Skip Navigation


Bioscience Horizons Advance Access originally published online on April 30, 2009
Bioscience Horizons 2009 2(2):134-146; doi:10.1093/biohorizons/hzp016
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrowOA All Versions of this Article:
2/2/134    most recent
hzp016v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Google Scholar
Right arrow Articles by Gregson, C.
PubMed
Right arrow Articles by Gregson, C.
Related Collections
Right arrow Biochemistry
Right arrow Cell Biology
Right arrow Medical Science
Right arrow Pathology
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© 2009 The Author(s)
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Optimization of MALDI tissue imaging and correlation with immunohistochemistry in rat kidney sections

Cassie Gregson*

Faculty of Sciences, Staffordshire University, Staffordshire, UK

* Corresponding author: 21 Williamson Road, Whaley Bridge, Derbyshire SK23 7AW, UK. Tel: +44 1663 733029. Email: cassie.e{at}gmail.com

Project Supervisor: Dr Sue Bird, Faculty of Sciences, Staffordshire University, Staffordshire, UK.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results and Discussion
 Conclusion
 References
 Author Biography 
 
Cancer is responsible for approximately 6.7 million deaths and 10.9 million newly diagnosed cases worldwide per year. Currently, the definitive diagnosis of a tissue or cell sample is determined by molecular pathology and histochemical techniques, which assess tumour type, grade and stage. This information also allows for the diagnosis, prognosis and available treatment options to be established. In addition, immunohistochemistry (IHC) in combination with histochemistry is used to determine the surgical margin status of tumours, which can be correlated to the likelihood of recurrence. Matrix-assisted laser desorption/ionization (MALDI) imaging is a mass spectrometry profiling technique, which can be used to simultaneously identify multiple species within a tissue section. The array format of the acquisition allows for the creation of an image that is viewed in a similar way to an IHC section. MALDI imaging could potentially provide an alternative diagnostic assay that could be used to provide cancer prognosis. To assess the suitability of MALDI imaging for this application, sample preparation and MALDI imaging methodology were developed using {alpha}2u-globulin as an example protein. This protein is known to be preferably expressed in the kidneys of male rats allowing a proof-of-principle study to be performed comparing the expression levels and localization between male and female rat kidney sections. The expression of {alpha}2u-globulin was localized to the cortex region of the kidney, with the levels of {alpha}2u-globulin shown to be significantly higher in the male than the female kidney sections. These findings were validated by comparison with IHC data. The proof-of-principle study therefore demonstrated that MALDI imaging could be a potential alternative to current molecular pathology and histochemical techniques for the determination of tumour type, grade and stage as well as the determination of surgical margin status.

Key words: MALDI, MALDI imaging, mass spectrometry


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results and Discussion
 Conclusion
 References
 Author Biography 
 
Cancer is responsible for approximately 6.7 million deaths and 10.9 million cases are newly diagnosed worldwide per year.1 Current limitations in molecular pathology and histochemical techniques may be adversely affecting the ability to definitively diagnose tumours to enable successful treatment of patients, therefore leading to a worse prognosis. An alternative to these current diagnostic techniques may be matrix-assisted laser desorption/ionization (MALDI) imaging, a new approach requiring successful proof-of-principle data to determine its applicability.

Molecular Pathology
Molecular pathology is the study of the molecular-level mechanisms involved in the pathogenesis of disease, including genes whose transcription is altered in tumour cells and the expression levels of the resultant proteins. The use of techniques that can identify these nucleic acids or proteins can therefore be used to help identify known cancer biomarkers and to determine tumour stage, which describes whether the tumour cells are invasive and/or aggressive.2 These techniques can also be used to detect molecular markers that can be used identify secondary tumours (metastases) of unknown primary origin.3 In addition, molecular methods can be used to identify the grade of the tumour, which determines the degree of resemblance of the tumour cells to those surrounding cells that are normal.4 Another important application of molecular pathology is the determination of residual cancer in surgical margins in patients who have undergone surgery to remove the tumour. A generalized workflow used in molecular pathology is shown in Fig. 1 and the molecular pathology techniques most commonly used are summarized in Table 1.


Figure 1
View larger version (23K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. Flow diagram of molecular pathology. Flow diagram shows the generalized molecular pathology workflow. FISH, fluorescence in situ hybridization; RT-PCR, real-time polymerase chain reaction.

 


View this table:
[in this window]
[in a new window]

  		Table 1. Summary of commonly used molecular pathology techniques, detailing applicable molecule and sample type and application to oncology

 
Of the techniques identified within Table 1, immunohistochemistry (IHC) has the ability to detect morphological alterations in tissues in addition to identifying specific molecules and their localization within the tissue, making it an extremely valuable technique in molecular pathology. Currently, IHC is the most widely used molecular pathology method for assessing tumour type, grade and stage and surgical margin status.

In IHC, the tumour markers analysed must be known and described in detail to enable the specific antibodies for this technique to be raised against the antigen and IHC is thus limited by the understanding and discovery rate, validation and success of tumour markers within the scientific community. This has prompted researchers to investigate other techniques that can overcome the limitations of IHC analysis, while providing equally valuable data. One of these techniques is mass spectrometry (MS), specifically MALDI5 imaging, developed by Caprioli et al.6

In MALDI imaging tissue sections are coated in matrix, and spectra (profiles) are acquired in an array format. Once all the data are collected, they are imported into specialized imaging analysis software, where ion intensity is plotted as a function of x and y co-ordinates to create two-dimensional ion maps (images). Each image created is representative of the relative abundance of a particular mass-to-charge ratio (m/z) value across the tissue, meaning that multiple images can be created from one MALDI imaging experiment. In this respect MALDI imaging provides information regarding localization of multiple molecules within a tissue. For MALDI imaging no prior knowledge regarding the molecular components of the tissue need to be known as MALDI-MS analysis does not require any molecularly specific exogenous compounds. In addition, hundreds of molecules can be visualized per imaging experiment. Coupled with a significantly decreased sample preparation time and potential for novel tumour marker identification, MALDI imaging may provide an alternative to IHC analysis, potentially allowing the current IHC limitations to be overcome.

A limitation of MALDI imaging currently is that the technique is still in its infancy and hence there is no well-defined MALDI imaging sample preparation or MALDI imaging method. The generalized MALDI imaging method is shown in Fig. 2.


Figure 2
View larger version (19K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2. MALDI imaging workflow. Generalized MALDI imaging method.

 
This study aims to successfully establish a sample preparation and MALDI imaging method for producing high-quality images of rat kidney and determine the suitability of MALDI imaging as a molecular pathology technique.

Rat {alpha}2u-globulin will be used as a model protein in the proof-of-principle study to establish the potential of MALDI in cancer molecular pathology. This protein is a male-specific protein localized in the proximal tubules of the rat kidney,7–9 and can thus be used to determine the ability of MALDI imaging to detect changes in expression levels of a protein (i.e. a tumour marker for diagnosis of tumour type, grade and stage) and localization of a protein (i.e. to determine surgical margin status of a tissue).

This study aims to successfully establish a sample preparation and MALDI imaging method for producing high-quality images of rat kidney, and to conduct a proof-of-principle study to establish the suitability of MALDI imaging for identifying tumour type, grade, stage and surgical margin status. The validity of the technique will be assessed by determining whether a difference in {alpha}2u-globulin expression levels can be visualized in MALDI images of male and female rat kidney sections (using the developed methodology) and whether the images produced are comparable to those seen with IHC analysis.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results and Discussion
 Conclusion
 References
 Author Biography 
 
Reagents
Acetonitrile (ACN) (A/0627/17), ethanol (EtOH) (E/0600/17), methanol (MeOH) (E/0600/17), trifluoroacetic acid (TFA) (J/8800/PB08), Tris (14050-0010), Tween 20 (BPE337-100), sodium chloride (BPE358-1) and hydrochloric acid (H/1000/PB17) were purchased from Fischer Scientific (Leicestershire, UK).

3,5-Dimethoxy-4-hydroxycinnamic acid (sinapinic acid) (49508) and {alpha}-cyano-4-hydroxycinnamic acid (alpha-cyano) (39468) were purchased from Sigma-Aldrich (St Louis, MO, USA).

Immunohistochemistry
Two tissue slices were cryosectioned from a fresh-frozen male rat kidney at 5 µm and two tissue slices were cryosectioned from a fresh-frozen female rat kidney at 5 µm using a Leica CM1900 cryostat (Leica Microsystems, Wetzlar, Germany), thaw-mounted onto a glass microscope slide and fixed in 100% acetone for 30 s. The sections were circled with a PAP pen (ab2601, Abcam, Cambridge, MA, USA) and endogenous peroxidase activity was blocked using peroxidase block (K4007, Dako, Glostrup, Denmark) for 10 min. The sections were then rinsed in distilled water and immersed in 50 mM Tris-buffered saline (3 M sodium chloride) with 0.05% Tween (TBST) for 5 min. A volume of 100 µl of a 1:20 dilution (in TBST) of normal rabbit serum (X0902, Dako, Glostrup, Denmark) was applied to each slide for 20 min. A volume of 100 µl of a 1:50 dilution (in TBST) of monoclonal anti-{alpha}2u-globulin primary antibody (MAB586 R&D Systems, Minneapolis, MN, USA) was added to one male and one female section, and incubated at room temperature for 60 min; the remaining sections (one male and one female) were omitted for this step to act as controls for this experiment. All sections were then washed twice by immersion in TBST for 5 min. A volume of 100 µl of a 1:50 dilution (in TBST) of rabbit anti-mouse antibody (Z0456, Dako, Glostrup, Denmark) was added to all the sections and incubated at room temperature for 30 min. The sections were then washed twice by immersion in TBST for 5 min. A volume of 100 µl of a 1:50 dilution (in TBST) of mouse peroxidase-anti-peroxidase (P0850, Dako, Glostrup, Denmark) was added to all the sections and incubated at room temperature for 30 min. The sections were then washed twice by immersion in TBST for 5 min. A volume of 200 µl of diaminobenzidine (HK153-5K, Biogenex San Ramon, CA, USA) was applied to each section for 2 min (to enable visualization of peroxidase). Sections were then washed in running tap water and then counter-stained with Carazzi's haematoxylin (H3136, Sigma-Aldrich, St Louis, MO, USA) for 1 min and developed in running tap water. The samples were then dehydrated in 95% EtOH and excess stain was removed by dipping in xylene. Hystomount (M124, TAAB Laboratories Equipment Ltd, Berkshire, UK) was used to mount a coverslip. The tissue section was then visually analysed using a light microscope.

Determination of Matrix Volume for MALDI Imaging of Rat Kidney Sections
Four tissue slices were cryosectioned from a fresh-frozen male rat kidney at 12 µm, using a Leica CM1900 cryostat, and thaw-mounted onto separate indium-tin oxide (ITO)-coated glass microscope slides (VisionTek Systems Ltd, Cheshire, UK). The tissues were then fixed in increasing concentrations of cold EtOH and allowed to fully dry in ambient air after each step. The tissue sections were then immediately coated in matrix solution. All four slides were spray coated with 5 ml of matrix solution (35 mg/ml sinapinic acid in 50:50:0.1 ACN:distilled water:TFA), using a airbrush (CM-C model) with a Sprint Jet compressor (Iwata, Portland, OR, USA), at 20 psi. The apparatus was set up as shown in Fig. 3.


Figure 3
View larger version (11K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 3. Diagrammatic representation of sprayed matrix application using an airbrush. For matrix application, the tissue section mounted on an ITO-coated glass slide is attached to a vertical surface and the airbrush clamped into place with the tip of the airbrush being 15 cm away from the surface of the slide.

 
The matrix solution was applied to the tissue sections by passing the airbrush from left to right over the tissue, while continuously spraying the matrix. The surface of the slide was then allowed to dry fully in ambient air (3 min). This was repeated until 5 ml of matrix had been applied. In addition, one plate had an extra 0.2 µl of matrix solution spotted directly onto the tissue, and a further 10 ml of matrix solution was sprayed onto another plate.

The ITO-coated slides were then attached to separate gold-coated stainless steel MALDI target plates (V700666, Applied Biosystems, Foster City, CA, USA), using Scotch brand double-sided tape (3 M Co., St Paul, MN, USA) and inserted into an Opti-TOF® holder (4347689, Applied Biosystems, Foster City, CA, USA). The plate was then loaded into the Voyager DE-STR MALDI-TOF mass spectrometer (Applied Biosystems, Foster City, CA, USA).

MMIST (v. 2.2.1) software, developed by Stoeckli et al. (1999)10 was used to acquire the imaging data by controlling the Voyager DE-STR. The data were acquired in positive linear MS mode with optimized delayed extraction, using the following parameters: 25 kV accelerating voltage, 95% grid voltage, 750 ns extraction delay time, m/z range 5000–25 000, low mass gate of 500 m/z, 75 laser shots per spectrum and the laser power attenuator set at 2800. In MMIST the area to be imaged, a 60 x 60 spot region (in which the area of interest was contained) with a 200 µm distance between each of the spots (horizontally and vertically), was defined. The calibration was performed externally and updated prior to each of the four analyses, using calibration 3 mixture. After the data had been acquired, they were imported into BioMap imaging software (v. 3.7.5.1), developed by Stoeckli et al. (2002) for image visualization and analysis.

Matrix Solvent Composition Determination
A protein mixture was prepared, using eight proteins over a mass range of 4.3–28.2 kDa at concentrations between 2.9 and 6.3 pmol/µl (Table 2), to test 24 different matrix solvent compositions. The matrix solvent compositions were composed of three different solvents (ACN, EtOH and MeOH), at four different concentrations (50:50, 60:40, 70:30 and 80:20 solvent:distilled water) with (0.1%) or without (0%) TFA; 35 mg/ml sinapinic acid was dissolved in each solvent composition.


View this table:
[in this window]
[in a new window]

  		Table 2. Components of the protein mixture used to determine the matrix solvent composition for MALDI imaging experiments

 
A volume of 0.5 µl of the protein mixture was co-crystallized with 0.5 µl of (one of) the matrix solvent compositions in triplicate, using the dried-droplet technique, onto a stainless steel MALDI target plate. This was repeated for each of the matrix solvent compositions, and alongside each triplicate a calibration 3 mixture (with 35 mg/ml sinapinic acid in 50:50:0.1 ACN:distilled water:TFA) was spotted. The plate was loaded into the Voyager DE-STR and the samples were analysed in positive linear MS mode with 200 laser shots per spectrum over a mass range of 4000–30 000 m/z, and external calibration was performed and updated before each sample. Once all the data were acquired, the spectra were imported into Data Explorer® software (Applied Biosystems, Foster City, CA, USA) and subjected to standard baseline correction and noise removal of 0.7 in order to remove maximal chemical noise without altering sample peaks. The heights for each of the protein peaks in the triplicate spectra for each matrix solvent composition were exported into Excel. Here the mean (n = 3) and standard deviation (SD) were calculated, and used to create a bar graph for each solvent type ( note that the mean and SD were calculated based on the number of peaks identified; where less than three peaks were recorded the calculations were adjusted accordingly to reflect the number of peaks identified).

Determination of Matrix Composition and Optimum Tissue Thickness for MALDI Imaging of Rat Kidney Sections
Six tissue slices were cryosectioned from a fresh-frozen male rat kidney, three slices at 12 µm and three slices at 5 µm, using a P2 Leica CM1900 cryostat. All six sections were thaw-mounted onto separate ITO-coated glass slides. The tissue sections were then fixed in increasing concentrations of cold EtOH. Three matrix solvent compositions were tested: 35 mg/ml sinapinic acid in 50:50:0.1 ACN:distilled water:TFA or 60:40:0.1 EtOH:distilled water:TFA or 80:20:0.1 for MeOH:distilled water:TFA. A volume of 15 ml of each matrix solvent composition was sprayed onto a 12 and 5 µm tissue section. Then the slides were mounted onto gold-plated stainless steel MALDI target plates using double-sided tape. The plates were loaded sequentially into the Voyager DE-STR and analysed using the imaging MS mode. The calibration was performed externally and repeated prior to each of the six analyses, using calibration 3 mixture. After the data had been acquired, they were imported into BioMap for image visualization and analysis.

Proof-of-principle MALDI Imaging of Male and Female Rat Kidney Sections
Five 5 µm tissue slices were cryosectioned from a fresh-frozen male rat kidney and five 5 µm tissue slices were cryosectioned from a fresh-frozen female rat kidney, using a Leica CM1900 cryostat, and thaw-mounted onto separate ITO-coated glass slides. The tissue sections were then fixed in increasing concentrations of cold EtOH. A volume of 15 ml of 35 mg/ml sinapinic acid in 50:50:0.1 ACN:distilled water:TFA was applied to each slide, using the matrix spraying application method. The slides were then mounted onto gold-plated stainless steel MALDI target plates using double-sided tape. The plates were loaded sequentially into the Voyager DE-STR and analysed using the imaging MS mode. The calibration was performed externally and repeated prior to each of the six analyses, using calibration 3. After the data had been acquired, they were imported into BioMap for image visualization and analysis.


   Results and Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results and Discussion
Conclusion
 References
 Author Biography 
 
MALDI Imaging Method Development
Immunohistochemistry
The IHC of the male rat kidney tissue section showed strong peroxidase staining (brown precipitate) in the cortex region of the kidney and very weak peroxidase staining in the female tissue section. In addition, results from both the male and female control sections (where the primary antibody was omitted) showed no peroxidase staining. These data agrees with Burnett et al.8 who state that {alpha}2u-globulin is located within the cortical region of the nephron, with the male rat kidney containing 100-fold more {alpha}2u-globulin (3% of total protein) than the female rat kidney (0.03% of total protein).

These IHC images (Fig. 4) provided comparisons for the male and female images from the proof-of-principle MALDI imaging experiment.


Figure 4
View larger version (119K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 4. Immunohistochemistry data. (A) Male rat kidney section showing strong peroxidase staining for alpha2u-globulin, (B) female rat kidney section showing weak strong peroxidase staining for alpha2u-globulin, (C) male rat kidney control showing no peroxidase staining and (D) female rat kidney control showing no peroxidase staining.

 
Qualitative Assessment Definitions of MALDI Imaging Results
There is currently no defined method of assessing the quality of the images produced from MALDI imaging experiments. However, it is necessary to define a method of qualitative assessment to enable comparison between the images created. The images will be placed into quality categories, which are based on a combination of the ability to distinguish the tissue section from the surrounding matrix and the number of resolved peaks visible on the spectrum (Table 3).


View this table:
[in this window]
[in a new window]

  		Table 3. Developed MALDI imaging quality assessment criteria used to determine the quality of the images produced during the method development and proof-of-principle experiments

 
Determination of Matrix Volume for MALDI Imaging of Rat Kidney Sections
Images were created in BioMap using the m/z 15.3 peak, which was shown to be present throughout the rat kidney and identified as haemoglobin subunit alpha-1/2. The protein peaks within the spectrum were identified based on mass region, where the cortex and medulla/papilla regions were separately subjected to micro-extraction (adapted from the method stated by Meistermann et al.11) followed by nano-high-pressure liquid chromatography/QTRAP MS/MS analysis (data not shown).

For each section, one image was created and two spectra were visualized; one by highlighting an area within the section and one by selecting an area outside of the tissue section. This then shows a spectrum within the tissue to validate the accuracy of the image created, and shows whether there is any delocalization outside of the tissue, in the surrounding matrix area. Overall quality of images for the sections is shown in Table 4 (images and spectra not shown).


View this table:
[in this window]
[in a new window]

  		Table 4. Image, spectrum and overall quality of the images of plates 1–4 created to assess the volume of matrix required to successfully acquire a rat kidney section MALDI image

 
The results from this experiment showed that plates 1 and 2 to which 5 ml of matrix were sprayed, produced an overall poor quality image. On closer inspection the fairly well-defined tissue section was a result of a higher matrix background in the spectra from within the tissue than that of the spectra from outside of the tissue area. Plate 3, on which 5 ml of matrix were sprayed and an additional volume of 0.2 µl was spotted on top according to the dried-droplet method, showed protein delocalization outside of the tissue boundaries and the image was deemed to be of poor quality. This is most likely due to the diffusion of the matrix spot over the tissue surface,12 which resulted in the physical movement of proteins that are soluble in the matrix solution across the surface of the plate. It is therefore highly plausible that protein delocalization occurred not only outside of the tissue but also within the tissue resulting in a loss of spatial integrity,6 essentially eliminating the one major advantage of this technique over a typical MS analysis. Plate 4, on which a total of 15 ml matrix was sprayed (5 ml and then 10 ml), resulted in a medium-quality image because the edges of the tissue were not well defined, but the spectra showed high-quality peaks.

From the results of plates 1 and 2, it was apparent that 5 ml of matrix were insufficient to ionize the proteins within the tissue and thus more matrix needed to be added. This theory was proved correct by the results of plates 3 and 4, where additional matrix was added and peaks were visualized. However, the addition of matrix according to the dried-droplet technique is unsuitable for MALDI imaging as it resulted in the delocalization of proteins. Therefore, these results define the most suitable matrix application technique as follows: 15 ml of matrix solution sprayed onto the tissue section using an airbrush, provided that the tissue is barely wetted during the spray application and is allowed to fully dry before the next spray coat.

This experiment showed that the imaging parameters are suitable for imaging of rat kidney sections. Although the matrix application for plate 4 was shown to be the best, the image produced was of medium/high quality, meaning that an improvement to the quality of the image could potentially be made by optimizing the matrix solvent composition. Therefore the next experiment was to determine the best possible matrix solvent composition for MALDI imaging of rat kidney sections.

Matrix Solvent Composition Determination
The graphs created for the ACN, EtOH and MeOH solvent compositions are in shown in Figs 5Go–7, respectively. In order to determine the best matrix solvent compositions, the mean peak heights were compared and, based on two criteria, the three best compositions were chosen. Firstly, the mean peak height of m/z 18.3 was considered to be the most important, because the aim of the matrix solvent for this study was to maximize the peak height in this region as {alpha}2u-globulin has a molecular mass of 18.7 kDa and secondly, to give the maximal height for the other seven peaks within the protein mixture. The raw spectra from this study showed that all of the proteins were detected with an accuracy of 0.01% (i.e. for the 1000 Da proteins the measured mass will be 1000 ± 0.1 Da) (data not shown).


Figure 5
View larger version (34K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 5. Matrix testing of ACN concentrations with and without TFA, where bars represent the mean and error bars represent ± 1 SD.

 


Figure 6
View larger version (31K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 6. Matrix testing of EtOH concentrations with and without TFA, where bars represent the mean and error bars represent ± 1 SD.

 


Figure 7
View larger version (32K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 7. Matrix testing of MeOH concentrations with and without TFA, where bars represent the mean and error bars represent ± 1 SD.

 
From the data reported in Figs 5Go–7, the following matrix solvent compositions were chosen: 50:50:0.1 ACN:distilled water:TFA, 60:40:0.1 EtOH:distilled water:TFA and 80:20:0.1 MeOH:distilled water:TFA. These three represented the compositions that gave the highest mean peak for m/z 18.3, while still showing high mean peaks for the remaining proteins in the mixture. Typically, those compositions containing 0.1% TFA gave increased mean peak heights when compared with those that contained 0% TFA.

The peak at m/z 25.8 was not detected using any of the 24 matrix compositions. This could be due to two reasons; first, the concentration of this protein may have been too low to detect (2.4 pmol) or secondly, that this particular protein does not ionize sufficiently in MALDI-MS.

For most of the data, the variation (SD) was high, which makes it difficult to confidently assess the best matrix solvent composition. However, the heterogeneity of matrix crystals within a spot is high particularly around the outside of the spot as indicated by Alterman et al.,13 and therefore it may not be possible to gain data with low variation.

From the data of the three chosen matrix solvent compositions, it seems likely that the ACN or EtOH compositions will result in high-quality images as the heights of the peaks were consistently higher than those for the MeOH composition (except for m/z 4.3). The next stage of the method development was to test these three matrix solvent compositions on rat kidney tissue.

Determination of Matrix Composition and Optimum Tissue Thickness for MALDI Imaging of Rat Kidney Sections
Images were created in BioMap for each plate using the m/z 15.3 peak; spectra are shown for each image within and outside of the tissue section. Here the various compositions are abbreviated by referring to the solvent they contain. The images of the 5 µm sections and spectra are shown in Fig. 8, and the overall quality of images for the sections is shown in Table 5.


Figure 8
View larger version (49K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 8. MALDI images and spectra for the determination of matrix composition and optimum tissue thickness for MALDI imaging of rat kidney sections. (A) ACN 5 µm, (B) EtOH 5 µm, (C) MeOH 5 µm and (D) intensity legend, where (i) image and spectrum at m/z 15.3 within the tissue section and (ii) spectrum outside of the tissue.

 


View this table:
[in this window]
[in a new window]

  		Table 5. Image, spectrum and overall quality of the images created to determine the matrix composition and optimum tissue thickness for MALDI imaging of rat kidney sections

 
Based on spectrum quality, this table indicates that the ACN and EtOH proved to be the best matrix solvent compositions; however substantial protein delocalization was detected in both of the EtOH plates, resulting in the overall quality of the image being classified as poor. This delocalization is most probably due to the greatly increased drying time for EtOH when compared to ACN and MeOH; as it was difficult to know whether the surface of the ITO-coated glass slide was completely dry before the next layer of matrix was applied, the EtOH slides may have become wet with matrix solution. Schwartz et al.14 demonstrated that increasing the amount of time a slide remains wet increased the probability of proteins delocalizing within and outside of the tissue section.

Although the highest spectra quality were obtained with the EtOH matrix solvent composition, the labour and time intensive nature of matrix coating (i.e. more than 6 min may be required in between application of each matrix coat) makes this solvent unsuitable for MALDI imaging of rat kidney sections, particularly when, as shown by Cornett et al.,15 the samples need to be analysed as promptly as possible to prevent molecular degradation. The MeOH matrix solvent composition was discounted due to the lower quality of the images when compared with the ACN composition. Therefore 50:50:0.1 ACN:distilled water:TFA was chosen as the best matrix solvent composition for MALDI imaging of rat kidney sections. Interestingly, this result is comparable to the matrix solvent composition of choice for protein MALDI imaging of several other research groups.12, 14, 16–18

In terms of tissue thickness, 5 µm was shown to produce consistently higher quality images when compared with 12 µm tissue sections. This data was comparable to that found by Sugiura et al.17 who stated that tissues <10 µm in thickness were optimal for MALDI imaging.

The matrix consisting of 35 mg/ml sinapinic acid in 50:50:0.1 ACN:distilled water:TFA solution, in combination with a tissue thickness of 5 µm generated highest quality images and so, this was applied to the proof-of-principle MALDI imaging experiment of male and female rat kidney sections.

Proof-of-principle MALDI Imaging Study of Male and Female Rat Kidney Sections
Five male rat kidney sections and five female rat kidney sections were imaged using the MALDI imaging method described above to determine whether the created MALDI images could be used to visualize differences in expression levels of {alpha}2u-globulin and whether the localization of this protein was comparable to the IHC data. A selection of the images and spectra for the male (3 and 5) and female (2 and 5) sections is shown in Fig. 9 and the overall quality of images for the sections are shown in Table 6.


Figure 9
View larger version (50K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 9. MALDI images and spectra of male and female rat kidney sections, from the proof-of-principle study. (A) Male 3, (B) Male 5, (C) Female 2, (D) Female 5 (E) intensity legend, where: (i) image at m/z 15.3 with spectra from area highlighted within tissue section, (ii) spectrum from outside of the tissue boundaries and (iii) image and spectrum at m/z 18.7.

 


View this table:
[in this window]
[in a new window]

  		Table 6. Image, spectrum and overall quality of the images created for the proof-of-principle study

 
The images constructed at m/z 15.3 for the 10 tissue sections showed that, except for female section 3, all of the images were of medium or high quality, and no protein delocalization was observed for any of the tissue sections. The spectra within the tissue sections of the male samples were all of high quality, whereas the female samples showed a variation in the quality of the spectra. The spectra obtained from the female sections showed lower peak intensity. This could be due, in part, to the overall size of the female kidney, and therefore tissue sections being smaller than those of the males, potentially resulting in lower levels of proteins, although, as the MALDI imaging method was developed using male kidney sections, it may mean that the conditions favour the ionization of the male kidney proteins.

The images created at m/z 18.7 showed that a peak at this m/z (that may correspond to {alpha}2u-globulin based on a micro-extraction experiment) was present in all of the male sections and was localized to the cortex region of the tissue, whereas none of the female sections contained any detectable peak at this m/z. This data was comparable to the IHC staining results. For the male sections, only one of the five images (male 3) showed a comparable image to that seen in the IHC experiment, which clearly shows the localization of {alpha}2u-globulin in the cortex. For the other four male sections, it was necessary to use the accompanying spectra to give a definitive localization of the protein. This demonstrates that the identification of surgical margin status could be established, but only for images that were of high quality. This means that, for the MALDI imaging method developed here, the differences in protein expression can only be visualized on a high-quality MALDI image. If the image is of medium or poor quality, it may neither be possible to definitively identify the protein nor its localization without viewing the spectra to manually validate the presence of the peak. Based on this information it is unlikely that small changes in proteins, such as a 2-fold change, would be identified in medium/poor quality images without visualizing the spectra.

The variation in image quality observed in the male and female sections could be due to the current inability to pre-process and normalize the spectra in MMIST or BioMap before creating the image. The accuracy of the data (5%) was highlighted when exporting the averaged spectra from the cortex regions of the tissue sections. This poor accuracy, which resulted from the inability to utilize internal calibration peaks within BioMap, could be a contributing factor to the observed variation in image quality. Therefore each plate has only a one-point external calibration prior to the collection of the imaging data, and it is highly likely that the spectra obtained close to the calibration spot will be of higher accuracy than those collected from further away. Fortunately, for this study {alpha}2u-globulin is the only peak in the m/z 18–20 region and is therefore not affected by the 5% variation in data. The use of a different MALDI imaging software (ProTS Data) that allows for pre-processing and normalization of spectra, along with internal calibration prior to image construction, could be used in future studies to help reduce the variability of the images.

Statistical Significance
A Mann–Whitney U-test was performed on the signal-to-noise ratio for m/z 18.7 to determine the statistical significance of the difference in {alpha}2u-globulin levels between the male and female sections.

To obtain the relevant data, the cortex regions of the 10 tissue images were highlighted in BioMap. The averaged spectra across this region were imported separately into Data Explorer® and subjected to standard baseline correction and noise removal of 0.7. From these spectra it was possible to observe an indication of a peak at m/z 18.7 in the majority of the female sections (1, 2, 3 and 5), however, the signal-to-noise ratios of these peaks were low and they were classified as background noise. The imported spectra also highlighted the poor accuracy of the imaging experiment, which in some cases was 5%.

Using a p value >0.05, the UCRIT value for the Mann–Whitney U-test was 4 and as U1 was below this value (0), resulting in accepting the alternative hypothesis, there was a significant difference between the {alpha}2u-globulin levels found in the male and female rat kidney sections.


   Conclusion
Top
 Abstract
 Introduction
 Materials and Methods
 Results and Discussion
 Conclusion
References
 Author Biography 
 
From the data presented it is possible to conclude that a successful sample preparation method for MALDI imaging was developed, where a volume of 15 ml of matrix (35 mg/ml sinapinic acid dissolved 50:50:0.1 ACN:distilled water:TFA) was applied onto the tissue using the airbrush spraying method. In addition, a tissue thickness of 5 µm was shown to produce higher quality images when compared with 12 µm tissue sections. MALDI imaging parameters were successfully established, where the tissue sections were analysed using a 75 shot/spectrum acquisition in positive linear MALDI-MS over an area of 60 x 60 pixel array at a distance of 200 µm (both horizontally and vertically). The combination of these methods allowed for the production of some high-quality spectra in the proof-of-principle MALDI imaging comparison of male and female rat kidney sections. However, variation in the quality of images was observed and would need to be addressed before further MALDI imaging experimentation.

A proof-of-principle study was also conducted, which showed a significant difference between the m/z 18.7 peak (which may correspond to {alpha}2u-globulin) expression levels observed in the male and female rat kidney sections thus suggesting that the detection of tumour markers could be used to differentiate between tumour type, grade and stage.

Overall, this study successfully achieved the aims with the development of sample preparation and MALDI imaging methodology, which when applied to male and female tissues sections showed a significant difference between the m/z 18.7 peak levels that was comparable to the IHC data.


   Acknowledgements
 
The author acknowledges Dr S.B. of the Faculty of Sciences, Staffordshire University, and J. Nickson and R. Rowlinson of AstraZeneca, for the use of their expertise and facilities during this investigation.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results and Discussion
 Conclusion
 References
Author Biography 
 

  1. Cancer Research UK. Cancer Worldwide—The Global Picture (2008) Available at http://info.cancerresearchuk.org/cancerstats/geographic/world/#source1 (Accessed April 2008).
  2. Ferrara N, Gerber HP. The role of vascular endothelial growth factor in angiogenesis. Acta Haematol (2001) 106(4):148–156.[CrossRef][Web of Science][Medline]
  3. Tímár J, Csuka O, Orosz Z, et al. Molecular pathology of tumour metastasis: II. Molecular staging and differential diagnosis. Pathol Oncol Res (2002) 8(3):204–219.[Medline]
  4. Moch H, Schraml P, Bubendorf L, et al. High-throughput tissue microarray analysis to evaluate genes uncovered by CDNA microarray screening in renal cell carcinoma. Am J Pathol (1999) 154:981–986.[Abstract/Free Full Text]
  5. Karas M, Hillenkamp F. Laser desorption ionization of proteins with molecular masses exceeding 10 000 daltons. Anal Chem (1988) 60:2299–2301.[Medline]
  6. Caprioli RM, Farmer TB, Gile J. Molecular imaging of biological samples: localization of peptides and proteins using MALDI-TOF MS. Anal Chem (1997) 69:4751–4760.[Medline]
  7. Roy AK, Neuhaus OW. Proof of the hepatic synthesis of a sex-dependant protein in the rat. Biochim Biophys Acta (1966) 127:82–87.[Medline]
  8. Burnett VL, Short BG, Swenberg JA. Localisation of {alpha}2u-Globulin within protein droplets of male rat kidney: immunohistochemistry using perfusion-fixed, GMA-embedded tissue sections. J Histochem Cytochem (1989) 37(6):813–818.[Abstract]
  9. McGregor D, Hard GC. Toxicological highlight: renal tubule tumour induction by tertiary-butyl alcohol. Toxicol Sci (2001) 61:1–3.[Abstract/Free Full Text]
  10. Stoeckli M. Automated mass spectroscopy imaging with a matrix-assisted laser desorption/ionisation time-of-flight instrument. J Am Soc Mass Spectrom (1999) 10:67–71.[CrossRef][Web of Science][Medline]
  11. Meistermann H, Norris JL, Aerni H-R, et al. Biomarker discovery by imaging mass spectrometry: transthyretin is a biomarker for gentamicin-induced nephrotoxicity in rat. Mol Cell Proteomics (2006) 5(10):1876–1886.[Abstract/Free Full Text]
  12. Sugiura Y, Shimma S, Setou M. Two-step matrix application technique to improve ionisation efficiency for matrix-assisted laser desorption/ionisation in imaging mass spectrometry. Anal Chem (2006) b 78:8227–8235.[Medline]
  13. Alterman MA, Kornilayev B, Duzhak T, et al. Quantitative analysis of cytochrome P450 isozymes by means of unique ioszyme-specifc tryptic peptides: a proteomic approach. Drug Metab Dispos (2005) 33(9):1399–1407.[Abstract/Free Full Text]
  14. Schwartz SA, Reyzer ML, Caprioli RM. Direct tissue analysis using matrix-assisted laser desorption/ionisation mass spectrometry: practical aspects of sample preparation. J Mass Spectrom (2003) 38:699–708.[CrossRef][Web of Science][Medline]
  15. Cornett DS, Reyzer ML, Chaurand P, et al. MALDI imaging mass spectrometry: molecular snapshots of biochemical systems. Nat Methods (2007) 4(10):828–833.[CrossRef][Web of Science][Medline]
  16. Chaurand P, Cornett DS, Caprioli RM. Molecular imaging of thin mammalian tissue sections by mass spectrometry. Curr Opin Biotechnol (2006) 17:431–436.[CrossRef][Web of Science][Medline]
  17. Sugiura Y, Shimma S, Setou M. Thin sectioning improves the peak intensity and signal-to-noise ratio in direct tissue mass spectrometry. J Mass Spectrom Soc Jpn (2006) a 54(2):45–48.
  18. Norris JL, Cornett DS, Mobley JA, et al. Processing MALDI mass spectra to improve mass spectral direct tissue analysis. Int J Mass Spectrom (2007) 260(2–3):212–221.[CrossRef][Web of Science][Medline]

   Author Biography 
Top
 Abstract
 Introduction
 Materials and Methods
 Results and Discussion
 Conclusion
 References
 Author Biography 
 
    Cassie Gregson graduated in 2008 with a first-class Honours degree in Biomedical Sciences from Staffordshire University, and was awarded the Best Undergraduate Biology Dissertation prize. Throughout the duration of her part-time degree course, and to the present day, Cassie works for AstraZeneca as a research bioinformatian. She hopes to develop her career within the company, by specialising and adding to her skills in bioinformatics.
Submitted on 30 September 2008; accepted on 18 December 2008


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrowOA All Versions of this Article:
2/2/134    most recent
hzp016v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Google Scholar
Right arrow Articles by Gregson, C.
PubMed
Right arrow Articles by Gregson, C.
Related Collections
Right arrow Biochemistry
Right arrow Cell Biology
Right arrow Medical Science
Right arrow Pathology
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?