Maldi Imaging Mass Spectrometry Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

MALDI imaging mass spectrometry provides the means to measure the distribution of analytes such as peptides and proteins across sectioned tissue. The strength of the method stems from the ability to measure hundreds of analytes in a single experiment without using antibodies and without characterization of the tissue prior to measurement. In this way molecular data can be obtained which allows tissue analysis based on mass spectrometric profiles. In the context of clinical application, the most pertinent MALDI imaging methods are those which allow analysis of formalin-fixed tissues. However, there is currently a lack of detailed, standardized MALDI imaging methods available in the literature. In order to promote the dissemination of protocols and provide start-up groups the ability to immediately begin analyzing FFPE tissues, the current manuscript presents, in detailed protocol format, updated methods required to successfully perform tryptic peptide MALDI imaging on FFPE tissues and integrate this data with peptide identifications made by LC-MS/MS.


Matrix assisted laser desorption/ionization imaging mass spectrometry (MALDI-IMS) directly measures the in situ distribution of analytes such as drugs {Prideaux, 2011 #1073}, lipids {Meriaux, 2010 #427}, peptides {Groseclose, 2008 #363;Lemaire, 2007 #328;Minerva, 2011 #1049;Schober, 2011 #995} and proteins {Balluff, 2011 #1041;Meding, 2012 #1044;Rauser, 2010 #440}. The advantages of the MALDI-IMS approach include the ability to analyze hundreds of analytes in a single experiment without prior knowledge of tissue composition, or the use of antibodies {Chaurand, 2006 #175}. Furthermore, MALDI-IMS can utilize as little as one tissue section per analysis, making it ideal for precious samples.

To date, the most clinically relevant application of MALDI-IMS has been the acquisition of in situ molecular profiles which can be used to classify tissue types and thereby provide patient specific information {Meding, 2012 #1044;Elsner, 2012 #1046;Hardesty, 2011 #996;Samsi, 2009 #783;Yanagisawa, 2003 #150;Caldwell, 2005 #777;Schwartz, 2005 #778}. Until now, most of those applications utilize frozen tissue, which is less than ideal given that the establishment and maintenance of large frozen tissue collections is challenging and expensive {Hood, 2005 #985}.

Clinical samples for histological studies are traditionally treated with fixatives to prevent the tissue from degrading and to maintain cellular structure. The most common fixative used worldwide in hospitals and pathology institutes for light microscopy is 10% neutral buffered formalin {Hood, 2005 #985;Hood, 2006 #936}. Formalin (methanol stabilized formaldehyde) creates intra and inter-protein cross-links between multiple amino acids, thereby halting tissue degradation and allowing storage at room temperature for decades {Gustafsson, 2011 #979;Addis, 2009 #934;Hood, 2006 #936}. The longevity of FFPE tissue has led to the accumulation of substantial patient sample archives, which have been collected alongside clinico-pathological data {Grantzdorffer, 2010 #935}. With suitable methods for accessing FFPE tissues, these archives represent a rich resource for tissue characterization using MALDI-IMS {Addis, 2009 #934}.

In order to circumvent the formalin-induced cross-links most groups have either adopted a proteolytic enzyme {Lemaire, 2007 #328;Aoki, 2007 #815} or antigen retrieval and enzyme combination treatment {Groseclose, 2008 #363}. Recently applied methods demonstrating the usage of antigen retrieval in combination with tryptic proteolytic digestion have shown marked success in reproducibly generating peptides which are discriminatory for known histological features {Groseclose, 2008 #363;Gustafsson, 2010 #399}. The methods necessary to prepare FFPE samples, perform MALDI-IMS, analyze data and characterize the resulting peptide features have not yet been standardized and are largely developed in-house by individual laboratories. The low number of detailed protocols in the field {Casadonte, 2011 #1024} prompted the current manuscript, in which the methods developed and published in two previous manuscripts by our group {Gustafsson, 2010 #399;Gustafsson, 2012 #1063} have been condensed into a single workflow combining tryptic peptide MALDI-IMS and subsequent LC-MS/MS identification of the peptides which localize to a region of interest (e.g. tumour tissue). As such, the current manuscript provides detailed descriptions of the sample preparation, acquisition and data analysis methods needed to repeat the aforementioned result. In addition, key explanations, method expansions and a bio-informatics tool have been provided to make the protocol as straightforward as possible.


3.1 Chemicals, consumables and equipment

See Table 1 for a comprehensive list of chemicals, consumables and equipment used in this protocol.

Protocol specific solutions

10 mM citric acid monohydrate pH 6

1.05 g citric acid monohydrate

Dissolve in 480 mL H2O

pH to 6.0 using 1M NaOH (~13 mL)

Make up volume to 500 mL with H2O

100 mM ammonium bicarbonate (NH4HCO3) stock

197 mg NH4HCO3 in 25 mL H2O

Internal calibrant solution (100 µL volume with 0.1% TFA v/v)

0.4 pmol/µL Angiotensin I34-43

0.4 pmol/µL [Glu1]-Fibrinopeptide B

2.0 pmol/µL Dynorphin A1-17

2.0 pmol/µL ACTH1-24

Matrix solutions (% is v/v)

MALDI-IMS (ImagePrep standard solution)

7 mg/mL α-cyano-4-hydroxycinnamic acid (CHCA) in 50% ACN, 49.8% H2O and 0.2% TFA

LC-MS/MS (Bruker Daltonics method).

748 µL 95% ACN, 4.9% H2O and 0.1% TFA

36 µL saturated CHCA in 90% ACN, 9.9% H2O and 0.1% TFA

8 µL 10% TFA in H2O

8 µL 100 mM NH4H2PO4

ImagePrep methods

Trypsin deposition (adapted from Bruker Daltonics method)

30 spray cycles

Spray time of 1.25 seconds

Dry time of 45 seconds

CHCA matrix (adapted from Bruker Daltonics method)

Phase 1 - 0.65V within 8-20 cycles, 10 s incubation and 90 s drying

Phase 2 - 30 s drying

Phase 3 - 0.1V within 4-10 cycles, 25% ± 35% spray power with 0.05V sensor controlled spray time, 30 s ± 30 s incubation and complete drying every cycle, safe dry set at 10 s

Phase 4 - 0.1V within 8-12 cycles, 25% ± 35% spray power with 0.1V sensor controlled spray time, 30 s ± 30 s incubation and grade 20% ± 40% complete dry every 2 cycles, safe dry set at 20 s

Phase 5 - 0.3V within 12-30 cycles, 25% ± 35% spray power with 0.2V sensor controlled spray time, 30 s ± 30 s incubation and grade 30% ± 40% complete dry every 3 cycles, safe dry set at 40 s

Phase 6 - 0.6V ± 0.5V within 20-64 cycles, 25% ± 35% power with 0.3V sensor controlled spray time, 30 s ± 30 s incubation and grade 30% ± 40% complete dry every 3 cycles, safe dry set at 40 s

Tissue block sectioning, water mounting and antigen retrieval

Prior to beginning, fill the water bath with H2O and set the temperature to 39°C.

If the temperature is set too high the water will melt the paraffin and increase the difficulty of mounting the sections.

Mount the FFPE block onto the sample holder and ensure the angle of the holder is appropriate. Then insert the microtome blade.

Set the sectioning thickness to 20 µm and trim the block until the tissue area(s) of interest is visible.

Change the sectioning thickness to between 4 and 6 µm. Begin collecting sections.

The thickness which maintains tissue section morphology best and gives the most consistent results should be the standard. Optimum section thickness for MS, in our hands, is 6 µm.

Use a fine-tip brush to lay the tissue section(s) gently onto the surface of the water in the pre-heated water bath.

Make sure the sections are not folded or rolled prior to floating in the water bath.

Water mount the sections onto an indium tin oxide (ITO) coated slide by immersing the slide beneath the section (at the desired location) and using it to lift the section out of the water.

Mount the section in one smooth motion to prevent section curling or folding.

The paraffin edge of the section will stick to the slide. This can be exploited to stabilize the section prior to final section mounting.

Finally, ensure that no air bubbles are caught beneath the section. This can cause the tissue section to detach from the slide during antigen retrieval.

Leave a ~1 cm wide space in the middle of the slide free of tissue. This space is required by the light scatter sensor in the ImagePrep sample preparation instrument.

Holding the prepared slide vertically, gently tap excess water onto a lint-free wipe.

Place the slide vertically into a container (e.g. coplin jar) and allow the sections to air dry at room temperature.

Optionally, slides can be dried by incubating at 56°C.

Lay the slide(s) [tissue side facing up] onto a heating block at 60°C for one hour.

Heating improves the adherence of tissue to the slide.

Wash slides twice in 100% xylene (5 min each) to remove paraffin (vertical slide orientation for all washes).

Wash the slides twice with 100% ethanol (EtOH) [2 min each] and allow the slides to dry.

Mark the slide(s) surface using a water-based white out (e.g. Tipp-Ex) or diamond tipped pen.

The marks are used to teach the tissue position in the mass spectrometer.

Place the slide(s) in a coplin jar and fill the empty slots with blank microscopy slides.

A full jar prevents formation of large bubbles during antigen retrieval and has been observed to reduce section fragmentation.

Rinse sections twice in 10 mM ammonium bicarbonate (NH4HCO3) for five minutes.

Discard the NH4HCO3 and re-fill the coplin jar with 10 mM citric acid mono-hydrate at pH 6 (see section 3.2.1) {Shi, 1993 #378;Gustafsson, 2010 #399}.

With the lid loosely placed on top, put the coplin jar in a microwave and use standard settings (i.e. quick start) to bring the citric acid solution to a boil.

The lid should be loosely placed on the coplin jar to prevent a dangerous pressure build up during microwave heating.

Once the solution reaches boiling point, set the microwave power such that solution temperature is maintained just below boiling point. Incubate at this power setting for 10 min.

Settings will vary depending on microwave model. Test settings using a set of blank slides first.

Our laboratory uses a LG 700W MS19496 microwave oven.

Following microwave incubation, transfer the coplin jar to a heating block set at 98°C for 30 minutes.

Remove the slide(s) following incubation and allow it to cool to room temperature.

Rinse the slide using two 1 min dips in 10 mM NH4HCO3 (use fresh NH4HCO3 for the second dip).

This step prepares the tissue for tryptic digestion by neutralizing the citric acid remaining on the tissue section. Trypsin activity is optimal at pH 8.5.

Dry the slide briefly (5-10 min) in a vacuum desiccator.

Our facility uses a 190 mm polypropylene vacuum desiccator (Kartell, Italy).

Tryptic digestion, internal calibrant application and matrix deposition

Scan the slide(s) which will be analyzed at a minimum resolution of 1400 dpi.

Our laboratory currently uses a CanoScan 5600F (Canon, Thailand) scanner.

Ensure that a computer is connected to the ImagePrep for light sensor signal read out.

NI VI Logger Lite (V2.0.1, National Instruments, Austin, TX) receives input from the analogue-digital converter attached to the ImagePrep (see ImagePrep user manual, V2.0).

Light sensor responsiveness can be tested by blocking light to the sensor. This should reduce the measured signal dramatically.

Perform a single cleaning cycle on the ImagePrep using MeOH and use lint-free wipes to clean the inside of the main chamber.

Load 200 µL distilled H2O directly onto the spray generator (see ImagePrep user manual, V2.0) and select the trypsin deposition method (see section 3.3.1).

Open the spray offset menu in the ImagePrep other menu and set the spray power and modulation such that the spray lasts 30-40 seconds at 38% spray power with 0% modulation.

If the spray does not start during the test, the loaded droplet has probably not reached the spray generator.

To fix this, stop the run, open the side door and tilt the solution vial slightly back out of its home position. Tap it back into position. This should move the droplet down onto the spray generator.

Dissolve 100 µg lyophilized trypsin gold in 200 µL 5 mM NH4HCO3.

Split the total volume into 40 µL aliquots and freeze these at -80°C until ready for use.

Mix 40 µL 0.5 µg/µL trypsin gold stock with 160 µL 25 mM NH4HCO3.

Load the full 200 µL of diluted trypsin gold (concentration = 100 ng/µL) onto the ImagePrep spray membrane (see ImagePrep user manual, V2.0).

Start the deposition and monitor the quality of the trypsin spray.

The trypsin spray should cover the tissue sections completely and dry (no droplets visible) within the 45 s drying time.

The spray should easily reach the end of the slide.

The entire volume of trypsin should be deposited over 25-35 spray cycles. Restart the method if required.

Once the entire volume has been deposited, immediately remove the slide from the ImagePrep and place it into an airtight container along with wet paper towels (creates a humid atmosphere).

Place the container at 37°C for 2 hours.

Following completion of the tryptic digest, load the internal calibrant solution (see section 3.2.3) onto the ImagePrep station spray generator.

Replace the slide in the ImagePrep station and use the trypsin deposition method to coat the tissue sections with the internal calibrant solution.

While the calibrants are being deposited, prepare ≥5 mL of 7 mg/mL α-cyano-4-hydroxycinnamic acid (CHCA) [see section 3.2.4].

At least half of the ImagePrep solution vial needs to be filled to prevent negative pressure in the vial during preparations. Negative pressure reduces spray power.

Change the spray offset, as before, to those indicated in the CHCA method (see section 3.3.2).

For phase 1 of the matrix method this is 20% power ± 35% modulation.

Perform a test spray. The matrix mist should just reach the far end of the ImagePrep and homogeneously cover the central raised area of the ImagePrep chamber.

If the spray power is not adequate change the spray offset to compensate.

To maintain spray quality, the spray offset may need to be increased in-between phases during sample preparation.

Prior to starting the automated preparation wipe away the deposited matrix using a small volume of MeOH and a lint-free wipe.

Place the ITO slide (with experimental sections) onto the central platform of the ImagePrep chamber.

The light scatter sensor should not be covered by tissue.

Place a glass cover-slip over the light sensor (prevents light scatter artifacts interfering with sensor controlled sample preparation).

Begin the matrix sample preparation.

Phase 1 is the only preparation phase which is not controlled by the light sensor. As a result, monitor the deposition to ensure that excessive amounts of matrix are not being deposited.

When phase 3 is reached, monitor the preparation using the LoggerVI software to ensure that the automated sample preparation proceeds correctly.

The sample preparation from phase 3 onwards relies on formation of a light scatter curve (see Figure 2).

If sample preparation is not proceeding correctly you will notice two things:

The spray power boost will increase incrementally (if it passes 20% stop the preparation and restart phase 3).

The cycle number will not change.

To correct this you can increase the number of minimum cycles in phase 1 or increase the spray offset for the preparation.

The critical step is to achieve a stable light scatter curve by the end of phase 1 (see Figure 2)

Once the preparation is stable in phase 3 it no longer needs to be actively monitored.

When the preparation is complete, remove the slide from the ImagePrep, dispose of the cover-slip and use a small volume of MeOH on a lint-free wipe to wipe away the matrix from a 0.5 cm wide space at both ends of the slide.

The space ensures that the metal on the slide adapter contacts the conductive slide surface.

Load the prepared slide into a MTP slide adapter II.

Re-suspend a tube of peptide calibration standard II in 125 µL 0.1% TFA to create a stock solution. Dilute this stock 1:50 with some of the remaining matrix solution (see section 3.2.4).

Sonicate the peptide calibration standard for five min and then deposit 0.5 µL of the solution onto the clean slide area left behind by the cover-slip.

Once the peptide calibration standard II spot has crystallized, load the MTP slide adapter into the MALDI-TOF/TOF instrument.

Our group uses an ultrafleXtreme MALDI-TOF/TOF instrument.

MALDI-TOF/TOF mass spectrometry data acquisition

NB: This protocol utilizes ultrafleXtreme MALDI-TOF/TOF specific programs and settings.

Load a suitable reflectron positive MALDI-IMS method in flexControl (V3.3).

Settings which can be optimized at this point include:

Laser power (instrument specific)

Laser offset (instrument specific - typically does not need to be altered for CHCA matrix preparations)

Laser repetition rate (typically maximum, 1 kHz for ultrafleXtreme)

Detector gain

m/z measurement range (at least m/z 1000 - 3500)

Matrix suppression (700 Da, usually 100 Da below m/z range minimum)

Acquisition rate (at least 1 GS/s for reflectron MS)

Open flexImaging (V2.1 or newer) and select create a new sequence in the window that appears. Use the wizard to set up all the relevant acquisition details:

Sequence name

Data directory

Sample preparation type (in this case uniformly distributed coating - 100 µm raster width)

AutoXecute method (edited through flexControl)

Use the scanned image from step 5.1 to teach the slide positions on the section image.

When the three teach marks have been assigned, select move sample carrier from the edit drop down menu. Move the sample carrier to the edges of the assigned teach marks to confirm accurate teaching.

Move the sample carrier to a region of prepared tissue. Optimize laser power on the tissue to ensure that sample preparation has been successful.

Note that this protocol analyses cellular proteomes. As such any acquired spectra are highly complex and as such slight arching of the spectrum is typical (see Figure 3) and overlap of peptide isotopic profiles can occur.

An optimum laser power has been achieved when a high intensity spectrum (≥ 5000 counts for base peak) can be acquired with 500 accumulated shots. The intensity will be variable and can exceed 104 for some spectra.

Test the selected laser power on several randomly selected spots.

All four internal calibrants should be detectable in a typical high S/N spectrum. Check this prior to starting a full automated acquisition.

Using the same method as steps 6.3 and 6.4, move the sample carrier to the peptide calibration standard II position of the slide. Acquire 500 shots from the calibrant position, ensuring that the peaks in the isotopic profile can be resolved to baseline.

Calibrate the instrument using the standard peptide calibration list provided by the proprietor and save the flexControl method.

Use the same or similar laser powers for both the external calibrants and sample spectra acquisitions.

Ensure that the flexControl (V3.3) autoXecute method contains the appropriate parameters.

The autoXecute tabs will have the following features:

General - flexControl method only selected

Laser - set optimum laser power and turn fuzzy control off

Evaluation - should contain no background list

Accumulation - fuzzy control off, 500 shots acquired in 500 shot steps, dynamic termination off

Movement - random walk off

Processing - flexAnalysis method required (see below)

MS/MS - no method

flexAnalysis method should perform the following processing (see Supplementary Figure 1 for flexAnalysis script required). The following additional settings need to be defined in the flexAnalysis Edit Processing Parameters and Edit Parameters windows.

Smoothing (Gaussian, 2 cycles with width of m/z 0.02)

Baseline subtraction (TopHat)

Peak detection (mono-isotopic, e.g. SNAP using averagine peptide composition)

Re-calibration (quadratic internal re-calibration, 300 ppm tolerance using a custom mass control list [.mcl] containing the following calibrants):

Angiotensin I [M+H]+: 1296.685

[Glu1]-Fibrinopeptide B [M+H]+: 1570.677

Dynorphin A [M+H]+: 2147.199

ACTH1-24 [M+H]+: 2932.588

Trypsin (autolysis) [M+H]+: 2211.104

Prior to starting the data acquisition, two important changes to the data loading settings need to be made. In flexImaging, go to the Edit dropdown menu and select sequence properties. Select read processed spectra, tick the box next to store reduced data for faster access and set the number of data points to 20 000 (only for reflectron data). Select ok. After data acquisition, make sure to save the imaging sequence, which creates a .dat file from the reduced spectra in the results directory.

Once the data acquisition is completed, flexImaging will automatically load all the spectral data files.

This is not the case if the autoXecute batch runner is used to run several imaging sequences. These sequences need to be loaded individually.

Start the imaging sequence from the flexImaging checklist menu or by selecting the start autoXecute run icon.

Once the run has completed, the slide can be removed from the MS instrument and washed in 70% EtOH to remove the matrix (5 min).

Stain the MALDI-IMS section(s) with an appropriate standard protocol. Haematoxylin and eosin (H&E) staining is preferred by our group.

Alternatively, a consecutive section can be used for histology.

Following staining, scan the tissue section. Our group uses a NanoZoomer automated slide scanner (Hamamatsu, Japan) to acquire images of stained sections for annotation (20 x objective lens).

Scans can be annotated by a pathologist to identify the cancerous/non-cancerous region(s).

MALDI-IMS cancer-specific peptide visualization and calculation of abundance weighted peptide m/z

NB: Although multivariate statistics are best used to select peaks of interest {Jones, 2011 #1029;Rauser, 2010 #440;Djidja, 2010 #792}, to keep the workflow as straightforward as possible, peak selection here is manual only and based on operator visualized ion intensity maps for peptides of interest.

NB: Steps 7.1 to 7.6 use flexImaging

NB: Steps 7.7 onwards use in-house data processing

When the MALDI-IMS data set loads completely in flexImaging (step 6.9), a sum spectrum will be generated in the Spectrum Display window.

Mass filters can be added to this sum spectrum to show the spectral intensity (normalized or non-normalized) within the filter mass range (see Figure 4).

Create a region of interest (ROI) around the tissue section area identified as cancerous (pathologist determined) using the polygon ROI tool.

If there is not sufficient contrast in the original section scan to match to the histological stain, co-register the scan of the stained section by selecting Co-Register Image… from the Edit drop down menu in flexImaging. Follow the wizard prompts to complete the co-registration.

In the Regions window select the ROI spectrum checkbox and use the slider at the far right of the Spectrum Display window to move to the corresponding spectrum.

From the cancer-only sum spectrum add mass filters to the mono-isotopic peaks of cancer specific peptides (see Figure 4).

To add a mass filter while in zoom mode, hold control and click on the sum spectrum at the desired m/z location.

To modify mass filter width (absolute or percentage) double click on the mass filter in the Results window list. Typical setting is between 0.5 Da.

To generate abundance weight mean (AWM) m/z values for the peptide features in the MALDI-IMS data set, the .xml files from each individual spectral folder are processed using an in-house developed software tool (available on request from authors).

The software tool (written in Java) uses similar processing work flows to that presented in a previous manuscript {Gustafsson, 2012 #1063}.

First, .xml peak lists are extracted from the folders containing each individual spectrum.

Peak lists are combined into a single list sorted by m/z.

Peak groups (i.e. peptides) are defined by applying a linkage distance, in this case 0.05, such that the distance between two adjacent peptide peak groups is always greater than 0.05.

The peaks in each group are then used to calculate an abundance weighted mean.


The larger peptide of interest mentioned, m/z approximately 1905.97 (Figure 4) was found to have an AWM m/z of 1905.981, as calculated from all spectra using a single linkage distance of 0.05 m/z.

The AWM m/z values for peaks of interest can now be matched to LC-MS/MS data generated from a tryptic digest of an adjacent section of cancer tissue.

nLC-MS/MS for tryptic peptide identification

NB: This protocol utilizes laser-capture micro-dissection (LCM) of FFPE material.

NB: The following methods refer to the use of an ultrafleXtreme MALDI-TOF/TOF MS instrument as well as the proprietary software required for this instrument.

For LCM, sections of the same FFPE tissue block must be water bath mounted onto PEN membrane slides.

Briefly, FFPE LCM uses a laser system to cut through a polymer membrane the tissue section is mounted on. The free polymer-tissue piece is then collected into a micro-centrifuge tube for further processing.

Our laboratory uses a Leica AS LCM microscope maintained by a collaborating facility (Adelaide Microscopy, Adelaide, South Australia).

Dry the water bath mounted sections overnight at room temperature.

Heat the PEN membrane slide(s) at 60°C for five minutes (heating block).

Remove paraffin using a quick 90 s dip in xylene followed by a 60 s wash in 100% EtOH.

Avoid longer incubations in xylene as the polymer membrane on the PEN slide will start to degrade.

No polymer contamination, attributable to the membrane, has been observed using the protocols described here.

Prepare a solution of 10 mM citric acid monohydrate at pH 6 (see section 3.2.1).

Pipette a small volume of the citric acid solution into the cap of a centrifuge tube suitable for the model of LCM microscope being used.

Load the prepared tubes and slides into the LCM instrument.

Perform LCM and ensure that the pieces have been collected into the cap of the tube.

Following completion of tissue excision and collection, centrifuge the tubes to ensure that all tissue pieces are in the solution at the bottom of the tube.

Add 10 mM citric acid (pH 6) to a volume of 200 µL and heat the tubes for 45 minutes at 98°C.

Allow the tubes to cool to room temperature and centrifuge briefly. Remove the citric acid and replace it with 10 µL 25 mM NH4HCO3.

Following a brief incubation (60 s), replace the NH4HCO3 with 10 µL fresh 25 mM NH4HCO3.

Add 10 µL 10 ng/µL Trypsin Gold and place the digest at 37°C for two hours.

Stop the proteolytic digest using 1 µL 10% TFA in H2O.

Use a reverse phase material to purify the peptides. Our laboratory uses C18 spin filter columns (Thermo-Fisher Scientific, Rockford, IL) to purify up 30 µg total peptide.

Use one spin column per sample, seated on top of a 1.5 mL micro-centrifuge tube.

Centrifuge speed is set to 3000 x g for one minute each.

To purify peptides, centrifuge the following solutions through the spin columns.

200 µL 100% ACN (2x)

200 µL 50% ACN, 49.5% H2O and 0.5% TFA (2x)

200 µL 2% ACN, 97.5% H2O and 0.5% TFA (3x)

40 µL 2% TFA in H2O (1x)

100 µL 2% TFA in H2O (1x)

Replace the micro-centrifuge tubes with a fresh tube.

Centrifuge the peptides through the spin column (x3). Each time pipette flow through onto the column membrane and re-centrifuge.

To wash the peptides, centrifuge through 200 µL 2% ACN and 0.5% TFA (3x).

Replace the micro-centrifuge tube with a fresh 1.5 mL tube.

Elute the peptides using 20 µL 70% ACN and 1% TFA (3x).

Transfer eluate to a HPLC vial and reduce the 60 µL volume to ~5 µL using a vacuum centrifuge.

Increase the volume of the reduced peptide solution to 10 µL using 2% ACN with 0.1% TFA (MALDI) or 0.1% FA (ESI).

The prepared sample can now be loaded onto a capillary or nano-HPLC system connected to an appropriate fraction collector, which can deposit fractions onto a solid sample support (e.g. MTP-384 800 µm AnchorChip target, Bruker Daltonics).

Our laboratory uses a nano-HPLC Ultimate 3500 RS Nano/Cap system (Dionex, Amsterdam) connected to a Proteineer fraction collector (Bruker Daltonics).

Trap column - 2 cm C18 Pepmap100 (3 µm, 100 Å)

Analytical column - 15 cm C18 Pepmap100 (3 µm, 100 Å)

Solvent A: 97.95% H2O, 2% ACN and 0.05% TFA

Solvent B: 80% ACN, 19.96% H2O and 0.04% TFA

2 µL sample injected and loaded for 10 min (3 µL/min) onto trap column at 0%B.

Main gradient 8-42%B over 46 minutes (can be increased to 8-50%B to encompass more hydrophobic peptides)

Followed by 42-90%B in 5 minutes

Flow rate set as 300 nL/min

Fraction collection every 15 seconds (184 fractions total)

Fractions were mixed with matrix in a micro-tee junction (Upchurch Scientific) prior to discontinuous deposition on MTP-384 800 target

CHCA matrix (see section 3.5.4) supplied at 150 µL/hour by a syringe pump (Cole-Palmer), combining 625 nL matrix with 75 nL of eluate every 15 s

Once all fractions have been collected dilute one µL of peptide calibration standard II with 49 µL of 30% ACN with 0.1% TFA.

Further dilute the standard using 160 µL of the CHCA matrix described above.

Spot 0.7 µL of the diluted calibrant onto each solid sample support calibration position.

Open flexControl on the MALDI-TOF/TOF acquisition computer, load the AnchorChip target into the MS instrument and create three autoXecute methods for acquisition of data on the instrument (see Table 1).

The three methods are for acquisition of calibrant, sample and MS/MS spectra. A general guide to their setup is provided below.

When the AnchorChip is loaded, ensure that prior to data acquisition, the source high vacuum has reached at least 1x10-6 mbar. Ideally, this value should be below 8x10-7 mbar.

The laser power for the individual flexControl methods for each autoXecute sequence should now be optimized to ensure acquisition of spectra resolved to the baseline with high S/N (ideally ≥5000).

Optimize the each method on its corresponding spot (i.e. calibrant method on the calibrant spots).

Laser power for MS/MS is typically 8-10% more than the MS.

Once you have set up the autoXecute methods, open WARP-LC (V1.2), select File and new autoXecute run.

Select the appropriate target geometry (e.g. MTP-384 800).

When the wizard prompts, tick the calibration and MS check boxes and select the autoXecute methods for both calibrant and MS sample.

These methods need to exist in flexControl before they can be selected here.

Select the data directory, enter the sample name, set the LC delay time (if applicable) and enter the time slice (fraction width in seconds).

Finish the wizard and save the autoXecute sequence in the same directory as the spectra.

Select Start AutoXecute Run (MS or MS/MS) in the WARP-LC main window.

Data acquisition will typically take anywhere from 15 to 25 minutes on an ultrafleXtreme system.

Once the MS acquisition is complete, select Edit for the WARP-LC method line in the main WARP-LC software window.

Save the method as and change the settings in various tabs as indicated in Table 2.

Select Calculate MALDI compound list. When the calculation is complete, the list can be displayed with the compounds selected for MS/MS indicated by a tick mark.

At this point the list can be manually modified to include or exclude peaks of interest, if they are known from previous experiments.

If the list is acceptable (MS/MS on at ≥1000 compounds for complex sample), select Extend AutoXecute Run.

Select Start AutoXecute Run (MS or MS/MS).

Make sure that the spectra being acquired are added to the flexControl sum buffer.

If accumulated spectra are not being added the laser intensity may be too high resulting in reduced resolution and quality of peaks. Laser power may also be low and parent masses of sufficient quality can not be acquired.

If either of these situations occur optimize the laser power to compensate.

Identifications of MALDI-IMS peptides of interest by nLC-MS/MS

Open ProteinScape (V2.1, Bruker Daltonics) and select New Project from the File drop down menu.

Name the new project, select Finish and right click on the project in the Project Navigator window.

Fill in the sample details and select Finish.

In the WARP-LC main screen select the ProteinScape button in the Send data to… section.

Select the project and sample for the data to be loaded into and select Ok.

Once your data has loaded, a list of MS/MS spectra will appear in the Project navigator window (Figure 5a - compounds). A 2D view of the LC-MS run (m/z versus retention time, LC-MS survey view) is also provided in the bottom left corner of the ProteinScape interface (Figure 5b).

The three levels shown in Figure 5a are the project name, sample name and data set name.

Right click on the data set name in the Project navigator window and select Protein search…

Save a method as and select the appropriate search parameters for your sample.

For example:

Search engine: MASCOT

Database/Taxonomy: Homo sapiens

Enzyme: Trypsin

Maximum missed cleavages: 3

Global modifications: None

Variable modifications: Oxidation (M)

Peptide (MS) tolerance: 100 ppm

Fragmentation type: LID/CID

MS/MS tolerance: 1.0 Da

Peptide charge: Instrument dependent (+1 for MALDI)

Select Start to submit the data and search parameters to MASCOT.

Under the data set in Figure 5a there is a search result, MS spectra and the number of compounds which underwent MS/MS.

In Figure 5a the compounds from a loaded .mgf (mascot generic file) file are shown (6184 in total).

Click on the protein search listing which appears (after about 30 s) under the data name in the Project navigator pane of ProteinScape. The listing has an icon that looks like two red bars superimposed on a mass spectrum (Figure 5a).

The main view tab will change to show a protein list (see Figure 5c). Click on a protein in the list and the table underneath will populate with the identified peptides from the protein. The LC-MS survey view will also update to show the locations of the identified peptides in the LC run (Figure 5b).

With a protein selected, hit Ctrl-A to select all proteins. The peptide list will update to show all peptides identified (see Figure 5d).

Scroll through the peptide list to find your peptide m/z of interest. Check the match error to ensure that it is within ± 20 ppm.

Any matches made are now considered tentative, and need to be confirmed either by:


In situ MS/MS (see next section)


The peptide of interest mentioned above - AWM m/z 1905.981 - was matched to heat shock protein beta 1 {Gustafsson, 2012 #1063} in Figure 6. The ppm error between the measured and theoretical (1905.992, ProteinProspector MS-digest {Clauser, 1999 #1052}) is 0.52 ppm.

Confirmation of identification by in situ MS/MS

NB: Two options are available - if the original sample used for MALDI-IMS is less than two days old, then in situ MS/MS can be attempted directly from this section. If not, then the sample preparation should be repeated for a consecutive section to maximize chances of a successful acquisition. The protocol below outlines in situ MS/MS following a repeat MALDI-IMS measurement of a small tissue region known to contain the peptide of interest (e.g. 1905.982).

Repeat sections 4 through 6 of this protocol using an FFPE section from the same patient block.

Once the measurement has completed and the data has been loaded in flexImaging, use the ROI tool to outline the region containing the peptide of interest, in this case the cancer region (see Figure 6a).

In the flexImaging interface tick the checkbox next the ROI that was just created (in the Regions window).

Right click the sum spectrum and change the display type to 2D All Scans. Zoom in on the m/z range of interest and confirm the presence of the peptide(s) of interest (see Figure 6b).

Apply a 0.5 Da mass filter to the mono-isotopic peak of the peptide and confirm the distribution is as expected (see Figure 6c).

Once the presence and distribution are confirmed, select the Show Single Spectrum tool in flexImaging and select a spectrum from the ROI which has relatively high intensity in the sum spectrum window and overlaps well with the sum spectrum isotopic peptide profile.

Right click on this tissue location and select Open Spectra in flexAnalysis.

In flexAnalysis, select Internal… in the Calibrate drop down menu. When the Internal Mass Calibration window appears, select the mass control list (.mcl) that contains the four internal calibrant peaks and one trypsin autolysis product (see section 6.8).

Set the User Defined Peak Assignment Tolerance to 300 ppm and the Mode to Quadratic. Click Automatic Assign and set the Zoom Range to ± 0.5%. Click on each calibrant found individually to zoom in and confirm correct assignment for all five peptides. If all five can not be found, select another spectrum from the same region.

Once calibration is confirmed, click Calibrate.

In the flexAnalysis Mass List find the peptide(s) of interest and right click. Select Add to MS/MS List. Right click again and select MS/MS List… to open the list. From the MS/MS list window, select Send to flexControl.

In flexControl, select a suitable LIFT method (see Table 1), then click the LIFT tab and select the peptide of interest from the MS/MS List drop down list. Once selected the instrument will automatically assign a Pre-Cursor Ion Selector (PCIS) Window Range.

Return to flexImaging and move the MALDI-TOF/TOF sample carrier (as in steps 6.3 and 6.4) to the locations on the tissue in which the peptide(s) of interest exists in high abundance.

NB: Use the non-normalized ion intensity maps for this as these provide the absolute peptide signal at all tissue locations.

Once the sample carrier has moved to the selected location, acquire a few hundred shots to collect between 3000 and 6000 counts for the parent mass.

It is important to zoom in on the parent and ensure that the profiles from multiple acquisitions overlap and are resolved almost to the baseline.

The sample carrier may need to be moved several times, as single locations will become exhausted of sample much faster than during LC-MALDI.

Once the parent has been measured, select Fragments in the LIFT tab. The laser power will automatically be adjusted by the software. Set the laser power to between 5 and 10% higher than the laser power required for the parent mass.

Begin measuring fragment spectra, increasing laser power and moving the sample carrier to adjacent areas of high peptide signal.

Upon sufficient fragmentation, select Save As… from the panel just below the MALDI CCTV display. In the window that appears, select a default LIFT processing method from the drop down list and tick the Open in flexAnalysis check box. Select a suitable data path and spectrum name (e.g. 1905_LIFT) and click Save.

Once the MS/MS spectrum is open in flexAnalysis and the processing script has finished, select ProteinScape from the Tools drop down menu.

NB: Spectrum can also be sent directly to BioTools.

In the ProteinScape window which appears, select Send selected Spectrum. A separate window will appear. Select the Export MALDI target tab and select the relevant project and sample name. Enter a name for the target and select Ok.


The peptide of interest mentioned above - AWM m/z 1905.981 - could be matched to heat shock protein beta 1 using both MASCOT database searches and by manual assignment of the sequence using BioTools Sequence Editor {Gustafsson, 2012 #1063}. Figure 6d and 6e compare the matching in situ MS/MS and LC-MALDI MS/MS fragments which were used to assign the peptides identity as Heat shock protein beta 1.

To submit a MASCOT search, in ProteinScape, open all the levels of the imported data (found under the selected project and sample). Right click on the CombinedLIFT level and select Protein Search...

Use the same search parameters as those for the LC-MALDI (step 9.5).

Select Start to start the database search. When complete, the peptide matches, if found, will be displayed.

Select the search result which appears next to the imported data (see step 9.8). If a positive match has been found, select Manual Validation in the Info tab of the Main View panel.

Compare the in situ MS/MS (Figure 6d) and LC-MS/MS (Figure 6e) spectrum to each other to confirm correct matching. If the initial match was correct there should be a high concordance between the ion series observed in both spectra.

If no match is possible the LC-MS/MS sequence can be directly compared to the in situ MS/MS using BioTools.

Send the in situ MS/MS spectrum from flexAnalysis to BioTools by selecting BioTools from the Tools drop down menu.

Once in BioTools, select the Start Sequence Editor icon.

In Sequence Editor, select File and New Sequence.

Enter the sequence of the tentatively identified peptide and modifications, if any.

Then select the Send to BioTools icon and return to BioTools.

The in situ MS/MS spectrum will now be annotated with any theoretical fragments from the entered peptide sequence.

Open the LC-MS/MS spectrum of interest (see step 9.13) in ProteinScape and select Manual Validation as before.

Compare the in situ MS/MS and LC-MS/MS in BioTools.

A match can be confirmed by the existence of unique MS/MS fragments in both the LC-MS/MS and in situ MS/MS spectra.


The in situ MS/MS to LC-MS/MS match in Figure 6d and 6e shows that the peptide identification by matching was correct, in this case, heat shock protein beta 1. Figure 6f shows the peptide sequence complete with fragments found in the LC-MS/MS spectrum (b and y ions).

Concluding remarks

Formalin-fixed tissues present a unique and rich resource for prospective biomarker and tissue classification studies {Groseclose, 2008 #363;Addis, 2009 #934}. Thus, much attention is currently focused on the capacity of MALDI-IMS to provide patient-specific and/or cancer-specific information from formalin-fixed archives and, in particular, tissue micro-arrays (TMAs) {Djidja, 2010 #792}. Dissemination of protocols, such as the one presented here, will hopefully lead to the formulation of an international standard MALDI-IMS method for clinical cohorts (e.g. TMAs). Furthermore, by application of the presented method, diagnostic or prognostic peptides can be identified using a combination of MALDI-TOF/TOF and LC-MS/MS: technologies which are readily available to many proteomic groups globally. Finally, the straightforward nature of the presented protocol ensures that groups new to MALDI-IMS can rapidly begin in situ characterization of formalin-fixed samples by MALDI-IMS. As a result, it will be possible for a growing number of laboratories to take advantage of the capacity to generate tissue-specific molecular data for application to clinical questions.


The authors would like to acknowledge Dr. Fergus Whitehead for annotating the ovarian cancer sections as well as Carmela Ricciardelli, Noor Lokman and Karina Martin for technical assistance with tissue collection and preservation by formalin-fixation. Martin Oehler and Peter Hoffmann acknowledge the support of the Australian Research Council (LP110100693) and Ovarian Cancer Research Foundation.