J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

ava i l ab l e a t www.sc i enced i r ec t . com

www.e l sev i e r . com/ loca te / j p ro t

JPROT-00374; No of Pages 16

Proteomic analysis of human cataract aqueous humour:Comparison of one-dimensional gel LCMS withtwo-dimensional LCMS of unlabelled andiTRAQ®-labelled specimens

Keiryn L. Bennetta,⁎, Marion Funkb, Marion Tschernuttera,b, Florian P. Breitwiesera,Melanie Planyavskya, Ceereena Ubaida Mohienc, André Müllera, Zlatko Trajanoskid,Jacques Colingea, Giulio Superti-Furgaa,⁎, Ursula Schmidt-Erfurthb,⁎aCenter for Molecular Medicine of the Austrian Academy of Sciences (CeMM), Vienna, AustriabDepartment of Ophthalmology, Medical University of Vienna, Vienna, AustriacInstitute for Genomics and Bioinformatics, Graz University of Technology, Graz, AustriadBiocenter, Division of Bioinformatics, Innsbruck Medical University, Innsbruck, Austria

A R T I C L E I N F O

⁎ Corresponding authors. Bennett and SupertiSciences (CeMM), Lazarettgasse 14, AKH BT 2Schmidt-Erfurth, Department of Ophthalmol40400 7931; fax: +43 1 40400 7932.

E-mail addresses: kbennett@cemm.oeawursula.schmidt-erfurth@meduniwien.ac.at (U

1874-3919/$ – see front matter © 2010 Elsevidoi:10.1016/j.jprot.2010.10.002

Please cite this article as: Bennett KL, edimensional gel LCMS with two-dimens

A B S T R A C T

Article history:Received 28 June 2010Accepted 5 October 2010

In this study, we report a comparative and quantitative analysis by mass spectrometry ofthe protein content of aqueous humour from cataract (control) patients. In addition toprotein profiling, the approach is layered with quantitative proteomics using the iTRAQ®methodology. Aqueous humour from ten clinically-matched patients was collected anddepleted of albumin and immunoglobulin G. Pairs of patient material were pooled anddivided into three aliquots for subsequent analysis by alternative proteomic approaches.Excluding keratin, trypsin, residual albumin and immunoglobulins, a total of 198 proteingroups were identified across the entire study. Relative protein quantitation with iTRAQ®revealed that 88% of the proteins had amaximal ±2-fold differential regulation between 3 ofthe 4 labelled samples, indicating minimal variation. The identified proteins werecategorised by gene ontology and one third of the proteins were annotated asextracellular. The major molecular functions of the proteins in aqueous humour arebinding (protein, metal ion, heparin, and DNA) and inhibition of proteolytic activity.Complementary to molecular function, the predominant biological processes for theproteins in aqueous humour are assigned to inflammatory and immune responses, andtransport.

© 2010 Elsevier B.V. All rights reserved.

Keywords:Aqueous humouriTRAQ®Orbitrap2D shotgunEye proteomics1D gel

-Furga are to be contacted at the Center for Molecular Medicine of the Austrian Academy of5.3, 1090 Vienna, Austria. Tel.: +43 1 40160 70001, +43 1 40160 70010; fax: +43 1 40160 970000.ogy, Medical University of Vienna, Währinger Gürtel 18-20, 1090 Vienna, Austria. Tel.: +43 1

.ac.at (K.L. Bennett), gsuperti@cemm.oeaw.ac.at (G. Superti-Furga),. Schmidt-Erfurth).

er B.V. All rights reserved.

t al, Proteomic analysis of human cataract aqueous humour: Comparison of one-ional LCMS of unlabelled..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

2 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

1. Introduction

One of the key developments that characterises the post-genomic era in biology is the ongoing experimental definitionof protein ensembles present in organisms, cells, organelles orextracellular body fluids [1–3]. When studies aimed towarddefining such protein complex ensembles [4] are performedquantitatively [5], an attempt can bemade to not only describethe protein component list of any given biological system, butalso to provide important quantitative information. To date,several body fluids have been analysed in depth by massspectrometric-basedmethods [6,7]. These include the vitreoushumour [8–10]; the tear [11], the urinary [12]; and the plasmaproteome [13]. Despite significant advances in this field,however, aqueous humour (AH) has not been thoroughlyinvestigated.

AH is the clear fluid filling the anterior segment (consistingof the anterior and posterior chambers of the human eye)between the corneal endothelium, the lens and the retrolentalcompartment (Fig. 1). AH is constantly generated by theepithelium of the ciliary body and is composed of a complexmixture of electrolytes, organic solutes and proteins. Thepermanent production of AH is compensated by drainagethrough the trabecular meshwork into Schlemm's canal andconveyance to the vascular system. AH is separated from thefirm, gel-like vitreous humour (VH) filling the posterior part ofthe eye by the lens and Zinn's membrane (Fig. 1). The vitreousgel consists of approximately 99% fluid, thus there is acontinuous exchange between the AH and VH derived froman identical ciliary origin. The major functions of AH are to: (i)supply the avascular tissues of the anterior segment with

Fig. 1 – Sagittal section of the human eye. The anterior chamber beand the firm, gel-like vitreous humour (VH) is located behind the

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

nutrients; (ii) remove metabolic waste from these oculartissues; (iii) participate in the immune response againstinvading pathogens; (iv) maintain the intraocular pressure ofthe eye; and (v) act as an antioxidant agent by transportingascorbic acid into the anterior segment [14–16].

Although AH is essential for the correct function of the eye,the protein content has not been characterised in detail byproteomic approaches and only a relatively small number ofpublications exist. In addition, not a single study could beidentified whereby quantitative information on the proteinsin AH was provided. The majority of the available studiesconducted on AH focused on the identification of specificproteins due to a disease state (uveal melanoma [17],glaucoma [18], myopia [19]) or acute corneal graft rejection[20]. There is one report in which AH samples from patientswith various ocular pathologies were analysed by on-linecapillary electrophoresis mass spectrometry [21], however,only average molecular masses were used to tentativelyidentify the proteins. Only four publications could be locatedin the literature on mass spectrometric profiling of AH andsubsequent protein identification [22–25]. The study byStastna et al. [22] was conducted on the AH from rabbits andis therefore difficult to correlate with human AH.

Richardson et al. [23] pooled AH from 12 cataract patients,depleted the combined sample of serum albumin andimmunoglobulin G, and analysed both the depleted andbound AH fractions by multi-dimensional protein identifica-tion technology (MudPIT). Fifty and 12 high-confidenceproteins (≥2 unique peptides per protein) were identified inthe depleted and bound AH fractions, respectively. Also viatheMudPIT technology, Marinach-Patrice et al. [24] comparedthe AH proteome from individual patients with pooled

tween the cornea and lens contains the aqueous humour (AH)lens in the posterior part of the eye.

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

3J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

samples. All analyses were performed on undepleted AH andoverall 71 proteins (≥1 unique peptides per protein) wereidentified.

The manuscript by Fautsch et al. is to date, the mostimpressive analysis of human AH. A total of eighty fivepatients were used in the study and AH samples were pooledinto three age-matched groups. The top 6 abundant proteinswere depleted from the AH pools and in total, 349 proteins (≥2unique peptides per protein, excluding immunoglobulins andresidual proteins from incomplete depletion) were identifiedby gel-based LCMS. The group complemented the proteomicdata with additional studies on low-abundance proteins byquantitative antibody-based protein arrays.

The aims of this study were to: (i) analyse the proteinquantity equivalent to that obtainable from the AH of a single

Fig. 2 – Overview of the methodology used in this study to charapatients. The number of protein groups identified by the three p1D-G-CID (one-dimensional gel collision-induced dissociation), 2dissociation) and 2D-SG-HCD (two-dimensional shotgun high-en

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

individual; (ii) assess the feasibility of labelling AH withiTRAQ® for relative quantitation; and (iii) establish cataractpatients as control subjects for subsequent studies on ocularpathologies. Usingmass spectrometry in conjunction with on-line and off-line high-performance liquid chromatography(HPLC) (Fig. 2), 198 protein groups (excluding keratin, trypsin,residual albumin and immunoglobulins) were unambiguouslyidentified in human AH (≥2 unique peptides per protein).Relative protein quantitation with the iTRAQ® methodologyrevealed that the majority of the identified proteins werepresent in three of the four samples at a maximum ±2-folddifferential regulation with a 95% confidence. The proteinsidentified in this study were analysed using the gene ontologyclassification [26] with respect to (i) subcellular localisation; (ii)molecular function; and (iii) biological processes.

cterise the proteome of human AH from control (cataract)roteomic approaches is shown. PI (protease inhibitors),D-SG-CID (two-dimensional shotgun collision-inducedergy collision-induced dissociation).

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

4 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

2. Materials and methods

2.1. Materials

Four-times concentrated Laemmli buffer (containing 10% β-mercaptoethanol), iodoacetamide, dithiothreitol, 1 M triethy-lammonium bicarbonate (TEAB), formic acid, protease inhib-itor cocktail (SIGMA-Aldrich, St. Louis, MO); NuPage 4–12%Bis–Tris gradient gels, NuPage MOPS SDS buffer and NuPageantioxidant (Invitrogen, Carlsbad, CA); trypsin (Promega Corp.,Madison, WI).

2.2. Surgical sample collection and preparation

Patients were recruited at the Department of Ophthalmologyat the Medical University of Vienna over an approximate6 week timeframe from January to March 2009. Informedwritten consent was obtained from all patients. The mean agewas 67.5 years (range 46–78 years) and the group was com-prised of 6 men and 4 women (Table 1). The study wasapproved by the Ethics Committee of theMedical University ofVienna, registered at the European clinical database(EUDRACT-2006-005684-26) and followed the tenets of theHelsinki protocol. AH was obtained from patients undergoingcataract surgery. Exclusion criteria were: any type of eyedisease (excluding cataract), previous eye surgery of any kind(including vitrectomy), laser coagulation, diabetes mellitus,use of immunosuppressive drugs, malignant tumours, andparticipation in any study of investigational drugs within a3 month timeframe preceding surgery. AH was obtained aftertopical anaesthesia and disinfection immediately prior to

Table 1 – Information on the 10 patients enrolled in this study.obtain 30 μg protein are included. Note that CAT10 and CAT11female; M, male.

SamplePool

Identifier Cataractclassification

Gender Age(YR)

AHtotal

volum(μL)

1 CAT1 Cataractacorticonuclearis

F 58 150

CAT12 Cataracta nucleariset subcapsularis

F 64 140

2 CAT14 Cataractacorticonuclearis

M 66 150

CAT16b Cataracta nucleariset subcapsularis

M 76 120

3 CAT19 Cataractacorticonuclearis

M 78 100

CAT21 Cataractacorticonuclearis

M 78 120

4 CAT17 Cataractacorticonuclearis

M 74 100

CAT22 Cataracta nucleariset subcapsularis

M 69 230

5 CAT10 Cataracta nucleariset subcapsularis

F 66 120

CAT11 Cataracta subcapsularisposterior

F 46 100

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

routine cataract surgical intervention. All surgical procedureswere performed by experienced ophthalmologists and nocomplications were encountered. Paracentesis was performedwith minimal contact to other intraocular structures, thuspreventing the potential introduction of contaminating pro-teins. Approximately 100–150 μL AH fluid was obtained bylimbal paracentesis of the anterior chamber using a 30 Gophthalmic cannula attached to an insulin syringe. Immedi-ately after collection, AH samples were transferred to dust-free Eppendorf tubes containing 3 μL 50× protease inhibitorcocktail, carefully mixed and stored at 4 °C for a maximum of4 h. Samples were centrifuged at 1600×g for 15 min at 4 °C in arefrigerated centrifuge (Eppendorf, Hamburg, Germany) topellet any intact cells. The supernatant was removed andcentrifuged at 16,000×g for 15 min at 4 °C to remove anycellular organelles and debris from apoptotic cells. Thesupernatant was removed and stored at −80 °C until required.

2.3. Protein quantitation and depletion

Total protein content of each of the ten AH samples wasdetermined with Qubit™ fluorometer and Quant-iT™ proteinassay kit according to the instructions provided by themanufacturer. Depletion of AH fluid was performed withProteoPrep® immunoaffinity albumin and IgG depletion kits(SIGMA-Aldrich, St. Louis, MO) according to the instructionsprovided, with the following exception. For compatibility withiTRAQ®derivatisation, 50mMTEAB (pH adjusted to 7.2with 5%formic acid) was used to elute the non-bound proteins from thecolumn. In order to compare data sets generated by differenttechnological approaches, pairs of the depleted patient sampleswere pooled, mixed, and split into three equal portions.

Total protein concentration plus normalised AH volumes tohad 27 μg and 21 μg, respectively (shown in bold face). F,

e

Averageprotein

concentration(ng/μL)

Totalprotein(μg)

Normalisedvolume (μL)(30 μg Totalprotein)

PercentageAH used

(%)

iTRAQ®Label

567±3 85.05 53 35 n/a

394±6 55.16 76.2 54

353±8 52.95 85 57 114

262±1 31.44 114.6 96

392±5 39.2 76.6 77 115

318.5±2.5 38.22 94.2 79

490.5±23.5 49.05 61.2 61 116

410±9 94.3 73.2 32

223.5±0.5 26.82 120 100 117

212±9 21.2 100 100

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

5J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

2.4. One-dimensional SDS-PAGE, silver staining and insitu tryptic digestion

The ocular fluid was mixed with 4× Laemmli buffer (contain-ing 10% β-mercaptoethanol), boiled for 3 min, cooled to roomtemperature and alkylated by incubation with iodoacetamide(final concentration, 13 μg/μL) for 20 min in the dark. Reducedand alkylated samples were separated by 1D-SDS-PAGE on a4–12% bis–Tris gel (NuPAGE, Invitrogen, Carlsbad, CA). Follow-ing visualisation of the proteins by silver staining [27], entirelanes were sliced into 20 pieces. Proteins in the gel slices werereduced with dithiothreitol, alkylated by incubation withiodoacetamide and digested in situ with modified porcinetrypsin [27] (Promega Corp., Madison, WI). The resultantpeptidemixturewas extracted from the gel slices and desaltedwith customised reversed-phase stage tips [28]. The volume ofthe eluted sample was reduced to approximately 2 μL in avacuum centrifuge and reconstituted to 10 μL with 5% formicacid. Depending on the intensity of the protein band staining,additional multiples of 8 μL 5% formic acid were added tospecific samples prior to analysis by LCMS.

2.5. Solution tryptic digestion and iTRAQ® derivatisation

Samples were reduced with dithiothreitol, alkylated byincubation with iodoacetamide and digested with modifiedporcine trypsin (Promega Corp., Madison, WI). Four of the fivesamples were chosen for derivatisation with 4-plex iTRAQ®reagent (ABI, Framingham, MA) [29] and labelled according tothe instructions provided.

2.6. Reversed-phase reversed-phase (RPRP) separation [30]

Tryptic digests were concentrated and purified by solid phaseextraction (SPE) (UltraMicroSpin columns 3–30 μg capacity,Nest Group Inc., Southboro, MA, USA) prior to injection onto aPhenomenex column (150×2.0 mm Gemini-NX 3 μm C18110 Å, Phenomenex, Torrance, CA, USA) on an Agilent 1200series HPLC (Agilent Biotechnologies, Palo Alto, CA) with UVdetection at 214 nm. HPLC solvent A consisted of 20 mMNH4OH pH 10.5 in 5% acetonitrile and solvent B consisted of20 mM NH4OH pH 10.5 in 90% acetonitrile. Peptides wereseparated at 35 °C with a flow rate of 100 μL/min and elutedfrom the columnwith a 41 min gradient ranging from 0 to 35%solvent B, followed by a 4 min gradient from 35 to 70% solventB and, finally, a 2 min gradient from 70 to 100% solvent B. Fiftyone time-based fractions were collected, acidified, and pooledinto 20 or 40 fractions. The sample volume was reduced toapproximately 2 μL in a vacuum centrifuge and reconstitutedto 10 μL with 5% formic acid.

2.7. Liquid chromatography and mass spectrometry

Mass spectrometry was performed on a hybrid LTQ OrbitrapXL mass spectrometer (ThermoFisher Scientific, Waltham,MA) using the Xcalibur version 2.0.7 coupled to an Agilent 1200HPLC nanoflow system (dual pump system with one pre-column and one analytical column) (Agilent Biotechnologies,Palo Alto, CA) via a nanoelectrospray ion source using liquidjunction (Proxeon, Odense, Denmark). Solvents for LCMS

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

separation of the digested samples were as follows: solventA consisted of 0.4% formic acid in water and solvent Bconsisted of 0.4% formic acid in 70% methanol and 20%isopropanol. From a thermostatted microautosampler, 8 μL ofthe tryptic peptide mixture was automatically loaded onto atrap column (Zorbax 300SB-C18 5 μm, 5×0.3 mm, AgilentBiotechnologies, Palo Alto, CA) with a binary pump at a flowrate of 45 μL/min. 0.1% TFA was used for loading and washingthe precolumn. After washing, the peptides were eluted byback-flushing onto a 16 cm fused silica analytical columnwithan inner diameter of 50 μm packed with C18 reversed-phasematerial (ReproSil-Pur 120 C18-AQ, 3 μm, Dr. Maisch GmbH,Ammerbuch-Entringen, Germany). The peptides were elutedfrom the analytical column with a 27 min gradient rangingfrom 3 to 30% solvent B, followed by a 25 min gradient from 30to 70% solvent B and, finally, a 7 min gradient from 70 to 100%solvent B at a constant flow rate of 100 nL/min. The analyseswere performed in a data-dependent acquisitionmode using atop 6 collision-induced dissociation (CID) method for peptideidentification alone; or a top 3 high-energy collision-induceddissociation (HCD) method for peptide identification plusrelative quantitation of iTRAQ® reporter ions. Dynamicexclusion for selected ions was 60 s. No lock masses wereemployed. Maximal ion accumulation time allowed on theLTQ Orbitrap in CID mode was 150 ms for MSn in the LTQ and1000 ms in the C-trap. Automatic gain control was used toprevent overfilling of the ion traps and were set to 5,000 (CID)in MSn mode for the LTQ, 106 ions for a full FTMS scan and 105

ions for HCD. Maximum ion time for HCD was set to 1000 msfor acquiring 1 microscan at a resolution of 7500. Injectionwaveforms were activated for both LTQ and Orbitrap. Intactpeptides were detected in the Orbitrap at 100,000 resolutionfor CID fragmentation and 60,000 for HCD fragmentationexperiments. The threshold for switching from MS to MSMSwas 2000 counts.

2.8. Data analysis

The acquired data were processed with Bioworks v3.3.1 SP1(ThermoFisher, Waltham, MA, USA), dta files merged with aninternally-developed program, and searched against thehuman SwissProt database version v57.4 (34,579 sequences,including isoforms as obtained from varsplic.pl) with thesearch engines MASCOT (v2.2.03, MatrixScience, London, UK)and Phenyx (v2.5.14, GeneBio, Geneva, Switzerland) [31].Submission to the search engines was via a Perl script thatperforms an initial search with relatively broad mass toler-ances (MASCOT only) on both the precursor and fragment ions(±10 ppm and ±0.6 Da, respectively). High-confidence peptideidentifications are used to recalibrate all precursor andfragment ion masses prior to a second search with narrowermass tolerances (±4 ppm and ±0.3 Da for CID and ±4 ppm and±0.005 Da for HCD, respectively). One missed tryptic cleavagesite was allowed. Carbamidomethyl cysteine and iTRAQ® 4-plex (N-terminii and lysine) were set as fixed modifications,and oxidised methionine was set as a variable modification.For the two-dimensional shotgun samples, all .dta files fromthe individual analyses were merged into a single .mgf fileprior submission to the search engines. To validate theproteins, MASCOT and Phenyx output files were processed

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

6 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

by internally-developed parsers. For MASCOT, two uniquepeptides with an ion score >18 (plus additional peptides fromproteins fulfilling the criteria with an ion score >10) arerequired. For Phenyx, two unique peptides with a z-score >4.5and a P-value <0.001 are required (plus additional peptidesfrom proteins fulfilling the criteria with a z-score >3.5 and a P-value <0.001).

The validated proteins retrieved by the two algorithms aremerged, any spectral conflicts discarded and grouped accord-ing to shared peptides. A false positive detection rate (FDR) of<0.25% and <0.1% (including the peptides exported with lowerscores) was determined for proteins and peptides, respective-ly, by applying the same procedure against a reverseddatabase. Comparisons between analytical methods involvedcomparisons between the corresponding sets of identifiedproteins. This was achieved by an internally-developedprogram that simultaneously computes the protein groupsin all samples and extracts statistical data such as the numberof distinct peptides, number of spectra, and sequencecoverage.

2.9. iTRAQ® relative quantitation

Identified peptides that are unique to a specific protein wereused to determine relative quantitation of a protein betweenthe four samples. The intensities of the iTRAQ® reporter ionm/z values (114.1112, 115.1083, 116.1116 and 117.1150) forthese peptides were extracted from the centroid data in themerged .mgf peak lists. The computation and analysis of therelative iTRAQ® ratios were performed with an internally-developed Perl script. The data was corrected for reagentimpurity (according to the values supplied by the manufac-turer). By using a sliding window approach, noise levels werecalculated relative to the retention time and subtracted. OnlyiTRAQ® reporter ion intensities>2000 counts were consideredfor the calculation of the ratios. Normalisation of the iTRAQ®signals was achieved by scaling to an equal sum of intensitiesin each channel. For the final relative quantitation of a protein,the median ratio of the relative intensities of the reporter ionsof all (specific) spectra for a protein is used.

3. Results and discussion

3.1. Total protein content in aqueous humour fromcataract (control) patients

Themean total protein content in the AH from 10 patients was0.36 mg/mL (range 0.21–0.57 mg/mL) (Table 1). Prior to deple-tion of albumin and immunoglobulin G, the total proteincontent for eight of the samples was normalised to 30 μg(Table 1). For two of the cataract patients (CAT10 and CAT11),the total protein of the entire AH sample was below this value(27 and 21 μg, respectively). Nonetheless, these patients wereincluded in the study to ascertain if any changes could beobserved in these suboptimal samples. Following depletion,the quantity of protein remaining in all 10 samples was verylow and an accurate protein determination could not beperformed without consuming the majority of the fluid. Thus,we made the assumption that 80% of the total protein was

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

depleted across all samples (according to the specifications ofthe manufacturer). In the study by Richardson et al. [23], thetotal protein from the depleted and bound AH fractions (12pooled samples) was 116 μg and 221 μg, respectively. Theprotein concentration of the patient pool prior to depletion isnot given. Assuming no losses, however, this indicates that66% of the proteins are removed from AH during depletion.Thus our assumption of 80% depletion is not completelyunreasonable.

Depleted pair-wise samples were pooled as indicated inTable 1, and split into three equal portions (i.e., equivalent to20 and 16 μg protein prior to depletion for pools 1 to 4 and pool5, respectively) for subsequent protein or peptide fractionationand analysis by LCMS. The rationale behind pooling twopatient samples and dividing into three equal aliquots, wasthat the protein concentration of each of the ten individualsamples was insufficient for the three intended proteomicapproaches. By pooling and dividing samples, the total proteincontent was sufficiently increased to perform the differentanalyses. Additionally, this allowed direct comparison of thegenerated data sets and provided a similar quantity of totalprotein that would be obtained from an individual patient.

The strategy in this study and in recent reports [23,25] wasto remove abundant proteins from AH. The debate over thebenefits of protein depletion versus loss of proteins bound to,e.g., albumin, however, continues to rage. We (unpublishedobservations) and others [32] have shown that the net gain inthe number of proteins identified is greater in depletedsamples than the loss of proteins via suppression effects inundepleted samples. Richardson et al. [23] also revealed thatLC–MSMS analysis of the depleted fraction from cataract AHconsisted primarily of abundant proteins. Most of the proteinsidentified in the bound fraction were present at high levels inthe depleted samples indicating that the loss during removalof serum albumin and immunoglobulin G was minimal.

3.2. One-dimensional gel-based protein separation

Albumin- and immunoglobulin G-depleted AH from one thirdof the five pooled samples were analysed by 1D-SDS-PAGE.Assuming 80% protein depletion from the normalised totalprotein input of 20 μg (Pools 1–4) and 16 μg (Pool 5), only 4 and3.2 μg total protein were loaded per lane. As can be seen by therepresentative diagram in Fig. 3, there is an overall lowabundance of protein and only seven bands are evident. Notangible difference in the staining pattern was observed forPool 5 (data not shown). As determined from spectral counts(Supplementary Table S1; 1D-G-CID), the major proteinsidentified in these regions were (A) serotransferrin (TRFE); (B)residual albumin from incomplete depletion; (C) α1-antitryp-sin (A1AT); (D) pigment epithelial-derived factor (PEDF); (E)apolipoprotein A-1 (APOA1); (F) immunoglobulin light chain;and (G) transthyretin (TTHY). The remainder of the gel is closeto background staining. Nonetheless, the entire lanes weresliced into twenty pieces, the proteins in each slice digested insitu with trypsin and analysed by LCMS. Excluding trypsin,keratin, immunoglobulins and residual albumin, 156 proteingroups were identified across the five gel-based analyses(Table 2 1D-G-CID; Groups 1, 2, 3 and 5). TRFE was the mostabundant protein in the depleted fluids with a maximum

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

Fig. 3 – Representative 1-dimensional SDS-PAGE image of theproteins present in albumin- and immunoglobulinG-depleted aqueous humour from control (cataract) patients.Prior to depletion, the samples were normalised on totalprotein content to 20 μg (Pools 1–4) and 16 μg (Pool 5).Assuming depletion efficiency of 80%, the total quantity ofprotein loaded per lane was 4 μg (Pools 1–4) and 3.2 μg(Pool 5). The bands were identified as (A) TRFE; (B) residualalbumin from incomplete depletion; (C) A1AT; (D) PEDF;(E) APOA1; (F) immunoglobulin light chain; and (G) TTHY .

7J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

sequence coverage of 68% (55 unique peptides identified from8318 spectra); followed by A1AT (55%, 26 unique peptidesidentified from 1678 spectra); and complement C3 (CO3) (48%,85 unique peptides identified from 1447 spectra). The latter,however, is not visually apparent on the silver-stained gel (Mr

188 kDa).The data set generated from gel-based proteomic analyses

performed by Fautsch et al. [25] resulted in the identification of349 proteins (excluding immunoglobulins, residual albuminand transferrin) from 30 (n=2) and 25 (n=1) pooled AHsamples (serum albumin-, haptoglobin-, antitrypsin-, trans-ferrin-, immunoglobulin G- and A-depleted). In the currentstudy, 156 protein groups were identified from one third of theAH from 2 (n=5) pooled samples (albumin- and IgG-depleted).

3.3. Liquid-based peptide separation (2-dimensionalshotgun CID fragmentation)

One third of the peptide mix derived from the tryptically-digested pooled samples (n=5) were firstly separated andfractionated into twenty off-line samples at pH 10, followed bya second dimension separation at pH 2 by on-line LCMS.

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

Twenty fractions were collected to be comparable to the 20slices that were excised from the gel-based experiments in theprevious section. Similar to the gel-based analyses, 4 μg (Pools1–4) and 3.2 μg (Pool 5) total protein were injected per off-lineanalysis. Shown in Fig. 4, is a representative chromatogram ofthe first dimension. As determined from spectral countabundance (Supplementary Table S1; 2D-SG-CID), the peptidesidentified in the peakswith high UV absorbancewere from: (A)CO3, prostaglandin-H2 D-isomerase (PTGDS), PEDF, TRFE andceruloplasmin (CERU); (B) TRFE and haptoglobin (HPT); (C)TRFE; (D) PTGDS and PEDF; (E) TRFE, PEDF, CERU, complementfactor B (CFAB) and neurosecretory protein VGF (VGF); (F) TRFE;and (G) complement factor C4-A (CO4A), A1AT, α1-antic-hymotrypsin (AACT), α2-macroglobulin (A2MG) and CO3.Excluding trypsin, keratin, residual albumin and immunoglo-bulins, 117 protein groups (c.f., 156 1D-G-CID) were identifiedacross the five liquid-based analyses fragmented by CID(Table 2 2D-SG-CID; Groups 1, 2, 3 and 5). As for the gel-based approach, TRFE was again the most abundant proteinwith a maximum sequence coverage of 66% (58 uniquepeptides identified from 19,713 spectra); followed by PEDF(43%, 18 unique peptides identified from 3393 spectra); andA1AT (60%, 24 unique peptides identified from 1967 spectra).CO3 was the fourth most abundant protein identified via thisproteomic approach. Four of the five samples were taintedwith haemoglobin (HBA and HBB) although no visible con-tamination was observed during sample collection. Thus weconcluded that the proteomic analyses of the AH samples inthis study were not overly compromised by a major influx ofproteins from blood. Rather the proteins identified are a truereflection of the protein content of AH alone.

Comparison of the data sets obtained from the gel-basedand 2D ‘shotgun’ (CID fragmentation) methods revealed that90 of the protein groups identified were common to bothapproaches (Table 2; Groups 1 and 2; Fig. 5) and there were 27protein groups present in the fractionated samples that werenot identified from the gels (Table 2; Groups 4 and 6; Fig. 5).The data set generated by Richardson et al. [23] with acomparative 2D-SG-CID approach resulted in the identifica-tion of 50 proteins from 12 pooled AH samples (serumalbumin- and immunoglobulin G-depleted). Marinach-Patriceet al. [24] also used 2D-SG-CID to identify a total of 71 proteinsfrom single and pooled undepleted AH samples. In the currentstudy, 117 protein groups were identified from one third of theAH from 2 pooled samples (albumin- and IgG-depleted).

3.4. Liquid-based peptide separation (2-dimensionalshotgun HCD fragmentation)

Pools 2 to 5 (Table 1) were chosen for derivatisation with theiTRAQ® reagent. Pools 2 to 4 and Pool 5 contained 4 μg and3.2 μg total protein, respectively. The rationale for includingone of the suboptimal samples was to ascertain the necessityof normalising protein input for iTRAQ labelling and determi-nation of relative protein ratios between samples. Followingmodification and pooling of the labelled samples, the resul-tant mixture was split into two equal portions and thepeptides were separated and fractionated as described inMaterials and methods. Note that in these experimentsapproximately twice as much total protein was fractionated

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

Table 2 – Overview of the proteins identified in this study. Proteins are grouped according to technology.Maximumand averagesequence coverage (SCV) between the different pair-wise patient poolswith standard deviation are compared. The global SCV isalso included, summarisedacross the threeproteomic approaches. Proteinsuniquely identified in this studyaredenotedwith anasterisk. 1D-G-CID (one-dimensional gel collision-induceddissociation), 2D-SG-CID (two-dimensional shotgun collision-induceddissociation) and 2D-SG-HCD (two-dimensional shotgun high-energy collision-induced dissociation).

Gene symbol 1D-G-CID 2D-SG-CID 2D-SG-HCD Global SCV

Total Mean SD Total Mean SD Total Mean SD

Group 1: 1D-G-CID+2D-SG-CID+2D-SG-HCDTRFE 0.68 0.46 0.14 0.66 0.57 0.06 0.65 0.62 0 0.78PEDF 0.61 0.38 0.16 0.43 0.33 0.04 0.4 0.37 0.01 0.62A1AT 0.55 0.46 0.07 0.6 0.44 0.06 0.3 0.25 0.03 0.67CO3 0.48 0.32 0.14 0.4 0.25 0.06 0.26 0.21 0 0.56PTGDS 0.31 0.27 0.03 0.5 0.37 0.05 0.37 0.37 0 0.54HEMO 0.42 0.23 0.11 0.39 0.25 0.07 0.26 0.22 0.04 0.46CERU 0.3 0.19 0.09 0.27 0.18 0.03 0.11 0.11 0.01 0.34VTDB 0.55 0.21 0.17 0.56 0.36 0.1 0.35 0.31 0.03 0.7A1AG1 0.41 0.27 0.14 0.39 0.3 0.07 0.26 0.25 0.02 0.45APOA1 0.74 0.56 0.21 0.59 0.44 0.14 0.47 0.43 0.04 0.75A2MG 0.4 0.19 0.12 0.23 0.14 0.04 0.12 0.09 0.02 0.42CFAB 0.3 0.21 0.06 0.32 0.22 0.05 0.15 0.13 0.04 0.33AACT 0.34 0.23 0.04 0.24 0.19 0.01 0.2 0.15 0.01 0.37IC1 0.31 0.24 0.02 0.2 0.14 0.04 0.13 0.13 0 0.33CO4A 0.26 0.15 0.05 0.24 0.11 0.06 0.17 0.13 0.01 0.36ANT3 0.42 0.32 0.05 0.38 0.22 0.04 0.28 0.25 0.01 0.52A1BG 0.29 0.17 0.07 0.33 0.19 0.07 0.26 0.26 0 0.34HPT 0.4 0.2 0.12 0.36 0.21 0.07 0.15 0.12 0.02 0.47TTHY 0.69 0.34 0.22 0.5 0.24 0.15 0.22 0.22 0 0.73A1AG2 0.33 0.17 0.13 0.22 0.19 0.02 0.11 0.08 0.01 0.37A2GL 0.27 0.16 0.1 0.22 0.14 0.05 0.13 0.13 0 0.31CYTC 0.31 0.22 0.05 0.31 0.26 0.06 0.23 0.18 0.08 0.35FETUA 0.18 0.09 0.06 0.19 0.13 0.04 0.11 0.1 0.01 0.27B2MG 0.29 0.22 0.12 0.24 0.23 0.02 0.22 0.22 0 0.29ZA2G 0.58 0.31 0.19 0.49 0.25 0.09 0.25 0.18 0.09 0.61OSTP 0.16 0.06 0.08 0.28 0.18 0.06 0.14 0.14 0 0.34OSTP 0.14 0.05 0.07 0.19 0.15 0.03 0.18 0.15 0.04 0.3CH3L1 0.44 0.25 0.16 0.34 0.2 0.07 0.21 0.16 0.08 0.52NPC2 0.36 0.19 0.15 0.28 0.21 0.12 0.15 0.08 0.11 0.36ENPP2 0.22 0.13 0.05 0.15 0.06 0.04 0.07 0.05 0.04 0.29APOH 0.23 0.06 0.1 0.29 0.15 0.1 0.12 0.09 0.03 0.34AFAM 0.15 0.06 0.04 0.26 0.14 0.03 0.16 0.1 0.02 0.36HRG 0.1 0.03 0.03 0.22 0.13 0.04 0.14 0.11 0.03 0.22ANGT 0.15 0.1 0.02 0.15 0.07 0.05 0.05 0.05 0 0.19CLUS 0.2 0.1 0.09 0.23 0.13 0.06 0.11 0.11 0 0.23GPX3 0.21 0.15 0.09 0.22 0.13 0.08 0.22 0.19 0.05 0.35CRBS 0.24 0.18 0.1 0.26 0.14 0.1 0.16 0.14 0.03 0.33DKK3 0.1 0.08 0.04 0.27 0.07 0.1 0.09 0.09 0 0.3AMBP 0.26 0.08 0.11 0.23 0.12 0.05 0.1 0.08 0.03 0.28PLMN 0.18 0.08 0.06 0.17 0.06 0.05 0.05 0.02 0.03 0.26CSTN1 0.15 0.08 0.02 0.13 0.05 0.04 0.06 0.03 0.02 0.2APOA4 0.4 0.18 0.15 0.32 0.11 0.11 0.13 0.09 0.02 0.49KNG1 0.23 0.08 0.09 0.22 0.09 0.06 0.11 0.08 0.01 0.32RET3 0.15 0.07 0.06 0.12 0.04 0.03 0.03 0.03 0.01 0.19SAP 0.09 0.04 0.03 0.05 0.03 0.02 0.05 0.05 0 0.09APOE 0.41 0.16 0.16 0.24 0.09 0.08 0.07 0.04 0.05 0.5ENOA 0.27 0.11 0.11 0.27 0.14 0.04 0.05 0.02 0.03 0.37CBG 0.1 0.03 0.03 0.11 0.07 0.02 0.09 0.05 0.06 0.17GELS 0.16 0.09 0.07 0.08 0.02 0.03 0.05 0.02 0.03 0.18SEM7A 0.05 0.02 0.02 0.15 0.05 0.03 0.11 0.08 0.01 0.19BGH3 0.16 0.06 0.07 0.09 0.03 0.04 0.04 0.02 0.03 0.17CFAI 0.25 0.06 0.08 0.11 0.02 0.05 0.08 0.06 0.03 0.32BTD 0.08 0.04 0.02 0.04 0.01 0.02 0.05 0.05 0 0.1OPT 0.14 0.04 0.06 0.11 0.04 0.06 0.12 0.09 0.04 0.19LG3BP 0.1 0.04 0.04 0.08 0.02 0.03 0.03 0.02 0.02 0.16SODC 0.23 0.05 0.1 0.25 0.07 0.09 0.1 0.05 0.07 0.33LUM 0.07 0.03 0.04 0.05 0.01 0.02 0.07 0.03 0.05 0.16FSTL1 0.07 0.01 0.03 0.09 0.02 0.04 0.06 0.03 0.04 0.18

8 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

Please cite this article as: Bennett KL, et al, Proteomic analysis of human cataract aqueous humour: Comparison of one-dimensional gel LCMS with two-dimensional LCMS of unlabelled..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

Table 2 (continued)

Gene symbol 1D-G-CID 2D-SG-CID 2D-SG-HCD Global SCV

Total Mean SD Total Mean SD Total Mean SD

Group 2: 1D-G-CID+2D-SG-CIDAPOA2 0.21 0.21 0 0.21 0.21 0 0 0 0 0.21HBB 0.56 0.23 0.22 0.31 0.16 0.1 0 0 0 0.56APLP2 0.16 0.06 0.06 0.12 0.05 0.05 0 0 0 0.2HORN 0.07 0.02 0.03 0.06 0.01 0.02 0 0 0 0.09BP7 0.26 0.08 0.11 0.31 0.16 0.1 0 0 0 0.43FAM3C 0.35 0.2 0.12 0.17 0.04 0.06 0 0 0 0.38RNAS1 0.25 0.1 0.14 0.12 0.02 0.05 0 0 0 0.25RBP4 0.23 0.09 0.09 0.19 0.07 0.09 0 0 0 0.24ITIH2 0.11 0.04 0.05 0.08 0.02 0.02 0 0 0 0.14DESP 0.05 0.02 0.02 0.01 0 0 0 0 0 0.05PGBM 0.03 0.01 0.01 0.02 0.01 0.01 0 0 0 0.03NRCAM 0.09 0.04 0.03 0.02 0 0.01 0 0 0 0.1DCD 0.23 0.13 0.12 0.13 0.03 0.06 0 0 0 0.23ITIH1 0.08 0.03 0.03 0.05 0.01 0.02 0 0 0 0.1ITIH4 0.12 0.04 0.05 0.04 0.01 0.01 0 0 0 0.14IBP6 0.1 0.02 0.04 0.13 0.04 0.06 0 0 0 0.13TPIS 0.2 0.11 0.07 0.15 0.04 0.06 0 0 0 0.25DPP2 0.04 0.01 0.02 0.15 0.03 0.05 0 0 0 0.15A4 0.08 0.02 0.04 0.04 0.01 0.02 0 0 0 0.09CATB 0.08 0.02 0.03 0.1 0.02 0.04 0 0 0 0.12ATRN 0.05 0.02 0.02 0.01 0 0.01 0 0 0 0.05APOD 0.23 0.08 0.08 0.13 0.03 0.06 0 0 0 0.23NCAM1 0.05 0.02 0.02 0.04 0.01 0.02 0 0 0 0.09CATZ 0.14 0.05 0.05 0.07 0.01 0.03 0 0 0 0.14B3GN1 0.06 0.01 0.03 0.07 0.01 0.03 0 0 0 0.12CADH2 0.04 0.02 0.02 0.03 0.01 0.01 0 0 0 0.05AL3A1 0.05 0.01 0.02 0.06 0.01 0.03 0 0 0 0.08FBLN3 0.04 0.01 0.02 0.06 0.01 0.03 0 0 0 0.09COL12* 0.03 0.01 0.01 0.04 0.01 0.02 0 0 0 0.04CFAH 0.02 0 0.01 0.01 0 0.01 0 0 0 0.03KAIN 0.07 0.01 0.03 0.1 0.03 0.04 0 0 0 0.15SPON1 0.04 0.01 0.02 0.02 0 0.01 0 0 0 0.05

Group 3: 1D-G-CID+2D-SG-HCDCAH3 0.19 0.08 0.05 0 0 0 0.15 0.1 0.03 0.23THBG 0.13 0.05 0.05 0 0 0 0.05 0.04 0 0.18SODE 0.23 0.07 0.09 0 0 0 0.1 0.05 0.07 0.23

Group 4: 2D-SG-CID+2D-SG-HCDTHRB 0 0 0 0.11 0.05 0.04 0.07 0.05 0.02 0.14

Group 5: 1D-G-CIDXLRS1 0.19 0.1 0.07 0 0 0 0 0 0 0.19CATD 0.35 0.11 0.15 0 0 0 0 0 0 0.35HBA 0.3 0.15 0.14 0 0 0 0 0 0 0.3UBIQ 0.58 0.29 0.28 0 0 0 0 0 0 0.58CO9A2 0.04 0.03 0.02 0 0 0 0 0 0 0.04ACTB 0.23 0.06 0.1 0 0 0 0 0 0 0.23CAH2 0.24 0.11 0.08 0 0 0 0 0 0 0.24LYSC 0.33 0.09 0.15 0 0 0 0 0 0 0.33CATL1 0.12 0.05 0.05 0 0 0 0 0 0 0.12ITIH5* 0.05 0.02 0.02 0 0 0 0 0 0 0.05CRGC 0.15 0.08 0.05 0 0 0 0 0 0 0.15PEBP1 0.27 0.16 0.11 0 0 0 0 0 0 0.27LRP2 0.02 0 0.01 0 0 0 0 0 0 0.02CRP 0.16 0.08 0.06 0 0 0 0 0 0 0.16CNTN1 0.09 0.02 0.04 0 0 0 0 0 0 0.09LDHA 0.16 0.06 0.06 0 0 0 0 0 0 0.16CRBB2 0.1 0.04 0.05 0 0 0 0 0 0 0.1CPMD8* 0.03 0.01 0.01 0 0 0 0 0 0 0.03CRGD 0.16 0.07 0.07 0 0 0 0 0 0 0.16

(continued on next page)

9J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

Please cite this article as: Bennett KL, et al, Proteomic analysis of human cataract aqueous humour: Comparison of one-dimensional gel LCMS with two-dimensional LCMS of unlabelled..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

Table 2 (continued)

Gene symbol 1D-G-CID 2D-SG-CID 2D-SG-HCD Global SCV

Total Mean SD Total Mean SD Total Mean SD

Group 5: 1D-G-CIDNRX3A 0.03 0.01 0.01 0 0 0 0 0 0 0.03MUC7* 0.06 0.01 0.03 0 0 0 0 0 0 0.06SPRL1 0.06 0.01 0.02 0 0 0 0 0 0 0.06CO5 0.03 0.01 0.01 0 0 0 0 0 0 0.03A2AP 0.05 0.01 0.02 0 0 0 0 0 0 0.05PRDX1 0.21 0.06 0.09 0 0 0 0 0 0 0.21GGCT 0.22 0.07 0.09 0 0 0 0 0 0 0.22S10A8 0.4 0.09 0.13 0 0 0 0 0 0 0.4CO2 0.04 0.02 0.01 0 0 0 0 0 0 0.04CO9A1* 0.03 0.01 0.02 0 0 0 0 0 0 0.03IPSP 0.08 0.03 0.04 0 0 0 0 0 0 0.08NGAL 0.1 0.02 0.05 0 0 0 0 0 0 0.1PLAK 0.07 0.01 0.03 0 0 0 0 0 0 0.07NEUS 0.08 0.02 0.04 0 0 0 0 0 0 0.08MDHC 0.09 0.02 0.03 0 0 0 0 0 0 0.09H2B1K* 0.29 0.06 0.13 0 0 0 0 0 0 0.29ARGI1 0.16 0.03 0.04 0 0 0 0 0 0 0.16S10A9 0.26 0.05 0.12 0 0 0 0 0 0 0.26FRIH 0.17 0.05 0.08 0 0 0 0 0 0 0.17SAMP 0.13 0.03 0.05 0 0 0 0 0 0 0.13H4* 0.29 0.06 0.13 0 0 0 0 0 0 0.29ZG16B* 0.12 0.02 0.05 0 0 0 0 0 0 0.12CHL1 0.03 0.01 0.01 0 0 0 0 0 0 0.03CASPE 0.05 0.02 0.02 0 0 0 0 0 0 0.05PGAM2* 0.08 0.02 0.04 0 0 0 0 0 0 0.08SIAE 0.04 0.01 0.02 0 0 0 0 0 0 0.04SAA4 0.15 0.06 0.08 0 0 0 0 0 0 0.15PIP 0.18 0.04 0.08 0 0 0 0 0 0 0.18PPIA 0.12 0.02 0.05 0 0 0 0 0 0 0.12SORC1 0.02 0 0.01 0 0 0 0 0 0 0.02NBL1* 0.1 0.02 0.04 0 0 0 0 0 0 0.1FILA2* 0.01 0 0 0 0 0 0 0 0 0.01DIAC 0.04 0.01 0.02 0 0 0 0 0 0 0.04NELL2 0.03 0.01 0.01 0 0 0 0 0 0 0.03LCN1 0.1 0.02 0.05 0 0 0 0 0 0 0.1SLPI* 0.15 0.03 0.07 0 0 0 0 0 0 0.15PLBL2* 0.03 0.01 0.01 0 0 0 0 0 0 0.03SPB12 0.06 0.01 0.03 0 0 0 0 0 0 0.06TIMP2 0.14 0.03 0.06 0 0 0 0 0 0 0.14SPB3 0.04 0.01 0.02 0 0 0 0 0 0 0.04CYTS 0.13 0.03 0.06 0 0 0 0 0 0 0.13ANXA2 0.06 0.01 0.03 0 0 0 0 0 0 0.06CALR 0.04 0.01 0.02 0 0 0 0 0 0 0.04MEGF8 0.01 0 0 0 0 0 0 0 0 0.01

Group 6: 2D-SG-CIDDYHC2* 0 0 0 0.01 0 0 0 0 0 0.017B2* 0 0 0 0.19 0.08 0.08 0 0 0 0.19VGF* 0 0 0 0.08 0.02 0.04 0 0 0 0.08GNS 0 0 0 0.04 0.02 0.02 0 0 0 0.04TETN 0 0 0 0.25 0.09 0.08 0 0 0 0.25TMCO7* 0 0 0 0.02 0 0.01 0 0 0 0.02OBSCN* 0 0 0 ‹0.01 0 0 0 0 0 ‹0.01SCG1 0 0 0 0.08 0.02 0.03 0 0 0 0.08PTHD2* 0 0 0 0.02 0 0.01 0 0 0 0.02CO9 0 0 0 0.06 0.02 0.03 0 0 0 0.06CBPB2 0 0 0 0.05 0.01 0.02 0 0 0 0.05CFAD 0 0 0 0.09 0.05 0.05 0 0 0 0.09VTN 0 0 0 0.05 0.02 0.03 0 0 0 0.05SYK* 0 0 0 0.05 0.01 0.02 0 0 0 0.05LTBP2 0 0 0 0.03 0.01 0.01 0 0 0 0.03OTOF* 0 0 0 0.01 0 0.01 0 0 0 0.01

10 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

Please cite this article as: Bennett KL, et al, Proteomic analysis of human cataract aqueous humour: Comparison of one-dimensional gel LCMS with two-dimensional LCMS of unlabelled..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

Table 2 (continued)

Gene symbol 1D-G-CID 2D-SG-CID 2D-SG-HCD Global SCV

Total Mean SD Total Mean SD Total Mean SD

Group 6: 2D-SG-CIDNUCB1 0 0 0 0.07 0.02 0.03 0 0 0 0.07FHR1 0 0 0 0.06 0.01 0.03 0 0 0 0.06MACF4* 0 0 0 ‹0.01 0 0 0 0 0 ‹0.01CMGA* 0 0 0 0.09 0.02 0.04 0 0 0 0.09AGRIN 0 0 0 0.01 0 0 0 0 0 0.01PCSK1* 0 0 0 0.14 0.03 0.06 0 0 0 0.14NBAS* 0 0 0 0.01 0 0 0 0 0 0.01SDC4* 0 0 0 0.12 0.02 0.05 0 0 0 0.12DACT1* 0 0 0 0.02 0 0.01 0 0 0 0.02LMAN2 0 0 0 0.06 0.01 0.03 0 0 0 0.06

Group 7: 2D-SG-HCDTITIN* 0 0 0 0 0 0 0.01 0 0 0.01CHLE* 0 0 0 0 0 0 0.05 0.02 0.04 0.05PGCP* 0 0 0 0 0 0 0.04 0.02 0.03 0.04CCD19* 0 0 0 0 0 0 0.07 0.03 0.05 0.07YR003* 0 0 0 0 0 0 0.03 0.02 0.02 0.03CLK4* 0 0 0 0 0 0 0.05 0.02 0.03 0.05ZN234* 0 0 0 0 0 0 0.03 0.01 0.02 0.03LBA1* 0 0 0 0 0 0 0.01 0 0.01 0.01LAMP2 0 0 0 0 0 0 0.05 0.05 0 0.05ADA21* 0 0 0 0 0 0 0.03 0.02 0.02 0.03ZN224* 0 0 0 0 0 0 0.03 0.01 0.02 0.03ECM1 0 0 0 0 0 0 0.04 0.02 0.03 0.04RNT2 0 0 0 0 0 0 0.07 0.03 0.05 0.07APOM* 0 0 0 0 0 0 0.07 0.03 0.05 0.07ALS 0 0 0 0 0 0 0.04 0.02 0.03 0.04

1D-G-CID

11J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

(7.6 μg) as compared to the individual 2D shotgun samples(4 μg for Pools 1–4 and 3.2 μg for Pool 5). To compensate for thedoubled quantity of total protein and maintain a level ofconsistency, forty fractions were collected as opposed totwenty in the previous section.

Fig. 4 – Representative chromatogram of the off-lineseparation of the proteins present in albumin- andimmunoglobulin G-depleted aqueous humour from control(cataract) patients. Prior to depletion, the samples werenormalised on total protein content to 20 μg (Pools 1–4) and16 μg (Pool 5). Assuming depletion efficiency of 80%, thequantity of protein injected for fractionation was 4 μg (Pools1–4) and 3.2 μg (Pool 5). Predominant peaks containedpeptides from (A) CO3, PTGDS, PEDF, TRFE and CERU; (B) TRFEand HPT; (C) TRFE; and (D) PTGDS and PEDF; (E) TRFE, PEDF,CERU, CFAB and VGF; (F) TRFE; and (G) CO4A, A1AT, AACT,A2MG and CO3.

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

Excluding trypsin, keratin, residual albumin and immuno-globulins, 77 protein groups (c.f., 156 1D-G-CID and 117 2D-SG-CID) were identified from the two liquid-based analyses

63

1526

3258

3

1

2D-SG-CID 2D-SG-HCD

Fig. 5 – Overview of the number of protein groups (total 198)observed via the three proteomic approaches. 1D-G-CID(one-dimensional gel collision-induced dissociation),2D-SG-CID (two-dimensional shotgun collision-induceddissociation) and 2D-SG-HCD (two-dimensional shotgunhigh-energy collision-induced dissociation).

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

Fig. 6 – Relative protein quantitation of one third of the AHfrom each of Pools 2 to 4 (Table 1) labelled with iTRAQ®reagents 114 to 116, respectively. Assuming depletionefficiency of 80%, the quantity of protein injected forfractionation was 7.6 μg. The red box indicates quantitationof the identified proteins relative to the 114 label with amaximum fold change of ±2.0 (±log10 0.3) in the 115 or 116channels (95% confidence). Proteins outside these values aredifferentially-regulated. The black line represents a Gaussianmodel and the blue line represents a Cauchy model. Cauchyis a better fit to the data and was used to estimate the ratioconfidence.

12 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

fragmented by HCD (Table 2 2D-SG-HCD; Groups 1, 3, 4 and 7).In accordance with all the previous analyses, TRFE was againthe most abundant protein with a maximum sequencecoverage of 65% (52 unique peptides identified from 1858spectra); followed by PEDF (40%, 17 unique peptides identifiedfrom 449 spectra); and A1AT (30%, 17 unique peptidesidentified from 330 spectra). Comparison of the data setsobtained from the gel-based approach (n=5) with the com-bined 2D ‘shotgun’ methods (CID plus HCD fragmentation)(n=7) revealed that 93 of the combined protein groupsidentified were common to both approaches (Table 2; Groups1, 2 and 3; Fig. 5. The gel-based method had 63 protein groupsthat were not identified by the 2D ‘shotgun’ approaches(Table 2; Group 5), however, there was a total of 42 proteingroups observed in the fractionated samples that were notidentified from the gels (Table 2; Groups 4, 6 and 7). Abreakdown comparison of the two off-line fractionationmethods revealed that there were 59 protein groups incommon (Table 2; Groups 1 and 4); 26 protein groups onlyidentified by the 2D-SG-CID approach (Table 2; Group 6); and15 protein groups unique to the 2D-SG-HCD analyses (Table 2;Group 7). Shown in Fig. 5 is the number of protein groupsidentified from each combination of the methodology. SeeSupplementary Table S1 for a complete overview of allproteins identified throughout this study.

Of the 198 proteins identified via the combined proteomicapproaches, 44, 157 and 52 were also identified in the studiesby Richardson et al. [23], Fautsch et al. [25], and Marinach-Patrice et al. [24], respectively. In addition, we identified 38protein groups not observed by any of these researchers.

As a generically-applicable approach, 1D-G-CID analysis ofAH provided approximately 25% more protein group identifi-cations (Fig. 5) than the identical samples analysed by 2D-SG-CID.When applied to the global analysis of a sample, however,1D-SDS-PAGE is unsuited to iTRAQ® labelling. Thus, ourmethodology necessitated the pursuance of an iTRAQ®-compatible strategy albeit at the loss of some proteins butwith a concurrent gain of proteins not identified from the gel-based analyses. The most feasible explanation as to whypeptide fractionation provided less protein identifications isas follows: if an overly-abundant peptide (e.g., from TRFE) ispresent in a fraction, then any low-abundance peptides in thesame fraction will be suppressed and thus neither fragmentedin the mass spectrometer nor identified. The collection ofadditional fractions should partly overcome dynamic rangesissues introduced by overly-abundant peptides and thusimprove the shortcomings of the approach. Note for example,the comparison of the spectral counts for TRFE from the 1D-G-CID analyses (n=5) compared to the 2D-SG-CID analyses(n=5); 8,318 c.f. 19,713 (Table 2, Group 1). That is, 2.4 timesmore spectral counts were obtained with 2D-SG-CID analysescompared to the 1D-G-CID analyses. Two-dimensional chro-matography vastly increases the resolution of peptide sepa-ration compared to one-dimensional methods, however, theapproach still suffers from high sample complexity at thepeptide level. Gilar et al. [33] reported that by increasingfraction frequency, the separation performance can be im-proved. Smaller fraction collection windows minimise thepotential overlap of abundant peptides across adjacentfractions and alleviates dynamic range issues during ESI of

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

abundant peptides, i.e., in this case, by TRFE suppressingpeptides from less abundant proteins. In addition, an increasednumber of fractions are advantageous for the time-demandingHCDduty cycle. For these reasons, forty fractionswere collectedfor the 2D-SG-HCD (n=2) experiments as opposed to thestandard twenty fractions employed in the 2D shotgun analysisof the underivatised samples (2D-SG-CID).

3.5. Relative quantitation of aqueous humour

A relative quantitative ratio could be assigned to 74 of the 77proteins identified by the 2D-SG-HCD approach (n=2). Com-pared to the iTRAQ® reporter ion at m/z 114.1112, 88% of theproteinshada foldchangeofwithin±2.0 (±log10 0.3) in the 115or116 channels with 95% confidence (Fig. 6). That is, the majorityof theproteins identified in cataract (control)AHwerepresent inan equimolar ratio. If the 117 channel is included, however, thenumber was reduced to less than a third. As noted above,sample Pool 5 modified with the 117 label was suboptimal asthere was less total protein than the other three sample Pools 2to 4 (Table 1). Thus, the data strongly highlights the importanceof standardising the total protein content prior to labellingsamples with the iTRAQ® reagent. Although the data werenormalised, computational methods are unable to rescue datasuch as that obtained from the 117 channel [34–36]. Low signalintensity leads topooraccuracyandultimately, lowerquantitiesof material (i.e., peptides from a decreased quantity of inputprotein) have two potential consequences: (i) a less precisesignal for quantitation; and/or (ii) complete loss of the iTRAQ®

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

13J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

signal below either the detection limit of themass spectrometeror below the subtracted noise level. When Pool 5 was excluded,the results of the iTRAQ® relative quantitation data weresurprisingly lessvariable thanenvisaged. Especially consideringthat even with particular care is taken in clinically-matchingpatient samples, there is always a high degree of human-to-human variability. Additionally, the data set consists of only4 samples. The majority of studies involving patients (e.g.,clinical drug trials) usually require a large cohort to establishacceptable significance and confidence levels. An overview ofthe relative quantitation of all proteins is given in Supplemen-tary Table S2.

The iTRAQ® reagent was used in this study to determinethe feasibility of labelling low quantities of protein from AHand to establish cataract patients as controls for comparisonwith other eye diseases. The use of the iTRAQ® labellingstrategy to propose hypotheses concerning this disease isoutside the scope of this manuscript. Nonetheless, theproteins that were differentially-regulated between Pools 2to 4 did reveal some interesting findings. Compared to the 114-labelled sample (Table 1, Pool 2), β-crystallin S (CRBS), carbonicanhydrase 3 (CAH3), transforming growth factor-β-inducedprotein ig-h3 (BGH3), HPT, α-1-acid glycoprotein 1 (A1AG1),apolipoprotein A-IV (APOA4), retinol-binding protein 3 (RET3),dickkopf-related protein 3 (DKK3) plus some of the immuno-globulin proteins were differentially-regulated. It is notsurprising that the expression of CRBS is variable as thepatients in this study have undergone cataract surgery. Thedata suggests that the patients have an altered expression ofCRBSwhichmay be attributable to the degree of progression ofthe disease. Decreased levels of carbonic anhydrase (CAH) inthe eye lens have been linked to the development of cataract[37] and the release of CAH from the lens into the AH mayindicate cataract development and opacification of the eyelens. Transforming growth factor β (TGFB) has also beenshown to induce cataractous changes in the rat lens in vitro[38] and intravitreal injection of active TGFB induced rat lensesto undergo cataractous changes [39].

3.6. Gene ontology annotation

Theproteins identified fromhumanAHwere groupedaccordingto the Gene Ontology (GO) terminology [26] using the bioinfor-matic tool MASPECTRAS [40]. An overview of the classificationsand the number of proteins in each GO term is given in Table 3(p<0.01, >2 proteins per GO classification). Keratin, immunoglo-bulins and residual albumin were excluded from the analyses.When catalogued according to cellular component, 33% of theproteins identified in this studyare secretedproteins commonlyfound in other body fluids, the extracellular space or extracel-lular matrix (Table 3A, GO:0005615 and GO:0005578). Thisobservation plus the absence of ribosomal and mitochondrialproteins, indicated that the collected samples were largelyuncontaminated with cellular debris. Some contamination,however, was apparent as evidenced by the 31 proteinsclassified as nuclear (Table 3A, GO:0005634).

Clustering the proteins according to molecular function(p<0.01, >2 proteins per GO classification) (Table 3B) revealedthat 70 to 80% of the proteins identified in AH exhibit differentbinding functions. Amongst these, protein binding (Table 3B,

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

GO:0005515, 48 proteins) was the most frequent functionfollowed by calcium ion (Table 3B, GO:0005509, 29 proteins);zinc ion (Table 3B, GO:0008270, 14 proteins); heparin (Table 3B,GO:0008201, 10 proteins); DNA (Table 3B, GO:0003677, 9proteins); and copper ion binding (Table 3B, GO:0005507, 6proteins). The next major category was proteins that inhibitproteolytic activity. Thirteen of the 23 proteins classified asexhibiting serine-type endopeptidase inhibitor activity belongto the family of SERPINS (Table 3B, GO:0004867). Approximate-ly two-thirds of human SERPINS have an extracellular role andcan undergo a unique change in conformation when inhibit-ing target proteases [41]. Although corticosteroid-bindingglobulin (CBG), thyroxine-binding globulin (THBG), and α-2-antiplasmin (A2AP) are classified by GO as inhibitors, the roleof these proteins is non-inhibitory; and are involved in cortisolbinding [42], thyroxine-binding [43], and anti-angiogenicproperties [44]. The thirdmain GO group comprised of proteinswith cysteine protease inhibitor activity (Table 3B, GO:0004869,6 proteins); and three of the proteins with endopeptidaseinhibitor activity were members of the complement cascade(Table 3B, GO:0004867).

Grouping the identified proteins according to involvementin biological processes (p<0.01, >2 proteins per GO classifica-tion) (Table 3C) provided a complementary picture to molec-ular function. The data revealed that the major classificationsof the proteins identified in human AH were: (i) acute-phaseresponse (Table 3C, GO:0006953, 11 proteins); (ii) complementactivation (Table 3C, GO:0006957 and GO:0006958, 14 proteins);and (iii) cholesterol homeostasis and reverse transport(Table 3C, GO:0042632 and GO:0043691, 12 proteins). Theacute-phase proteins act in a positive fashion in response toinflammation and have different physiological functions forthe immune system. Some act to destroy or inhibit growth ofmicrobes (e.g., C-reactive protein, CRP; serum amyloid A-4protein, SAA4). Other proteins provide a negative feedback onthe inflammatory response, e.g., A1AT and AACT. A1ATprotects tissues from enzymes of inflammatory cells (inparticular elastase) [45]. AACT inhibits the activity of certainenzymes called proteinases (e.g., cathepsin G that is found inneutrophils); and chymases (found in mast cells) via cleavageand alteration of protein shape or conformation. This activityprotects tissues from damage caused by proteolytic enzymes[46]. A2MG and coagulation factors (e.g., prothrombin andfibrinogen) affect coagulation.

Several proteins from the complement family were iden-tified in this study. These proteins fall into two categories: (i)the classical pathway that is activated by a group of bloodproteins that mediate the specific antibody response triggeredby antigen-bound antibody molecules (CO3; CO4A; comple-ment factor I, CFAI; complement C5, CO5; complementcomponent C9, CO9; complement C2, CO2); and (ii) thealternative pathway that is an innate component of thenatural defence of the immune system against infectionswhich can operate without antibody participation (CO3;complement factor B, CFAB; CO5; CO9; complement factor D,CFAD; CO2). The alternative pathway is known to play animportant role in the development of age-related maculardegeneration and genetic analysis has shown a strongassociation between AMD and specific polymorphisms ofcomplement factors.

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

Table 3 – Gene ontology classification of the proteinsidentified in this study grouped according to: (A) cellularcomponent; (B) molecular function; and (C) biologicalprocesses (p<0.01, >2 proteins per GO classification).

Geneontology

identification

No.proteins

Gene ontology term Geneontologyp value

AGO:0005615 58 Extracellular space 0.001GO:0005634 31 Nucleus 0.001GO:0016021 14 Integral to membrane 0.001GO:0005578 12 Proteinaceous

extracellular matrix0.001

GO:0005764 11 Lysosome 0.001GO:0031093 7 Platelet alpha granule

lumen0.001

GO:0034361 6 Very low densitylipoprotein particle

0.001

GO:0016020 6 Membrane 0.001GO:0034364 4 High-density lipoprotein

particle0.001

GO:0042627 4 Chylomicron 0.001GO:0005856 3 Cytoskeleton 0.001GO:0042470 5 Melanosome 0.004GO:0005625 11 Soluble fraction 0.005

41 Unannotated

BGO:0005515 48 Protein binding 0.001GO:0005509 29 Calcium ion binding 0.001GO:0004867 23 Serine-type

endopeptidase inhibitoractivity

0.001

GO:0008270 14 Zinc ion binding 0.001GO:0008201 10 Heparin binding 0.001GO:0003677 9 DNA binding 0.001GO:0005507 6 Copper ion binding 0.001GO:0004869 6 Cysteine protease

inhibitor activity0.001

GO:0005529 6 Sugar binding 0.001GO:0001540 4 Beta-amyloid binding 0.001GO:0060228 4 Phosphatidylcholine-

sterol O-acyltransferaseactivator activity

0.001

GO:0004866 3 Endopeptidase inhibitoractivity

0.001

GO:0005319 3 Lipid transporter activity 0.001GO:0016918 3 Retinal binding 0.001GO:0003676 3 nucleic acid binding 0.001GO:0005200 3 Structural constituent of

cytoskeleton0.001

GO:0004252 8 Serine-typeendopeptidase activity

0.003

GO:0005520 3 Insulin-like growth factorbinding

0.003

GO:0005044 3 Scavenger receptoractivity

0.006

GO:0005543 3 Phospholipid binding 0.006GO:0004872 3 Receptor activity 0.007

24 Unannotated

CGO:0006953 11 Acute-phase response 0.001GO:0006958 8 Complement activation,

classical pathway0.001

GO:0006957 6 Complement activation,alternative pathway

0.001

Table 3 (continued)

Geneontology

identification

No.proteins

Gene ontology term Geneontologyp value

GO:0042632 6 Cholesterol homeostasis 0.001GO:0043691 6 Reverse cholesterol

transport0.001

GO:0007275 5 Multicellular organismaldevelopment

0.001

GO:0033344 5 Cholesterol efflux 0.001GO:0006879 5 Cellular iron ion

homeostasis0.001

GO:0006955 5 Immune response 0.001GO:0045087 5 Innate immune response 0.001GO:0006954 4 Inflammatory response 0.001GO:0033700 4 Phospholipid efflux 0.001GO:0006350 4 Transcription 0.001GO:0006355 3 Regulation of

transcription,DNA-dependent

0.001

GO:0008544 3 Epidermis development 0.001GO:0007596 5 Blood coagulation 0.002GO:0006096 4 Glycolysis 0.002GO:0016525 3 Negative regulation of

angiogenesis0.002

GO:0030212 3 Hyaluronan metabolicprocess

0.002

GO:0007165 5 Signal transduction 0.003GO:0006508 18 Proteolysis 0.005GO:0008285 5 Negative regulation of

cell proliferation0.008

42 Unannotated

14 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

The GO analysis performed in this study showed substan-tial overlap and agreement with Richardson et al. [23]. Thisand the current studies plus the investigations from Mar-inach-Patrice et al. [24] and Fautsch et al. [25] reveal a highnumber of previously-unidentified proteins from AH thatundoubtedly play important roles in eye physiology.

4. Concluding remarks

This manuscript presents an overview of the proteins presentin AH collected during routine cataract surgery from otherwisehealthy eyes, thus providing a representative profile for aregular age-related condition of the adult human eye. Insummary, the 198 unique protein groups identified via thecombined proteomicmethods are involved in diverse biologicalprocesses and function such as binding (protein, metal ion,nucleic acid, lipid, sugar and steroid), inhibition of proteolyticactivity, transport, immune, and inflammatory response. Quan-titative information provided by the iTRAQ® methodologyrevealed that when standardised to total protein input, themajority of the identified proteinswere present in an equimolarratio. In addition, some of the differentially-expressed proteinsappear tobe implicated in theprogressionof cataract formation.

We have shown that it is feasible to qualitatively andquantitatively compare the protein quantity that can beobtained from an individual AH sample. This study providesthe foundation for future proteomic investigations where wewill use the proteins identified from the AH of cataract

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

15J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

patients as a baseline to compare with changes in proteinexpression levels in the AH from patients suffering fromintraocular diseases. Such studies include comparison of theexpression profiles between healthy individuals (cataract) andpatients suffering from widespread eye diseases; and trackingproteomic changes in AH from individual patients withprogression of eye disease or during treatment.

Supplementary materials related to this article can befound online at doi:10.1016/j.jprot.2010.10.002.

Acknowledgements

The authors wish to thank Denis Hochstrasser for consultationonappropriate collectionand storageof body fluids; Stefan Sacuand Michael Georgopoulos for surgical removal of AH fromcataract patients; RenateAmonandMarionMunk for assistancewith sample collection and preparation for mass spectrometricanalysis; GoranMitulović for advice on alternativeHPLC solventsystems; Bernhard Kuster for consultation and assistance inestablishing high-energy collision-induced dissociation (HCD)on the Orbitrap XL mass spectrometer for relative quantitationusing iTRAQ® reagents; and Christoph Binder for criticalevaluation of the manuscript. Work in our laboratory issupported by the Austrian Academy of Sciences, the AustrianFederal Ministry for Science and Research (Gen-Au projects,APP, BIN and Dragon) and by the Austrian Science Fund FWF.

R E F E R E N C E S

[1] Yates III JR, Gilchrist A, Howell KE, Bergeron JJ. Proteomics oforganelles and large cellular structures. Nat Rev Mol Cell Biol2005;6:702–14.

[2] Andersen JS, Mann M. Organellar proteomics: turninginventories into insights. EMBO Rep 2006;7:874–9.

[3] Schmidt A, Aebersold R. High-accuracy proteome maps ofhuman body fluids. Genome Biol 2006;7:242.

[4] Kocher T, Superti-Furga G. Mass spectrometry-basedfunctional proteomics: from molecular machines to proteinnetworks. Nat Methods 2007;4:807–15.

[5] Bantscheff M, Schirle M, Sweetman G, Rick J, Kuster B.Quantitative mass spectrometry in proteomics: a criticalreview. Anal Bioanal Chem 2007;389:1017–31.

[6] Steen H, Mann M. The ABC's (and XYZ's) of peptidesequencing. Nat Rev Mol Cell Biol 2004;5:699–711.

[7] Cravatt BF, Simon GM, Yates III JR. The biological impact ofmass-spectrometry-based proteomics. Nature 2007;450:991–1000.

[8] Gao BB, Chen X, Timothy N, Aiello LP, Feener EP.Characterization of the vitreous proteome in diabetes withoutdiabetic retinopathy and diabetes with proliferative diabeticretinopathy. J Proteome Res 2008;7:2516–25.

[9] Kim T, Kim SJ, Kim K, Kang UB, Lee C, Park KS, et al. Profiling ofvitreous proteomes from proliferative diabetic retinopathyand nondiabetic patients. Proteomics 2007;7:4203–15.

[10] Yu J, Liu F, Cui SJ, Liu Y, Song ZY, Cao H, et al. Vitreousproteomic analysis of proliferative vitreoretinopathy.Proteomics 2008;8:3667–78.

[11] de Souza GA, Godoy LM, Mann M. Identification of 491proteins in the tear fluid proteome reveals a largenumber of proteases and protease inhibitors. Genome Biol2006;7:R72.

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

[12] Adachi J, Kumar C, Zhang Y, Olsen JV, Mann M. The humanurinary proteome contains more than 1500 proteins,including a large proportion of membrane proteins. GenomeBiol 2006;7:R80.

[13] Issaq HJ, Xiao Z, Veenstra TD. Serum and plasma proteomics.Chem Rev 2007;107:3601–20.

[14] Apte RS, Niederkorn JY. Isolation and characterization of aunique natural killer cell inhibitory factor present in theanterior chamber of the eye. J Immunol 1996;156:2667–73.

[15] Kramer M, Goldenberg-Cohen N, Axer-Siegel R,Weinberger D,Cohen Y, Monselise Y. Inflammatory reaction in acute retinalartery occlusion: cytokine levels in aqueous humor andserum. Ocul Immunol Inflamm 2005;13:305–10.

[16] Niederkorn JY. See no evil, hear no evil, do no evil: the lessonsof immune privilege. Nat Immunol 2006;7:354–9.

[17] Missotten GS, Beijnen JH, Keunen JE, Bonfrer JM. Proteomics inuveal melanoma. Melanoma Res 2003;13:627–9.

[18] Grus FH, Joachim SC, Sandmann S, Thiel U, Bruns K, LacknerKJ, et al. Transthyretin and complex protein pattern inaqueous humor of patients with primary open-angleglaucoma. Mol Vis 2008;14:1437–45.

[19] Duan X, Lu Q, Xue P, Zhang H, Dong Z, Yang F, et al. Proteomicanalysis of aqueous humor from patients with myopia. MolVis 2008;14:370–7.

[20] Funding M, Vorum H, Honore B, Nexo E, Ehlers N. Proteomicanalysis of aqueous humour from patients with acute cornealrejection. Acta Ophthalmol Scand 2005;83:31–9.

[21] Rohde E, Tomlinson AJ, Johnson DH, Naylor S. Comparisonof protein mixtures in aqueous humor by membranepreconcentration-capillary electrophoresis-massspectrometry. Electrophoresis 1998;19:2361–70.

[22] Stastna M, Behrens A, Noguera G, Herretes S, McDonnell P,Van Eyk JE. Proteomics of the aqueous humor in healthy NewZealand rabbits. Proteomics 2007;7:4358–75.

[23] Richardson MR, Price MO, Price FW, Pardo JC, Grandin JC, YouJ, et al. Proteomic analysis of human aqueous humor usingmultidimensional protein identification technology. Mol Vis2009;15:2740–50.

[24] Escoffier P, Paris L, Bodaghi B, Danis M, Mazier D,Marinach-Patrice C. Pooling aqueous humor samples:bias in 2D-LC–MS/MS strategy? J Proteome Res2010;9:789-97.

[25] Roy Chowdhury U, Madden BJ, Charlesworth MC, Fautsch MP.Proteome analysis of human aqueous humor. InvestOphthalmol Vis Sci 2010;51:4921–31.

[26] Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, CherryJM, et al. Gene ontology: tool for the unification of biology.The Gene Ontology Consortium. Nat Genet 2000;25:25–9.

[27] Shevchenko A,WilmM, VormO, MannM.Mass spectrometricsequencing of proteins silver-stained polyacrylamide gels.Anal Chem 1996;68:850–8.

[28] Rappsilber J, Ishihama Y, Mann M. Stop and go extraction tipsfor matrix-assisted laser desorption/ionization,nanoelectrospray, and LC/MS sample pretreatment inproteomics. Anal Chem 2003;75:663–70.

[29] Ross PL, Huang YN, Marchese JN, Williamson B, Parker K,Hattan S, et al. Multiplexed protein quantitation inSaccharomyces cerevisiae using amine-reactive isobaric taggingreagents. Mol Cell Proteomics 2004;3:1154–69.

[30] Gilar M, Olivova P, Daly AE, Gebler JC. Two-dimensionalseparation of peptides using RP–RP-HPLC system withdifferent pH in first and second separation dimensions. J SepSci 2005;28:1694–703.

[31] Colinge J, Masselot A, Giron M, Dessingy T, Magnin J. OLAV:towards high-throughput tandem mass spectrometry dataidentification. Proteomics 2003;3:1454–63.

[32] Baussant T, Bougueleret L, Johnson A, Rogers J, Menin L, HallM, et al. Effective depletion of albumin using a newpeptide-based affinity medium. Proteomics 2005;5:973–7.

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002

16 J O U R N A L O F P R O T E O M I C S X X ( 2 0 1 0 ) X X X – X X X

[33] Gilar M, Olivova P, Daly AE, Gebler JC. Orthogonality ofseparation in two-dimensional liquid chromatography. AnalChem 2005;77:6426–34.

[34] Hu J, Qian J, Borisov O, Pan S, Li Y, Liu T, et al. Optimizedproteomic analysis of a mouse model of cerebellardysfunction using amine-specific isobaric tags. Proteomics2006;6:4321–34.

[35] Hundertmark C, Fischer R, Reinl T, May S, Klawonn F, JanschL. MS-specific noise model reveals the potential ofiTRAQ in quantitative proteomics. Bioinformatics 2009;25:1004–11.

[36] Karp NA, Huber W, Sadowski PG, Charles PD, Hester SV, LilleyKS. Addressing accuracy and precision issues in iTRAQquantitation. Mol Cell Proteomics 2010;9:1885–97.

[37] Bakker A. Carbonic anhydrase and cataracta lentis. Br JOphthalmol 1948;32:910–2.

[38] Liu J, Hales AM, Chamberlain CG, McAvoy JW. Induction ofcataract-like changes in rat lens epithelial explants bytransforming growth factor beta. Invest Ophthalmol Vis Sci1994;35:388–401.

[39] Hales AM, Chamberlain CG, Dreher B, McAvoy JW. Intravitrealinjection of TGFbeta induces cataract in rats. InvestOphthalmol Vis Sci 1999;40:3231–6.

Please cite this article as: Bennett KL, et al, Proteomic analysisdimensional gel LCMS with two-dimensional LCMS of unlabelle

[40] Mohien CU, Hartler J, Breitwieser F, Rix U, Rix LR, Winter GE,et al. MASPECTRAS 2: an integration and analysis platform forproteomic data. Proteomics 2010;10:2719–22.

[41] Huntington JA, Read RJ, Carrell RW. Structure of aserpin–protease complex shows inhibition by deformation.Nature 2000;407:923–6.

[42] Klieber MA, Underhill C, Hammond GL, Muller YA.Corticosteroid-binding globulin, a structural basis for steroidtransport and proteinase-triggered release. J Biol Chem2007;282:29594–603.

[43] Zhou A, Wei Z, Read RJ, Carrell RW. Structural mechanism forthe carriage and release of thyroxine in the blood. Proc NatlAcad Sci USA 2006;103:13321–6.

[44] Dawson DW, Volpert OV, Gillis P, Crawford SE, Xu H, BenedictW, et al. Pigment epithelium-derived factor: a potent inhibitorof angiogenesis. Science 1999;285:245–8.

[45] Kushner I. The acute phase response: an overview. MethodsEnzymol 1988;163:373–83.

[46] Kalsheker NA. Alpha 1-antichymotrypsin. Int J Biochem CellBiol 1996;28:961–4.

of human cataract aqueous humour: Comparison of one-d..., J Prot (2010), doi:10.1016/j.jprot.2010.10.002