomics, resulting in coarse spatial resolution within the brain.4,5 This however does not sufficiently consider heterogeneities within the structures.3 In contrast, using Matrix-Assisted Laser Desorption and Ionization Mass Spectrometry Imaging (MALDI–MSI), it is possible to analyze different omics modalities while maintaining high spatial information.6−8 A laser beam scans a single tissue section in a grid-wise manner and allows a two-dimensional (2D) image of spatial protein or peptide distribution for proteomic analysis.9 Depending on the proximity of measured points, the resolution can be increased to 10 μm.7 Furthermore, computational methods reconstruct and stack the 2D images of single tissue sections into a final 3D MSI data set to investigate multidimensional protein distribution.7 While MALDI–MSI is well suited for the analysis of spatial peptide distribution, a drawback of this approach is that only a small part of the tissue proteome is detected due to suppression effects.8,10,11 Using MALDI–MSI only, a few hundred proteins can be identified, whereas LC-MS/MS-based methods for bottom-up proteomics reveal several thousands of proteins with quantitative information.8,10

Another approach used in spatial proteomics is laser-capture microdissection (LCM). This tool utilizes a pulsed laser for the dissection and isolation of a region of interest (ROI) on a tissue section in the submicrometer scale.12 To define the ROI for tissue sectioning, staining and visualization with a light microscope is required.3,13 The dissected sample had to be further processed according to the bottom-up proteomics workflow. Here, one crucial step is the sample homogenization, releasing the biomolecules from the original tissue sample. During this, enzymatic or mechanical degradation can occur, resulting in alterations of the protein composition.14

A novel approach to overcome the above-mentioned limitations is the nanosecond infrared laser (NIRL) for a fast simultaneous tissue sampling and homogenization.15−18 The process is based on tissue irradiation with ultrashort infrared laser pulses at a wavelength of 2940 nm. During this process, the energy is absorbed and converted into vibrational motion of the OH stretch band of water molecules in the cell. This causes an explosion, wherein the biomolecules are released in a homogenized aerosol and captured for further sample processing.19,20 As the biomolecules remain in their native form, it offers an insight into the proteome, which is closer to the original composition in the intact tissue.14,21,22 Furthermore, during the ablation, only the irradiated tissue region is removed without damaging the surrounding area,23 and thus, the method is well suited for its application in spatial proteomics. Recently, it was used to spatially resolve the layered structure of the murine intestine with a lateral resolution of 117 μm by subsequent bottom-up proteomics.16

In this study, we used NIRL-based tissue sampling to ablate directly from an intact, fresh frozen embryonic mouse head. Targeting the forebrain region, we gained even higher axial resolution with consecutive layers of 40 μm thick from the scalp to the cerebral cortex. It is known that corticogenesis is a highly orchestrated process and depends on gradients of specific markers or pathways. In the ventricular zone (VZ), the deepest layer of the cerebral cortex, radial glial cells (RGCs) proliferate and generate neurons known as direct neurogenesis. Alternatively, RGCs can give rise to intermediate progenitors (IPs) in the second germinal layer, the subventricular zone (SVZ). These IPs can further divide to generate pairs of neurons (indirect neurogenesis). Postmitotic cells leave the germinal layer and migrate radially along the RGCs to more superficial regions to form the subplate (SP), which later becomes the cortical plate (CP). Early-born projection neurons remain in deeper layers, while later-born neurons migrate in more superficial cortex layers.24,25 Spatially defined gradients of single factors or pathways are crucial for intact cortical layering. One well-known protein that regulates migration is Reelin, expressed by Cajal-Retzius cells in the marginal zone (MZ), the first layer of the cortex.26 Mutations in the human RELN gene revealed impaired neuron migration leading to irregular cortex layering associated with neural migration disorders (NMD) like agyria and cortical dysplasia.27,28 Characterization of layer-specific signatures in development and disease is therefore crucial to unravel the pathogenesis of cortical malformations.

Comprehensive analysis of proteome composition within the embryonic mouse head structures and especially for the heterogeneity of cortical layers has not yet been performed. Our study provides a three-dimensional map of protein abundances, enabling new insights into these complex structures at a developmental stage.

Experimental Section

Animal Handling and Preparation

All animals were kept at 12 h/12 h light/dark cycle with accessible water and food supply. The animals were handled according to local government guidelines. We used four control mice (FB1–4) with a C57BL/6 background at embryonic day E14.5. The mouse head was flash-frozen in liquid nitrogen-cooled 2-methylbutane. The base of the head was mounted with Tissue-Tek O.C.T. compound onto a glass slide. To standardize the ROI, the midline and left eye were marked with red color. The ROI for ablation was then placed near the marked regions. The mounted tissue was stored at −80 °C until laser ablation.

Laser Ablation and Sample Collection

A nanosecond infrared laser (NIRL) system was used for tissue ablation. The general laser system build-up is published,16 the main experiment-specific parameters will be further described. The wavelength of the laser was set to 2940 nm using a pulse width of 7 ns, pulsing with the maximum repetition rate of 20 Hz. A pulse energy of 650 μJ was measured at the sample position. A scan lens with a focal length of 100 mm was used for focusing the beam. The glass slide with the tissue was placed on a cooling stage, which was set to −5 °C during the ablation. The cooling stage was mounted onto a translation stage composed of two motorized linear stages (MLT25, Newport Corp., CA). A PTFE-coated glass slide with 12 wells (Epredia X5XER202WAD1, catalog-no. 17342650) was placed at a short distance (<1 mm) above the ROI (800 μm × 800 μm). Single wells were used to collect the plume material of a single layer. In total, nine layers were ablated from five hemispheres, and each layer was 2 × 25 μm deep and consisted of 112 laser pulses.

Thickness Measurement and Optical Coherence Tomography (OCT)

To estimate the number of layers to ablate, H&E-stained FFPE frontal sections of the mouse head were used. To determine the actual ablation volume, OCT measurements were performed before and after NIRL ablation of nine consecutive tissue layers. Additional information can be found in Supporting Information.

Sample Processing

Further sample processing was done with the protocol published by Tsai et al.29 The condensed sample aerosol was resuspended in 10 μL of 0.01% DDM (n-dodecyl β-D-maltoside) and transferred from the well into a protein low binding tube (Protein LoBind Tubes, Eppendorf SE, catalog-no. 0030108116). All other sample preparation steps were adapted from the mentioned protocol with the change to 20 ng trypsin for tryptic digestion.29 The samples were dried in a vacuum centrifuge and stored at −20 °C. Prior to mass spectrometric measurement, tryptic peptides were resuspended in 10 μL of 0.1% FA.

Fraction Library Preparation

Besides NIRL-based laser ablation of four different mouse heads, brain tissue was conventionally homogenized for building a fraction library to improve the search engine results in later data processing by matching between runs function. A more detailed description can be found in the Supporting Information.

Mass Spectrometric Data Acquisition

Liquid chromatography–tandem mass spectrometry (LC-MS/MS) measurements were performed on a Tribrid Mass spectrometer (Orbitrap Fusion, Thermo Fisher Scientific, Waltham, MA) coupled to a nano-UPLC (Dionex Ultimate 3000 UPLC system, Thermo Fisher Scientific, Waltham, MA)

Separation and elution of peptides were achieved by a linear gradient from 2 to 30% solvent B in 30 min at a flow rate of 0.3 μL/min. Eluting peptides were ionized by using a nanoelectrospray ionization source (nano-ESI) with a spray voltage of 1800 V, transferred into the MS, and analyzed in data-dependent acquisition (DDA) mode. More information in the Supporting Information.

Data Processing and Analyses

Database search was performed with Proteome Discoverer (version 2.4.1.15). Spectra were searched using Sequest HT against a reviewed murine Swissprot FASTA database obtained in December 2021 containing 17,085 entries. Further data processing and analysis were done in R studio (version 4.2.3). The log2-transformed data was used for the HarmonizR framework (version 0.0.0.9000)30 for batch effect reduction of pseudo batches (p batch) with the ComBat method (mode 1). More information about data processing and statistical analyses including prior batch effect reduction can be found in the Supporting Information.

Histological Staining

Hematoxylin-Eosin (H&E) staining was performed on acetone-fixed cryosections (8 μm) and on formalin-fixed and paraffin-embedded (FFPE) tissue sections (2 μm). Immunohistochemistry (IHC) with 3, 3-diaminobenzidine (DAB) was performed on FFPE sections with the Ventana Benchmark XT machine (Ventana, Tuscon). First antibodies used: MAP2C (M4403, 1:3000), NESTIN (611658, 1:200), KI67 (ab15580, 1:100), and SOX2 (ab97959, 1:200). The stained sections were scanned and digitalized with Hamamatsu Photonics K.K. and NDP.view (version 2.8.24). The in situ hybridization images (ISH) of embryonic mice at E13.5 were obtained from http://developingmouse.brain-map.org, accessed on 21.12.2022.31 The corresponding experiment IDs and sources are provided in the Supporting Information. All Images were processed with Photoshop Elements 15 and Fiji (ImageJ 2.1.032).

Results and Discussion

Laser Ablation Parameters and Quantification

In this study, we utilized NIRL-based 3D sampling and homogenization to directly ablate consecutive layers from fresh frozen embryonic mouse head for mass spectrometry-based spatial bottom-up proteomics. To estimate the required total laser ablation depth from the scalp to the ventricular zone, we used H&E-stained FFPE frontal head sections and measured the thickness from the skin surface to the ventricle margin (total thickness) at three different embryonic time points of development (Figure 1a, each time point n = 6–9, Supporting Information). As expected, there was no significant difference between the left (l) and right (r) hemispheres at any time point, while the mean total thickness was increasing over time, E14.5: x̅ = 294.6 μm (min = 203 μm, max = 445.7 μm), E16.5: x̅ = 623.3 μm (min = 548 μm, max = 683.9 μm), and for E18.5: x̅ = 957.8 μm (min = 613.4 μm, max = 1204 μm) (Figure 1a).

Figure 1.

Figure 1

Open in a new tab

Experimental setting for laser ablation-based sampling and homogenization of the embryonic mouse head at E14.5. (a) Quantification of skin and mouse cortex thickness on frontal H&E-stained sections at embryonic day E14.5, E16.5, and E18.5 (n = 6–9). r = right; l = left (Supporting Information). Mean and SD are shown. (b) Schematic illustration of the laser ablation setup. Laser wavelength: 2940 nm; repetition rate: 20 Hz; pulse energy: 650 μJ. NIRL = nanosecond infrared laser (c) The region of interest (ROI) was set to an area of 800 μm × 800 μm and ablated by 112 laser shots per layer. Nine layers were obtained in total. (d) Pictures of the specimen. The ROI was set on the hemisphere between the midline and the eye. (e) Frontal section of the mouse head. Left: scheme. Right: H&E-staining of a frozen section after the laser ablation. H = hippocampus, V = ventricle, VZ = ventricular zone, LGE = lateral ganglionic eminence, and MGE = medial ganglionic eminence. (f) Optical coherence tomography (OCT) analysis before and after the ablation with the manually segmented ROI in red (Supporting Information). Mean values are displayed (n = 3).

The ablation setup with the NIRL system (Figure 1b) allows automated ablation of the focused laser beam over the tissue surface in the defined ROI (Figure 1c). Based on preliminary experiments, it was known that one single laser shot removes a cylindrical tissue fraction with a diameter of about 100 μm and a height of 25 μm on average (Figure 1c). Due to the given ablation scope (Figure 1a–c), we decided to ablate nine consecutive layers of the E14 mouse head. Each layer is covering a theoretical ablation depth of 50 μm (2 × 25 μm), estimating 450 μm depth in total. Each ablated layer consisted of 112 laser pluses distributed over an ablation area of 800 × 800 μm (Figure 1c). For future studies, the ablation depth and area can be dynamically adjusted to the required resolution. Here, we aim to target the cerebral cortex, avoiding the midline region. Prior to ablation, we marked the midline and the left eye to standardize the ablation region and placed the ROI in equal proximity to the marked region (Figure 1d).

In order to determine the actual depth of the ablated volume, we scanned the surface of the specimens with optical coherence tomography (OCT) before and after the ablation. The results show that the actual average ablation depth of all nine layers in total reached about 370 μm and a volume of 240 nL (n = 3) (Figure 1f). The deviation from the initially estimated 50 μm/layer to the actual coverage of 41.1 μm/layer may occur due to decreasing laser energy over ablation depth and the sample being out of laser focus. As further refinement for the ablation system, an automated adjustment of the laser settings after each layer ablation could be implemented. To verify the ablation region and depth using histology, we prepared fresh frozen sections of the specimen after the ablation. Respective frontal sections were stained via H&E and showed that the nine ablated layers cover most of the cortex area but were insufficient to reach the border to the ventricle (V) (Figure 1e). Overall, we covered a large part of the cortex with a high spatial resolution for further differential quantitative proteomics.

Sample Processing

The principle of NIRL-based sampling and homogenization is the cold evaporation of tissue samples, where the biomolecules are released within an aerosol. For targeted aerosol collection, we used glass slides with 12 wells predefined by PTFE coating. Each well was manually placed above the ROI on the embryonic mouse head to collect the aerosol of the single layers. Based on microscope images, we reassured that the collected aerosol is evenly distributed and well homogenized. Only a few larger chunks were observed, most likely formed during the drying process of the tissue aerosol (Supplementary Figure 1). For subsequent sample preparation, we adapted to a protocol using n-dodecyl β-D-maltoside (DDM) as a nonionic detergent. DDM is compatible with mass spectrometry and brings membrane proteins into solution particularly well.29 After the resuspension of the condensed sample aerosol with DDM, all sample preparation steps were performed in a single low protein binding tube, thus avoiding unnecessary transfer steps that could potentially lead to adsorption effects. These effects also occur during resuspension of the sample. Using predefined wells for targeted aerosol collection, the wetted area was limited during resuspension by the hydrophobic properties of PTFE coating. This additionally minimized the loss of sample due to adsorption effects.

Altogether, we were able to reduce the ablation volume by a factor of 20, from the prior 60016 to 30 nL per layer, and still identify 3126 proteins with quantitative information. In terms of spatial resolution, this means that we effectively enhanced the z-resolution from 11716 to 41 μm per layer. The potential to further increase the resolution to the laser beam dimensions (Figure 1c) depends on further sample processing improvements and measurements with highly sensitive mass spectrometers to accomplish resolutions like conventional methods for spatial proteomics such as MALDI–MSI7 or laser capture microdissection (LCM).3,12,13 Both methods MALDI–MSI and LCM contain several sample preparation steps like tissue sectioning, fixation, matrix application, and staining, risking degradation processes resulting in altered protein composition.14 In contrast, the simple application of NIRL requires neither prior tissue sectioning nor complex sample preparation. Instead, concurred homogenization allows fast sampling, ensuring that biomolecules remain close to their native form (Figure 1b). Moreover, direct 3D sampling offers more flexibility to orientate the sample and reach respective regions of interest.

Data Preprocessing and Batch Effect Reduction

In our study, we ablated five hemispheres from four mouse heads (FB1–4), the right hemisphere from FB1–3 as biological replicates, and the right and left hemispheres of FB4 as technical replicates (n = 5 in total). The ablation of each animal was performed on four different days, and each day was labeled as a laser ablation batch (LA batches). Altogether, 45 samples were acquired (nine layers per hemisphere and five hemispheres in total). These samples were measured in five different batches by mass spectrometry (m batches), and we identified in total 5031 proteins including 3126 proteins with quantitative information. These batches can introduce technical bias, the so-called batch effects, which we want to investigate and reduce. Therefore, the LA batch and m batch were combined and assigned into six pseudo batches (p batches, Supplementary Table 1). To examine whether a batch effect is present within the six pseudo batches, nonlinear Iterative Partial Least Squares principal component analysis (NIPALS-PCA) was performed using relative protein abundances of 88 detected housekeeping proteins.33 Before applying the HarmonizR framework30 for batch effect reduction, scatterplot visualization of NIPALS-PCA results in strong pseudo batch effects. Whereas after batch effect reduction, no separation is apparent for the pseudo batches (Figure 2a). Additionally, based on housekeeping proteins ACTB, GAPDH, and TUBB3, it was shown that relative protein abundances were adjusted to similar mean values across the pseudo batches after HarmonizR30 (Supplementary Figure 2). For further validation of the impact of batch effects regarding the entire protein data set, hierarchical clustering was performed with 70% valid values of both data sets, before and after HarmonizR. Without batch effect reduction, the samples clustered strongly according to p batches associated with the replicates. For the harmonized data, no clustering was observed for p batches, whereas layerwise distribution was enhanced (Figure 2b). All together, we showed, by combining technical and biological replicates and reducing technical batch effects, that the laser ablation method is reproducible and applicable for analyzing layerwise signatures of the samples.

Figure 2.

Figure 2

Open in a new tab

Batch effect reduction by integration into pseudo batches. (a) Nonlinear Iterative Partial Least Squares principal component analysis (NIPALS-PCA) based on 88 housekeeping proteins.33 (b) Hierarchical clustering with 70% valid values before (1369 proteins) and after (1350 proteins) HarmonizR.30

Layer Proteome Signature

After applying the HarmonizR framework30 for batch effect reduction, we maintained 3040 proteins with quantitative information. We observed a higher number of detected proteins in deeper layers, suggesting a positive correlation of cell number and protein density (Supplementary Figure 3). The harmonized data (Supplementary Table 2) was used to study the proteome signature of the ablated layers. We used two different approaches for dimension reduction to analyze whether the samples can be differentiated based on the layers. First, the NIPALS-PCA allows calculations with missing values. Therefore, we used proteins with 70% valid values along the 45 samples, resulting in 1350 proteins. The scatter-plot visualization of PCA shows a gradient-wise distribution from superficial to deeper layers in the first two components, as expected (Figure 3a). This observation can be explained by the layer composition of different cell types and factors. To confirm this observation, we additionally used UMAP (Uniform Manifold Approximation and Projection) and proteins with 100% valid values (570 proteins) and observed a similar outcome (Figure 3b).

Figure 3.

Figure 3

Open in a new tab

Dimension reduction analysis to resolve sample distribution after batch effect correction. (a) Scatter-plot visualization of the first two principal components (PC) of the NIPALS-PCA with 70% valid values (1350 proteins). (b) UMAP for dimension reduction with 100% valid values (570 proteins). Annotation color is based on layers for (a) and (b). (c) Top 100 proteins with 70% valid values were selected based on ANOVA between ablated layers (p < 0.05) and demonstrated in a heatmap with Euclidean distance-based hierarchical clustering.

To show which proteins drive this distribution along the layers, we selected the top 100 proteins based on ANOVA testing between the layers (70% valid values, p < 0.05) and visualized the abundances in a heatmap (Figure 3c, Supplementary Table 3). These results indicated distinct proteome signatures in each layer with the formation of two superordinate protein signatures involving layers 1–4 and layers 5–9 wherein gradient-wise abundance can be observed (Figure 3c).

Next, we compared the proteins from each layer with gene sets of anatomical structures (Figure 4a, Supplementary Table 4). The mean abundance for all given proteins (Supplementary Table 5) within a gene set showed that layers 1–3 have a signature highly associated with skin and bone structures. This shows the potential to further investigate multilayered organs like the skin. Proteome analysis of the layered skin structure will improve the understanding of wound healing processes34 and diverse skin conditions such as dermatitis35 or psoriasis.36 Layer 4 displayed a high abundance of proteins associated with the meninges (Figure 4b). Usually, during brain dissection, the two outer layers dura and arachnoidea mater are detached from the brain and impede investigating the meningeal system.37 With our layer ablation method, we captured the meninges within a layer. This provides the possibility to investigate altered conditions of the blood-brain barrier in different neurological diseases.38 Moreover, global proteomic analysis enables to find the potential target proteins for drug delivery and treatment of cancer and other diseases.34,39 Layer 5 showed a high abundance of proteins in the marginal zone. The abundance profiles of layers 6–9 corresponded with the signature from the cortical plate, including the cortical-preplate and -subplate, the subventricular zone and ventricular zone (Figure 4b). We showed that, despite the relatively simple histological composition at E14.5, we were able to show spatial differences on the proteomic level and relate the layers to anatomical structures.

Figure 4.

Figure 4

Open in a new tab

Analysis of layer-specific signature and marker proteins. (a) Schematic illustration of the cortex structure after the second trimester of embryonic development. RG = radial glial cells. (b) Gene set enrichment analysis of cortex structures applied to ablated layers. The layerwise calculated mean abundance of present proteins for each gene set is plotted. (c) IHC staining of mouse E14.5 cortex with known marker proteins and the corresponding protein abundance for each layer. Arrows indicate the corresponding layer with a high relative protein abundance. Scale bar: 100 μm. (d) Variant protein abundances are compared to RNA in situ hybridization (ISH) images of mouse cortex at E13.5 obtained from http://www.developingmouse.brain-map.org.31 Arrows indicate the corresponding layer with high relative protein abundance. Scale bar: 110 μm.

To further validate our data, we compared layerwise protein abundances with immunohistochemical (IHC) staining of the E14.5 mouse cortex. We investigated proteins that are commonly used as laminar markers (Figure 4c). The brain tissue signature in layers 5–9 is supported by the high MAP2 (microtubuli associated protein)40 protein abundance and the positive IHC staining of this dendritic marker in layers 4–9. NESTIN, a radial glia (RG) marker, showed increasing abundance toward the ventricular zone (VZ) where the somata of apical RG are located, indicating that we attained parts of the VZ.25,41 KI67 is an established marker for proliferating cells, which predominantly reside in the ventricular zone.42 Moreover, the progenitor cell marker SOX2 is also expected to be present in the VZ.41,43,44 Both proteins were indeed highly abundant in layer 4 and in deeper layers 8 and 9. This indicates high proliferation activity of RGs in the VZ,25 a region where tumors potentially develop.27 Further analysis could reveal molecular environmental factors that may facilitate tumor formation.

In order to find potential new protein markers for certain regions, we analyzed highly variant proteins among the ablated layers and compared the candidates (DLG4, APOE, NMRAL1, and MTA1) to in situ hybridization (ISH) images of mouse cortex at E13.5 (http://www.developingmouse.brain-map.org.31). We showed that both proteins, DLG4 and APOE, are associated with superficial layers on the proteomic level, which could be validated by the ISH images (Figure 4d). The protein DLG4 (Disks large homolog 4) also known as PSD95 is a postsynaptic scaffold protein regulating the excitability of glutamatergic synapsis. Deficiency of this protein was shown to be related to psychiatric disorders.45,46 APOE (Apolipoprotein E) is a protein expressed in the brain and is associated with astrocytic cells, which are in close proximity to the pial surface.47−49 NMRAL1 and MTA1 were shown to be more abundant in layers 5–9 (Figure 4d). Whereas these proteins have not been reported in the context of cortical development yet, MTA1 (Metastasis-associated protein) is known to be expressed in different cancer types and to regulate the stability of the oncosuppressor p53.50−52 Additionally, NMRAL1 (NmrA-like family domain-containing protein 1) is a largely unknown protein, and only a NADPH-binding function is assumed.53

The results of our study confirm the potential to spatially resolve the proteomic signatures of embryonic mouse head structures and forebrain regions using NIRL-based 3D sampling and subsequent bottom-up proteomics. Direct ablation from a 3-dimensional tissue sample like the mouse head with an axial resolution of about 40 μm paves the way for numerous research questions focusing on spatial resolution in proteomics with the additional possibility to achieve even higher resolution. The high spatial resolution with 3D NIRL-based sampling for bottom-up proteomics allows future incorporation of multiomics modalities like lipidomics54 or transcriptomics. The translation from the genomic to protein level is crucial to understand the functional and organizational complexity during development, disease pathogenesis, or malformations and enable a better translation from disease models to the clinics.39,55,56

Conclusions

In this work, we used NIRL-based 3D sampling to directly ablate consecutive tissue layers from the embryonic mouse head to the forebrain. By optimizing sample processing for subsequent quantitative, bottom-up proteomics of very small ablation volumes (<30 nL), we were able to significantly increase the spatial resolution in depth (∼40 μm). Further adaptation will allow to sample and analyze smaller volumes with higher spatial resolution. Our findings highlight the potential of NIRL-based sampling to spatially resolve three-dimensional tissue heterogeneities without the need for prior elaborate sample preparation steps. Thus, it is possible to specifically sample regions of interest like the scalp or meninges of the embryonic mouse head to address related research questions. Common approaches to dissociate structural layered tissue samples using proteases (e.g., dispase to separate the epidermis and dermis) will no longer be necessary.

Acknowledgments

The authors would like to thank Shweta Godbole, Matthias Dottermusch and Tasja Lempertz for their great support in working with mice, Antonia Gocke for helping perform sample preparation and Anton Walter for the support in OCT analysis. Further, the authors would like to thank the Institute of Neuropathology and Kristin Hartmann from the Mouse Pathology facility for the technical assistance and service. The authors thank Ursula Müller, Jasmin Seydler, Alexandra Gröss, Anke Dorendorf, Beate Miksche, and the other members of the animal facility team for their support and service. The authors used images and templates for parts of Figures 1 and 4 from http://smart.servier.com, licensed under a Creative Common Attribution 3.0 Generic License.

Data Availability Statement

Data is available via ProteomeXchange with the identifier PXD043011.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.3c02637.

  • Experimental section; visualization of condensed aerosol (Figure S1); batch effect reduction by integration of pseudo batches using the HarmonizR framework (Figure S2); cell density of extracranial and cerebral cortex structures compared to the number of detected proteins with valid values per layer (Figure S3) (PDF)

  • Sample information and unprocessed quantitative data (Table S1) (XLSX)

  • Harmonized data (Table S2) (XLSX)

  • ANOVA significant proteins (Table S3) (XLSX)

  • Gene list input (Table S4) (XLSX)

  • Protein list (Table S5) (XLSX)

Author Contributions

⊥ J.N. and M.M. contributed equally to this work and share the first authorship. J.E.N. and J.H. contributed equally to this work and share senior authorship. The manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript.

This research was funded by the “Behörde für Wissenschaft, Forschung, Gleichstellung und Bezirke der Freien und Hansestadt Hamburg” (Public Authority for Science, Research and Equality of the Free and Hanseatic City of Hamburg; grant number LFF-FV-75 and LFF-GK-10). This study was supported by grants from the Deutsche Forschungsgemeinschaft (DFG) (INST 337/15–1, INST 337/16–1, INST 152/837–1, INST 152/947–1 FUGG, and SCHL 406/21–1). Furthermore, this work was supported by intramural funding from the faculty. J.E.N. is supported in the Emmy Noether program (DFG project number 416054672).

The authors declare no competing financial interest.

Supplementary Material

ac3c02637_si_004.pdf (440.2KB, pdf)
ac3c02637_si_005.xlsx (10.5MB, xlsx)
ac3c02637_si_006.xlsx (1.3MB, xlsx)
ac3c02637_si_007.xlsx (10.6KB, xlsx)
ac3c02637_si_008.xlsx (56.7KB, xlsx)
ac3c02637_si_009.xlsx (29.2KB, xlsx)

References

  1. Maes E.; Cools N.; Willems H.; Baggerman G. FACS-based proteomics enables profiling of proteins in rare cell populations. Int. J. Mol. Sci. 2020, 21, 6557. 10.3390/ijms21186557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Dyring-Andersen B.; Løvendorf M. B.; Coscia F.; et al. Spatially and cell-type resolved quantitative proteomic atlas of healthy human skin. Nat. Commun. 2020, 11, 5587 10.1038/s41467-020-19383-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Mao Y.; Wang X.; Huang P.; Tian R. Spatial proteomics for understanding the tissue microenvironment. Analyst 2021, 146, 3777–3798. 10.1039/D1AN00472G. [DOI] [PubMed] [Google Scholar]
  4. Sharma K.; Schmitt S.; Bergner C. G.; et al. Cell type– and brain region–resolved mouse brain proteome. Nat. Neurosci. 2015, 18, 1819–1831. 10.1038/nn.4160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Giansanti P.; Samaras P.; Bian Y.; et al. Mass spectrometry-based draft of the mouse proteome. Nat. Methods 2022, 19, 803–811. 10.1038/s41592-022-01526-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Schwamborn K.; Caprioli R. M. MALDI Imaging Mass Spectrometry - Painting Molecular Pictures. Mol. Oncol. 2010, 4, 529–538. 10.1016/j.molonc.2010.09.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Zhu X.; Xu T.; Peng C.; Wu S. Advances in MALDI Mass Spectrometry Imaging Single Cell and Tissues. Front. Chem. 2022, 9, 782432. 10.3389/fchem.2021.782432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Ščupáková K.; Balluff B.; Tressler C.; et al. Cellular resolution in clinical MALDI mass spectrometry imaging: the latest advancements and current challenges. Clin. Chem. Lab. Med. 2020, 58, 914–929. 10.1515/cclm-2019-0858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Thiele H.; Heldmann S.; Trede D.; et al. 2D and 3D MALDI-imaging: Conceptual strategies for visualization and data mining. Biochim. Biophys. Acta - Proteins Proteomics 2014, 1844, 117–137. 10.1016/j.bbapap.2013.01.040. [DOI] [PubMed] [Google Scholar]
  10. Dilillo M.; Pellegrini D.; Ait-Belkacem R.; et al. Mass Spectrometry Imaging, Laser Capture Microdissection, and LC-MS/MS of the Same Tissue Section. J. Proteome Res. 2017, 16, 2993–3001. 10.1021/acs.jproteome.7b00284. [DOI] [PubMed] [Google Scholar]
  11. Lin M.; Eberlin L. S.; Seeley E. H. Reduced Hemoglobin Signal and Improved Detection of Endogenous Proteins in Blood-Rich Tissues for MALDI Mass Spectrometry Imaging. J. Am. Soc. Mass Spectrom. 2022, 33, 296–303. 10.1021/jasms.1c00300. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Xu R.; Tang J.; Deng Q.; et al. Spatial-Resolution Cell Type Proteome Profiling of Cancer Tissue by Fully Integrated Proteomics Technology. Anal. Chem. 2018, 90, 5879–5886. 10.1021/acs.analchem.8b00596. [DOI] [PubMed] [Google Scholar]
  13. Davis S.; Scott C.; Ansorge O.; Fischer R. Development of a Sensitive, Scalable Method for Spatial, Cell-Type-Resolved Proteomics of the Human Brain. J. Proteome Res. 2019, 18, 1787–1795. 10.1021/acs.jproteome.8b00981. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kwiatkowski M.; Wurlitzer M.; Omidi M.; et al. Ultrafast extraction of proteins from tissues using desorption by impulsive vibrational excitation. Angew. Chem. - Int. Ed. 2015, 54, 285–288. 10.1002/anie.201407669. [DOI] [PubMed] [Google Scholar]
  15. Hahn J.; Moritz M.; Voß H.; et al. Tissue sampling and homogenization in the sub-microliter scale with a nanosecond infrared laser (Nirl) for mass spectrometric proteomics. Int. J. Mol. Sci. 2021, 22, 10833. 10.3390/ijms221910833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Voß H.; et al. Tissue Sampling and Homogenization with NIRL Enables Spatially Resolved Cell Layer Specific Proteomic Analysis of the Murine Intestine. Int. J. Mol. Sci. 2022, 23, 6132. 10.3390/ijms23116132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dong C.; Donnarumma F.; Murray K. K. Infrared Laser Ablation Microsampling for Small Volume Proteomics. J. Am. Soc. Mass Spectrom. 2022, 33, 1003–1010. 10.1021/jasms.2c00063. [DOI] [PubMed] [Google Scholar]
  18. Dong C.; Richardson L. T.; Solouki T.; Murray K. K. Infrared Laser Ablation Microsampling with a Reflective Objective. J. Am. Soc. Mass Spectrom. 2022, 33, 463–470. 10.1021/jasms.1c00306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Franjic K.; Miller R. J. D. Vibrationally excited ultrafast thermodynamic phase transitions at the water/air interface. Phys. Chem. Chem. Phys. 2010, 12, 5225–5239. 10.1039/c000359j. [DOI] [PubMed] [Google Scholar]
  20. Amini-Nik S.; Kraemer D.; Cowan M. L.; et al. Ultrafast mid-IR laser scalpel: Protein signals of the fundamental limits to minimally invasive surgery. PLoS One 2010, 5, e13053. 10.1371/journal.pone.0013053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Sampson J. S.; Murray K. K.; Muddiman D. C. Intact and Top-Down Characterization of Biomolecules and Direct Analysis Using Infrared Matrix-Assisted Laser Desorption Electrospray Ionization Coupled to FT-ICR Mass Spectrometry. J. Am. Soc. Mass Spectrom. 2009, 20, 667–673. 10.1016/j.jasms.2008.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kwiatkowski M.; Wurlitzer M.; Krutilin A.; et al. Homogenization of tissues via picosecond-infrared laser (PIRL) ablation: Giving a closer view on the in-vivo composition of protein species as compared to mechanical homogenization. J. Proteomics 2016, 134, 193–202. 10.1016/j.jprot.2015.12.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Böttcher A.; Kucher S.; Knecht R.; et al. Reduction of thermocoagulative injury via use of a picosecond infrared laser (PIRL) in laryngeal tissues. Eur. Arch. Oto-Rhino-Laryngol. 2015, 272, 941–948. 10.1007/s00405-015-3501-4. [DOI] [PubMed] [Google Scholar]
  24. Nadarajah B.; Parnavelas J. G. Modes of neuronal migration in the developing cerebral cortex. Nat. Rev. Neurosci. 2002, 3, 423–432. 10.1038/nrn845. [DOI] [PubMed] [Google Scholar]
  25. Florio M.; Huttner W. B. Neural progenitors, neurogenesis and the evolution of the neocortex. Development 2014, 141, 2182–2194. 10.1242/dev.090571. [DOI] [PubMed] [Google Scholar]
  26. Ogawa M.; Miyata T.; Nakajimat K.; et al. The reeler gene-associated antigen on cajal-retzius neurons is a crucial molecule for laminar organization of cortical neurons. Neuron 1995, 14, 899–912. 10.1016/0896-6273(95)90329-1. [DOI] [PubMed] [Google Scholar]
  27. Copp A. J.; Harding B. N. Neuronal migration disorders in humans and in mouse models - An overview. Epilepsy Res. 1999, 36, 133–141. 10.1016/S0920-1211(99)00047-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hevner R. F. Layer-specific markers as probes for neuron type identity in human neocortex and malformations of cortical development. J. Neuropathol. Exp. Neurol. 2007, 66, 101–109. 10.1097/nen.0b013e3180301c06. [DOI] [PubMed] [Google Scholar]
  29. Tsai C. F.; Zhang P.; Scholten D.; et al. Surfactant-assisted one-pot sample preparation for label-free single-cell proteomics. Commun. Biol. 2021, 4, 265. 10.1038/s42003-021-01797-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Voß H.; et al. HarmonizR enables data harmonization across independent proteomic datasets with appropriate handling of missing values. Nat. Commun. 2022, 13, 3523 10.1038/s41467-022-31007-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Allen Reference Atlas – Developing Mouse Brain. http://developingmouse.brain-map.org/.
  32. Schneider C. A.; Rasband W. S.; Eliceiri K. W. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 2012, 9, 671–675. 10.1038/nmeth.2089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Uhlén M.; Fagerberg L.; Hallström B. M.; et al. Tissue-based map of the human proteome. Science 2015, 347, 1260419. 10.1126/science.1260419. [DOI] [PubMed] [Google Scholar]
  34. Zomer H. D.; Trentin A. G. Skin wound healing in humans and mice: Challenges in translational research. J. Dermatol. Sci. 2018, 90, 3–12. 10.1016/j.jdermsci.2017.12.009. [DOI] [PubMed] [Google Scholar]
  35. Rizzo J. M.; Oyelakin A.; Min S.; et al. ΔNp63 regulates IL-33 and IL-31 signaling in atopic dermatitis. Cell Death Differ. 2016, 23, 1073–1085. 10.1038/cdd.2015.162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Schön M. P.; Manzke V.; Erpenbeck L. Animal models of psoriasis—highlights and drawbacks. J. Allergy Clin. Immunol. 2021, 147, 439–455. 10.1016/j.jaci.2020.04.034. [DOI] [PubMed] [Google Scholar]
  37. Louveau A.; Filiano A. J.; Kipnis J. Meningeal Whole Mount Preparation and Characterization of Neural Cells by Flow Cytometry. Curr. Protoc. Immunol. 2018, 121, 678–687. 10.1002/cpim.50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Munji R. N.; Soung A. L.; Weiner G. A.; et al. Profiling the mouse brain endothelial transcriptome in health and disease models reveals a core blood–brain barrier dysfunction module. Nat. Neurosci. 2019, 22, 1892–1902. 10.1038/s41593-019-0497-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Ha J.; Park H.; Park J.; Park S. B. Recent advances in identifying protein targets in drug discovery. Cell Chem. Biol. 2021, 28, 394–423. 10.1016/j.chembiol.2020.12.001. [DOI] [PubMed] [Google Scholar]
  40. Caceres A.; Binder L.; Payne M.; et al. Differential subcellular localization of tubulin and the microtubule-associated protein MAP2 in brain tissue as revealed by immunocytochemistry with monoclonal hybridoma antibodies. J. Neurosci. 1984, 4, 394–410. 10.1523/JNEUROSCI.04-02-00394.1984. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Penisson M.; Ladewig J.; Belvindrah R.; Francis F. Genes and Mechanisms Involved in the Generation and Amplification of Basal Radial Glial Cells. Front. Cell. Neurosci. 2019, 13, 1–21. 10.3389/fncel.2019.00381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Li L. T.; Jiang G.; Chen Q.; Zheng J. N. Predic Ki67 is a promising molecular target in the diagnosis of cancer (Review). Mol. Med. Rep. 2015, 11, 1566–1572. 10.3892/mmr.2014.2914. [DOI] [PubMed] [Google Scholar]
  43. Miyagi S.; Nishimoto M.; Saito T.; et al. The Sox2 regulatory region 2 functions as a neural stem cell-specific enhancer in the telencephalon. J. Biol. Chem. 2006, 281, 13374–13381. 10.1074/jbc.M512669200. [DOI] [PubMed] [Google Scholar]
  44. Amador-Arjona A.; Cimadamore F.; Huang C. T.; et al. SOX2 primes the epigenetic landscape in neural precursors enabling proper gene activation during hippocampal neurogenesis. Proc. Natl. Acad. Sci. U.S.A. 2015, 112, E1936–E1945. 10.1073/pnas.1421480112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Keith D.; El-Husseini A. Excitation control: Balancing PSD-95 function at the synapse. Front. Mol. Neurosci. 2008, 1, 200. 10.3389/neuro.02.004.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Coley A. A.; Gao W. J. PSD95: A synaptic protein implicated in schizophrenia or autism?. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 2018, 82, 187–194. 10.1016/j.pnpbp.2017.11.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Abnet K.; Fawcett J. W.; Dunnett S. B. Interactions between meningeal cells and astrocytes in vivo and in vitro. Dev. Brain Res. 1991, 59, 187–196. 10.1016/0165-3806(91)90099-5. [DOI] [PubMed] [Google Scholar]
  48. Sun Y.; Wu S.; Bu G.; et al. Glial Fibrillary Acidic Protein–Apolipoprotein E (apoE) Transgenic Mice: Astrocyte-Specific Expression and Differing Biological Effects of Astrocyte-Secreted apoE3 and apoE4 Lipoproteins. J. Neurosci. 1998, 18, 3261–3272. 10.1523/JNEUROSCI.18-09-03261.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Pfrieger F. W. Cholesterol homeostasis and function in neurons of the central nervous system. Cell. Mol. Life Sci. 2003, 60, 1158–1171. 10.1007/s00018-003-3018-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Gururaj A. E.; Singh R. R.; Rayala S. K.; et al. MTA1, a transcriptional activator of breast cancer amplified sequence 3. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 6670–6675. 10.1073/pnas.0601989103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Li D. Q.; Divijendra Natha Reddy S.; Pakala S. B.; et al. MTA1 coregulator regulates p53 stability and function. J. Biol. Chem. 2009, 284, 34545–34552. 10.1074/jbc.M109.056499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Liu J.; Xu D.; Wang H.; et al. The subcellular distribution and function of MTA1 in cancer differentiation. Oncotarget 2014, 5, 5153–5164. 10.18632/oncotarget.2095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Zang W.; Zheng X. Structure and functions of cellular redox sensor HSCARG/NMRAL1, a linkage among redox status, innate immunity, DNA damage response, and cancer. Free Radic. Biol. Med. 2020, 160, 768–774. 10.1016/j.freeradbiomed.2020.09.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Stadlhofer R.; et al. Lipidome Analysis of Oropharyngeal Tumor Tissues Using Nanosecond Infrared Laser (NIRL) Tissue Sampling and Subsequent Mass Spectrometry. Int. J. Mol. Sci. 2023, 24, 7820. 10.3390/ijms24097820. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Pradet-Balade B.; Boulmé F.; Beug H.; Müllner E. W.; Garcia-Sanz J. A. Translation control: Bridging the gap between genomics and proteomics?. Trends Biochem. Sci. 2001, 26, 225–229. 10.1016/S0968-0004(00)01776-X. [DOI] [PubMed] [Google Scholar]
  56. Wang T.; Cui Y.; Jin J.; et al. Translating mRNAs strongly correlate to proteins in a multivariate manner and their translation ratios are phenotype specific. Nucleic Acids Res. 2013, 41, 4743–4754. 10.1093/nar/gkt178. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ac3c02637_si_004.pdf (440.2KB, pdf)
ac3c02637_si_005.xlsx (10.5MB, xlsx)
ac3c02637_si_006.xlsx (1.3MB, xlsx)
ac3c02637_si_007.xlsx (10.6KB, xlsx)
ac3c02637_si_008.xlsx (56.7KB, xlsx)
ac3c02637_si_009.xlsx (29.2KB, xlsx)

Data Availability Statement

Data is available via ProteomeXchange with the identifier PXD043011.

Articles from Analytical Chemistry are provided here courtesy of American Chemical Society

RESOURCES