A Method to Reuse Archived H&E Stained Histology Slides for a Multiplex Protein Biomarker Analysis

Protocol A Method to Reuse Archived H&E Stained Histology Slides for a Multiplex Protein Biomarker Analysis

James P. Hinton 1,2 , Katerina Dvorak 2, Esteban Roberts 2, Wendy J. French 2, Jon C. Grubbs 2, Anne E. Cress 1,3,* , Hina A. Tiwari 4 and Raymond B. Nagle 5

1 Cancer Biology Graduate Interdisciplinary Program, University of Arizona Cancer Center, Tucson, AZ 85724, USA; [email protected] 2 Ventana/Roche Tissue Diagnostics, Tucson, AZ 85755, USA; [email protected] (K.D.); [email protected] (E.R.); [email protected] (W.J.F.); [email protected] (J.C.G.) 3 Department of Cellular and Molecular Medicine, University of Arizona, Tucson, AZ 85724, USA 4 Department of Medical Imaging, College of Medicine, University of Arizona, Tucson, AZ 85724, USA; [email protected] 5 Department of Pathology, College of Medicine, the University of Arizona, Tucson, AZ 85724, USA; [email protected] * Correspondence: [email protected]; Tel.: +1-520-626-7553

Received: 14 September 2019; Accepted: 8 November 2019; Published: 15 November 2019

Abstract: Archived Hematoxylin and Eosin (H&E) stained pathology slides are routinely stored to index formalin-fixed paraffin-embedded (FFPE) sample tissue blocks. FFPE blocks are clinically annotated human tumor specimens that can be valuable in studies decades after the tissue is collected. If stored properly, they have the potential to yield a valuable number of serial sectioned slides for diagnostic or research purposes. However, some retrospective studies are limited in scope because the tissue samples have been depleted or not enough material is available in stored blocks for serial sections. The goal of these studies was to determine if archived H&E-stained slides can be directly reutilized by optimizing methods to de-stain and then re-stain the H&E stained slides to allow the detection of several biomarkers of interest using a conjugated antibody with chromogen multiplex immunohistochemistry procedure. This simple but innovative procedure, combined with image analysis techniques, demonstrates the ability to perform precise detection of relevant markers correlated to disease progression in initially identified tumor regions in tissue. This may add clinical value in retaining H&E slides for further use.

Keywords: immunohistochemistry; H&E stained slides; de-stain; re-stain; image analysis

1. Introduction In immunohistochemistry, several types of tissue immunostains are utilized to analyze morphological features, cellular structures, cell type, and the presence or absence of microorganisms. The most popular of the staining methods for diagnostic potential is the utilization of hematoxylin and eosin (H&E) staining [1]. H&E stains reveal structural information, with speciﬁc functional implications. H&E staining of tissue is used to assess cellular and morphological structures, identify the type of tissue, morphological variability, cell type, and pathological changes. The use of H&E staining has been the most eﬀective and utilized procedure for pathological diagnosis of patient neoplasia for over a century [2,3]. It has allowed pathologists to pinpoint focal areas of a specimen-containing aggressive tissue and foster a proper diagnosis [4]. Therefore, developing procedures to re-utilize these archived samples to determine individual biomarker expression levels (and potential protein–protein association) could assist in determining disease progression and directions for appropriate treatments.

Methods Protoc. 2019, 2, 86; doi:10.3390/mps2040086 www.mdpi.com/journal/mps Methods Protoc. 2019, 2, 86 2 of 15

H&E staining is used in conjunction with a variety of tissue ﬁxatives and allows the display of various cellular and tissue components, including the extracellular matrix, the cellular cytoplasm, and the nuclear structures [3]. The hematoxylin is converted into its oxidization product hematein, which is a basic dye that stains acidic (basophilic) tissue components (ribosomes, nuclei, and rough endoplasmic reticulum) a darker purple color and the acidic eosin dye stains other protein structures of the tissue (stroma, cytoplasm, muscle ﬁbers) a pink color [2,4,5]. They are also valuable in distinguishing normal structural components from neoplastic regions. However, with the current procedures, H&E staining is utilized along with sequential sections stained with antibody. Serial sectioning may cut through the region of intent and may result in the loss of regions necessary for critical diagnosis. This is particularly an issue with smaller core needle biopsies (CNBs) that are of a limited size and number. These samples are considered “precious” in regard to availability and require the utmost accuracy in testing procedures to result in proper diagnoses. A major advantage of a method that allows the reuse of the H&E-stained slide is that it will alleviate the need for additional sequentially sectioned slides, particularly with the diminutive CNBs. Due to the of size of CNBs, they are also subject to tissue sample exhaustion with the loss of the diagnostic lesion. This method would present a major practicality when a particular region of interest is no longer available in the sample block due to sequential cuts. The ability to reuse the initial H&E containing the lesion could be critical. De-staining these H&E-stained tissue slides could also potentially reduce the need for re-biopsy. Another advantage of re-staining archived H&E-stained slides is due the rapidly expanding use of whole-slide imaging (WSI), also known as digital pathology (DP) or virtual pathology. DP is a technology that involves the high-speed and high-resolution digital acquisition of images representing entire stained tissue sections from glass slides in a format that allows them to be viewed by a pathologist on a computer monitor [6]. This streamlines the ability of a surgical pathologist to make a primary diagnosis utilizing digitized images of the H&E-stained slide, allowing digital preservation while the H&E and other stains are fresh [7]. As the validation of this technology becomes widespread, the method reported here could be used for analysis of stored H&E-stained slides for subsequent diagnosis of tumor subtypes within a patient sample or future discovery of novel target proteins.

2. Experimental Design This procedure details steps to de-stain and reutilize archived H&E stained slides for antibody immunostaining modalities. For our research, prostate cancer was initially chosen due to frequent limitations of tissue in sample biopsies and the requirement for biomarker study. Prostate cancer (PCa) is also known to express variable levels of several markers associated with disease progression, such as phosphatase tensin homolog (PTEN) and ETS-Related Gene (ERG), making it a viable target for testing this procedure. A link between the PTEN pathway and ERG protein expression has previously been evaluated in various prostate cancer studies [8–13]. In studies investigating the trend of PTEN loss in tumors of prostatectomies and locally recurring castrate-resistant prostate cancers (CRCPs) with ERG overexpression, the data showed that the loss of PTEN was significantly associated with ERG positivity [11]. Another study indicated that the combination of ERG overexpression and PTEN deletion is common in aggressive capsular penetrating lesions [14]. Therefore, we decided that using antibodies targeting PTEN and ERG would be the validated markers in this study. We first used archived H&E-stained slides from PCa resections or biopsies stored for at least one year (with film coverslips) to demonstrate proof that the H&Es could be reutilized for biomarker stain using an H&E de-staining procedure with standard laboratory equipment and reagents. De-identified patient tissue samples were provided with no link to information that can be used to identify patients. Oversight for tissue acquisition was managed by the UA Cancer Center, an NCI designated Comprehensive Cancer Center. De-identified FFPE prostate tissue multi-Array (TMA), adenocarcinoma, lung, colon, and skin tissue slide samples (Table1) were acquired from Roche Tissue Diagnostics (RTD)/Ventana Medical Systems Inc. (VMSI or Ventana), Tucson, Arizona. De-identified PCa CNBs Methods Protoc. 2019, 2, 86 3 of 15

(Table1) used for initial feasibility testing of these procedures, were provided and serial sectioned by the University of Arizona Cancer Center Tissue Acquisition and Cellular/Molecular Analysis Resource (TACMASR) core support service with the approval of the University of Arizona Institutional Review Board (IRB). Initial H&E analysis (Table1), was provided by Dr. Ray Nagle. Several of the additional sample tissue specimens used within this study (liver, lung, normal colon and PCa TMA) were exhausted and only the H&E stained slide remained available for testing. For the samples in the cohort that retained tissue availability, sample slides for each tissue were H&E-stained and cover-slipped using the reagents and staining procedures on the Sakura Tissue-Tek Prisma Plus & Film Automated Slide Stainer & Coverslipper at VMSI and allowed to air dry in a fume hood for approximately 10 min.

Table 1. List of initially assess sample H&E stained slides with specimen parameters, de-stain results and antibodies tested.

H&E De-Stain Tissue Specimen Antibodies Used Initial H&E Analysis Result Malignant Primary_Adenocarcinoma, Gleason Prostate Resection + p40 3 + 3 = 6 Prostate Resection a + CK5/14: PTEN NA Malignant Primary_Adenocarcinoma, Gleason Prostate Resection a + CK5/14: p504s 3 + 3 = 6 Liver (Pancreas CNB a + PTEN/p504s/CK5/14 Mock CNBs (due to the cut) Met) * Liver Resection +/- HMWCK + p63 NA Lung Resection +/- HMWCK + p63 NA N. colon Resection + HMWCK + p63 NA Skin Resection - E-cadherin NA Skin Resection - E-cadherin NA Prostate TMA + ERG NA Prostate CNB + CD49f 3+3 Prostate CNB + CD49f no cancer Prostate CNB + CD49f no cancer Prostate CNB + PTEN/ERG b High grade growing into normal glands 3+3 Prostate CNB + PTEN/ERG b little bit of tumor grade 3+3 Prostate CNB + PTEN/ERG b 3+3, area of tumor, fragmented tumor Prostate CNB + PTEN/ERG b no tumor Prostate CNB + PTEN/ERG b grade 4 and 5 cancer, High grade Prostate CNB + PTEN/ERG b low grade 3, Lot of PIN Prostate CNB + PTEN/ERG b atrophy, inﬂammation Prostate CNB + PTEN/ERG b no cancer, small nerve area Prostate CNB + PTEN/ERG b a little fragment tumor Prostate CNB + PTEN/ERG b HGPIN, few basal cells left, some cancer Atypical adenomatous hyperplasia and Prostate CNB + PTEN/ERG b atrophic glands, Central zone lesion Prostate CNB + PTEN/ERG b 1 mm grade 3 tumor Prostate CNB + PTEN/ERG b small tumor area Prostate CNB + PTEN/ERG b small tumor area high grade cancer 4 and 5 trying to make Prostate CNB + CD49f/HMWCK b glands invading into norm glands Prostate CNB + CD49f/HMWCK b grade 5 cancer (high grade), PIN Prostate CNB + CD49f/HMWCK b small amount of tumor no basal cells Prostate Resection + CK 8 &18 NA Prostate CNB + CD49f little bit of tumor grade 3+3 Prostate CNB + HMWCK + p63 grade 4 and 5 cancer, High grade Prostate CNB + p504s low grade 3, Lot of PIN Prostate CNB + CK 8 &18 cancer (3 + 3 with normal) Prostate CNB + PTEN 3 + 3 ERG positive tumor, folded over Prostate CNB + CD49f 3 + 3 lesion: 3 cores Prostate CNB + ERG 3 + 3 involving 2/2 cores Methods Protoc. 2019, 2, 86 4 of 15

Table 1. Cont.

H&E De-Stain Tissue Specimen Antibodies Used Initial H&E Analysis Result Prostate CNB + CD49f 2 cores: 3 + 3 involving 2/2 cores 3 + 3 lesion in one frag 1mm heterogeneous Prostate CNB + CK 8 & 18 chromatin Prostate CNB + HMWCK + p63 tumor 3 + 3 Atrophic glands, edge normal Prostate CNB + HMWCK + p63 no tumor Prostate CNB + HMWCK + p63 Atrophy, inﬂammation Prostate CNB + CK 8&18 atrophy, inﬂammation Prostate CNB + CD49f no cancer Prostate CNB + ERG 3 + 3 fragmented tumor lost basal cells Abbreviations: Met, Metastasis; CNB, Core needle biopsy: NA, Not applicable: a Multiple H&Es prepared. b Dual chromogen immunostaining. * Mock needle cores: Appropriate H&E de-stain, (+); Moderate de-stain, (+/-); retention of H&E, (-).

2.1. Reagents and Materials Acetone (VWR, Visalia, CA, USA; BDH1101-4LP) • Xylene (Millipore, Billerica, MA, USA; 58235) • 100%–95% Ethanol • Distilled water (DI) • Reaction Buffer (Tris based, 7.6 0.2 pH) (Proprietary reagent, Ventana/RTD, Tucson, AZ, USA) • ± H&E stained tissue slides • Paraffin embedded sample tissues • Microscope slides (Matsunami TOMO®, Bellingham, WA or Superfrost™, Thermo Fischer • Scientific, Waltham, MA, USA) Slide coverslips (VWR, Visalia, CA, USA; 48393 251) • Mounting medium (coverslip sealant) (Thermo Scientific, Waltham, MA, USA; 8312-4) • 2.2. Equipment Slide Baskets (Sakura Finetek, Torrence, CA, USA; 4768) • Plastic staining dish (container) [with lids preferably] (Sakura Finetek, Torrence, CA, USA) • Forceps • Kimwipes™ • Brightfield microscope (Olympus BX40) • Parafilm (optional) ((Parafilm M® Laboratory film, VWR, Visalia, CA, USA) • Slide Scanner (Leica Aperio AT2; Leica Biosystems, Buffalo Grove, IL, USA, DP200 Slide Scanner • RTD, Tucson, AZ, USA) Incubator (60 C 5 C) (VWR, Visalia, CA, USA) • ◦ ± ◦ BenchMark ULTRA (optional) • Slide Coverslipper (Sakura Tissue -Tek Prisma Plus & Film Automated Slide stainer and • Coverslipper; Sakura Finetek, Torrence, CA, USA)

The initial antibodies chosen for the proof-of-concept testing were the on-market products VENTANA anti-p40 (B28) mouse monoclonal antibody (data not shown), anti-cytokeratin 5/14 (CK5/14) (EP1601Y/LL002) rabbit and mouse monoclonal antibody cocktail from Cell Marque (Figure1) and a rabbit polyclonal antibody (Ab) against the laminin-binding extracellular domain of integrin alpha 6 (CD49f) from the lab of Dr. Anne Cress at the University of Arizona, Department of Cellular and Molecular Medicine. The CD49f antibody was formulated and optimized from a 1 mg/mL stock concentrate to a 1:800 dilution in a pH = 7.3 Phosphate Avidin antibody diluent containing a proprietary B5 blocker, goat globulins, and 55 mM NaCl concentration with Proline preservative. Methods Protoc. 2019, 2, x FOR PEER REVIEW 5 of 15

according to the procedure listed in this report and immunostained with the various antibodies along with the corresponding sequential slide for each sample (see Results section). IHC DAB detection was accomplished by utilizing a VENTANA OptiView DAB IHC Detection kit and VENTANA ultraView DAB IHC Detection Kits. Chromogen detection was accomplished by utilizing a VENTANA Discovery Chromomap Detection Kit to target and detect the anti-HMWCK + p63 mouse monoclonal antibody cocktail with the secondary antibody linker conjugated with hydroxyquinazoline and anti-hydroxyquinazoline (HQ) horseradish peroxidase (HRP) enzymes for detection with a VENTANA Discovery Purple Chromogen. The detection for the anti-CD49f rabbit polyclonal antibody included an anti-rabbit nitrolpirazole (NP) conjugated secondary and anti-NP HRP for detection of the VENTANA Discovery Teal Chromogen. The detection of the sample slides Methodswas accomplished Protoc. 2019, 2, 86 utilizing the LEICA AT2 slide scanner and the VENTANA Digital Pathology 2005 of 15 (DP200) slide scanner.

Methods Protoc. 2019, 2, x FOR PEER REVIEW 5 of 15

p63 mouse monoclonal antibody cocktail with the secondary antibody linker conjugated with FigureFigure 1. 1Malignant. Malignant primary primary prostate prostate adenocarcinoma adenocarcinoma tissue tissue sample. sample. The The first first image image shows shows the the hydroxyquinazoline and anti-hydroxyquinazoline (HQ) horseradish peroxidase (HRP) enzymes for prostateprostate tissue tissue stained stained with with H&EH&E ((AA).). TheThe areasareas with tumor tumor and and normal normal prostate prostate gland gland tissue tissue are are detection with a VENTANA Discovery Purple Chromogen. The detection for the anti-CD49f rabbit labeled.labeled. The The H&E H&E stained stained tissue tissue slideslide waswas de-stainedde-stained and and anti anti-CK5-CK5/14/14 mouse mouse monoclonal monoclonal antibod antibodyy polyclonalcocktailcocktail antibody was was applied applied incl to touded determine determine an anti feasibilityfeasibility-rabbit ofnitrolpirazole the proposed protocol(NP) protocol conjugated (B (B).). A A sequential sequential secondary sample sample and slide slide anti -NP HRPwas forwas staineddetection stained with with of the the same sameVENTANA anti-CK5 anti-CK5/14/ 14Discovery markermarker using Teal a aChromogen. standard standard protocol protocol The procedure proceduredetection (right (rightof the panel), panel), sample for for slides was comparisonaccomplishedcomparison of of stain utilizingstain intensity intensity the to toLEICA initial initialde-stain AT2de-stain/re slide/re-stain- stainscanner procedure and the results results VENTANA (C (C).). [10 [10× magnification] Digitalmagnification]. Pathology. 200 × (DP200) slide scanner. 3. ProcedureThe antibodies utilized for additional immunostaining of five selected de-stained archived H&E stained slides to analyze and compare immunostaining intensities with corresponding sequential slides wereDe- theStaining VENTANA the H&E anti-HighSlides. Time Molecular to Completion: Weight ~2 h Cytokeratin (HMWCK) and p63 (34βE12 + 4A4, respectively)Perform mouse the de basal-staining cell cocktail, procedures Cell in Marque a fume anti-CK hood using8&18 (B22.1 manual &23.1) wash rabbit stations monoclonal, (baths) VENTANAcontaining anti-E-cadherin the reagent solvents (36) mouselisted in monoclonal, reagents and anti-CD49f materials section. and VENTANA These procedures anti-ERG negate (EPR3864) the rabbitneed monoclonal.for heat to remove Each sealed reused coverslips (re-stained) and result H&E in stained safe removal slide waswithout de-stained tissue damage. according to the procedure listed in this report and immunostained with the various antibodies along with the 1. Place the H&E archived index slides into slide baskets to allow manual rinsing. corresponding sequential slide for each sample (see Results section). 2. Manually soak slides in Acetone for 10 min to remove coverslip. IHC DAB detection was accomplished by utilizing a VENTANA OptiView DAB IHC Detection kit and VENTANA CRITICALultra STEPView Cover DAB the IHC container Detection of Kits. acetone Chromogen with either detection container was lid accomplished or parafilm to by utilizingreduce aacetone VENTANA evaporation Discovery (evaporation Chromomap reduce Detection effectiveness Kit of to coverslip target and removal). detect the anti-HMWCK + p63 mouse monoclonal antibody cocktail with the secondary antibody linker conjugated with 3. Remove the coverslip with forceps slowly to reduce damage to tissue (longer intervals allow hydroxyquinazolinecoverslip to slide and off anti-hydroxyquinazoline independently). Discard coverslip. (HQ) horseradish peroxidase (HRP) enzymes for Figure 1. Malignant primary prostate adenocarcinoma tissue sample. The first image shows the detection4. Moderately with a VENTANA rinse slides Discovery3 times in xylene Purple bath Chromogen. to remove The any detectionremaining for adhesive, the anti-CD49f allow slides rabbit prostate tissue stained with H&E (A). The areas with tumor and normal prostate gland tissue are polyclonalto sit antibody in the bath included to remove an all anti-rabbit sealant (Approximately nitrolpirazole 1 (NP)-min conjugatedhold times should secondary suffice and between anti-NP labeled. The H&E stained tissue slide was de-stained and anti-CK5/14 mouse monoclonal antibody HRP forrinses). detection [Rinse of = the 30+ VENTANA Dips in reagent] Discovery [Hold Teal= allowing Chromogen. slides to The sit in detection reagent container of the sample between slides cocktail was applied to determine feasibility of the proposed protocol (B). A sequential sample slide was accomplishedrinses]. utilizing the LEICA AT2 slide scanner and the VENTANA Digital Pathology 200 was stained with the same anti-CK5/14 marker using a standard protocol procedure (right panel), for (DP200)5. Rinse slide slides scanner. (5–6 times) with 3 min hold intervals between rinses in 95% EtOH for ~30 min to comparison of stain intensity to initial de-stain/re-stain procedure results (C). [10× magnification]. remove eosin stain. [Total time for 95% EtOH procedure is approx. 30 min, however intervals 3. Proceduremay be increased for removal of eosin from larger tissue sections.] 3. Procedure

De-Staining the H&E Slides. Time to Completion: ~2 h De-Staining the H&E Slides. Time to Completion: ~2 h Perform the de-staining procedures in a fume hood using manual wash stations (baths) containing the reagentPerform solvents the de listed-staining in reagents procedures and materials in a fume section. hood These using procedures manual negate wash the stations need for (baths heat) tocontaining remove sealed the reagent coverslips solvents and listed result in in reagents safe removal and materials without tissue section. damage. These procedures negate the need for heat to remove sealed coverslips and result in safe removal without tissue damage. 1. Place the H&E archived index slides into slide baskets to allow manual rinsing. 1. Place the H&E archived index slides into slide baskets to allow manual rinsing. 2. Manually soak slides in Acetone for 10 min to remove coverslip. 2. Manually soak slides in Acetone for 10 min to remove coverslip.

CRITICAL STEP Cover the the container container of of acetone acetone with with either either container container lid lid or or paraﬁlm parafilm to to reducereduce acetoneacetone evaporationevaporation (evaporation(evaporation reducereduce eeffectivenessﬀectiveness of of coverslip coverslip removal). removal).

3.3. Remove the coverslip with forceps slowly to reducereduce damagedamage toto tissuetissue (longer(longer intervalsintervals allowallow coverslip to slide ooffﬀ independently).independently). Discard coverslip. 4. Moderately rinse slides 3 times in xylene bath to remove any remaining adhesive, allow slides to sit in the bath to remove all sealant (Approximately 1-min hold times should suffice between rinses). [Rinse = 30+ Dips in reagent] [Hold = allowing slides to sit in reagent container between rinses]. 5. Rinse slides (5–6 times) with 3 min hold intervals between rinses in 95% EtOH for ~30 min to remove eosin stain. [Total time for 95% EtOH procedure is approx. 30 min, however intervals may be increased for removal of eosin from larger tissue sections.]

Methods Protoc. 2019, 2, x FOR PEER REVIEW 5 of 15

according to the procedure listed in this report and immunostained with the various antibodies along with the corresponding sequential slide for each sample (see Results section). IHC DAB detection was accomplished by utilizing a VENTANA OptiView DAB IHC Detection kit and VENTANA ultraView DAB IHC Detection Kits. Chromogen detection was accomplished by utilizing a VENTANA Discovery Chromomap Detection Kit to target and detect the anti-HMWCK + p63 mouse monoclonal antibody cocktail with the secondary antibody linker conjugated with hydroxyquinazoline and anti-hydroxyquinazoline (HQ) horseradish peroxidase (HRP) enzymes for detection with a VENTANA Discovery Purple Chromogen. The detection for the anti-CD49f rabbit polyclonal antibody included an anti-rabbit nitrolpirazole (NP) conjugated secondary and anti-NP HRP for detection of the VENTANA Discovery Teal Chromogen. The detection of the sample slides was accomplished utilizing the LEICA AT2 slide scanner and the VENTANA Digital Pathology 200 (DP200) slide scanner.

Figure 1. Malignant primary prostate adenocarcinoma tissue sample. The first image shows the prostate tissue stained with H&E (A). The areas with tumor and normal prostate gland tissue are labeled. The H&E stained tissue slide was de-stained and anti-CK5/14 mouse monoclonal antibody cocktail was applied to determine feasibility of the proposed protocol (B). A sequential sample slide was stained with the same anti-CK5/14 marker using a standard protocol procedure (right panel), for comparison of stain intensity to initial de-stain/re-stain procedure results (C). [10× magnification].

Methods Protoc. 2019 2 3. Procedure , , 86 6 of 15

4.De-StainingModerately the H& rinseE Slides. slides Time 3 times to Completion: in xylene bath ~2 toh remove any remaining adhesive, allow slides to sit inPerform the bath the to remove de-staining all sealant procedures (Approximately in a fume 1-min hood hold using times manual should su washﬃce betweenstations rinses).(baths) containing[Rinse the= 30 reagent+ Dips solvents in reagent] listed [Hold in reagents= allowing and slides materials to sit section. in reagent These container procedures between negate rinses]. the 5.needRinse for heat slides to remove (5–6 times) sealed with coverslips 3 min hold and intervalsresult in safe between removal rinses without in 95% tissue EtOH damage. for ~30 min to remove eosin stain. [Total time for 95% EtOH procedure is approx. 30 min, however intervals 1. Place the H&E archived index slides into slide baskets to allow manual rinsing. may be increased for removal of eosin from larger tissue sections.] 2.Methods Manually Protoc. 2019 soak, 2, x slidesFOR PEER in AcetoneREVIEW for 10 min to remove coverslip. 6 of 15

CRITICAL STEP Cover the container of acetone with either container lid or parafilm to CRITICAL STEP Cover with caps or paraffin paraffin between rinses to reduc reducee reagent evaporation. reduce acetone evaporation (evaporation reduce effectiveness of coverslip removal). ModeratelyModerately rinserinse slidesslides duringduring holdhold timestimes toto expediteexpedite eosineosinremoval. removal. 3. Remove the coverslip with forceps slowly to reduce damage to tissue (longer intervals allow 6.6. Rinse slides in distilled (DI)(DI) HH22O and lightly tap on KimwipesKimwipes™™ to remove excess.excess. coverslip to slide off independently). Discard coverslip. 7.7. Rinse slidesslides inin Reaction Reaction Bu Bufferffer (3–4 (3– rinses4 rinses with with 2 min 2 min hold hold intervals) intervals) to remove to remove Hematoxylin Hematoxylin (cap 4. Moderately rinse slides 3 times in xylene bath to remove any remaining adhesive, allow slides or(cap para or ffiparaffinn cover cover to reduce to reduce evaporation). evaporation). to sit in the bath to remove all sealant (Approximately 1-min hold times should suffice between 8.8. OPTIONAL STEP Apply extraextra rinsesrinses inin reactionreaction bubufferffer toto expediteexpedite removalremoval ofof hematoxylin.hematoxylin. rinses). [Rinse = 30+ Dips in reagent] [Hold = allowing slides to sit in reagent container between 9.9. Allow slides to dry inin hoodhood inin fumefume hoodhood forfor 55 minmin (do(do notnot allowallow tissuetissue toto drydry completelycompletely thisthis rinses].will impact immunostaining intensity). will impact immunostaining intensity). 5. Rinse slides (5–6 times) with 3 min hold intervals between rinses in 95% EtOH for ~30 min to 10.10. Run antibody IHC assay assay detection detection protocols protocols on on instruments instruments or or manual manual procedures procedures (will (will be be removevariable, eosin depending stain. [Totalon biomarker). time for 95% EtOH procedure is approx. 30 min, however intervals variable,may be increased depending for on removal biomarker). of eosin from larger tissue sections.] 11. Wash completed slides in water and dawn dish detergent mix to remove any residual reagents 11. Wash completed slides in water and dawn dish detergent mix to remove any residual reagents 12. Dehydrate slides and coverslip on Sakura coverslipper or manual dehydrate rinse with xylene 12. Dehydrate slides and coverslip on Sakura coverslipper or manual dehydrate rinse with xylene > > acetone > 80% EtOH > 90% EtOH > 100% EtOH > xylene and coverslip using mounting acetone > 80% EtOH > 90% EtOH > 100% EtOH > xylene and coverslip using mounting medium. medium.

PAUSE STEPSTEP SlidesSlides maymay remainremain inin xylenexylene andand reactionreaction bubufferﬀer stepssteps forfor extendedextended periodsperiods withoutwithout damagingdamaging thethe tissue.tissue.

4.4. ExpectedExpected ResultsResults TheThe methodmethod describeddescribed inin thisthis studystudy utilizedutilized forty-nineforty-nine samplesample H&E-stainedH&E-stained resectionresection andand CNBCNB slidesslides thatthat werewere analyzedanalyzed andand commentedcommented onon byby aa board-certiﬁedboard-certified pathologistpathologist forfor normalnormal oror neoplasticneoplastic status,status, GleasonGleason grade,grade, preservation status, status, and and any any distinguishing distinguishing features features for for the the categorization categorization of ofpotential potential aggressiveness. aggressiveness. The The initial initial testing testing was was accomplished accomplished using using DAB DAB IHC IHC detection kits to to determinedetermine retentionretention ofof markermarker stainstain intensity.intensity. DuringDuring thethe initialinitial stagesstages ofof thisthis study,study, multiplemultiple testtest samplessamples demonstrateddemonstrated lowerlower intensitiesintensities asas aa resultresult ofof utilizingutilizing anan un-optimizedun-optimized protocolprotocol (data(data notnot shown).shown). However,However, continuedcontinued editingediting andand updatesupdates toto thethe initialinitial procedureprocedure onon re-utilizedre-utilized H&EH&E indexindex slidesslides resultedresulted in in viable viable stain stain intensity, intensity demonstrating, demonstrating the the feasibility feasibility of theof the procedure procedure and and potential potential for optimizationfor optimization to culminate to culminate in stain in intensity stain intensity comparable comparable to that of to sequential that of sequential slides utilizing slides standard utilizing proceduresstandard procedures (Figure2). (Figure 2).

Figure 2. Sequential slides of human prostate tissue exhibiting cancer (Ca) invading into normal glands and high-grade prostatic intraepithelial neoplasia (HGPIN). There is an aggressive carcinoma invading glandular structures that have retained normal basal cells. The retained basal cells are positive for HMWCK + p63 (A) (grey arrowheads), positive for phosphatase tensin homolog (PTEN) (B) (black arrows), and negative for ETS-Related Gene (ERG) (C) (green arrowheads). The cancer is negative for HMWCK + p63 (A) (red arrowheads), PTEN (B) (red arrows), but positive for ERG (C) (black arrowheads). [10× magnification].

The initial testing procedures resulted in moderate but visible stain intensity providing proof- of-concept. At this stage, further optimization and repeat testing was warranted to increase the stain intensity to comparable levels of those occurring using the standard antibody staining methods and to ensure reproducibility. The procedure methods were improved by four steps: 1. Applying timed

Methods Protoc. 2019, 2, x FOR PEER REVIEW 6 of 15

CRITICAL STEP Cover with caps or paraffin between rinses to reduce reagent evaporation. Moderately rinse slides during hold times to expedite eosin removal.

6. Rinse slides in distilled (DI) H2O and lightly tap on Kimwipes™ to remove excess. 7. Rinse slides in Reaction Buffer (3–4 rinses with 2 min hold intervals) to remove Hematoxylin (cap or paraffin cover to reduce evaporation). 8. OPTIONAL STEP Apply extra rinses in reaction buffer to expedite removal of hematoxylin. 9. Allow slides to dry in hood in fume hood for 5 min (do not allow tissue to dry completely this will impact immunostaining intensity). 10. Run antibody IHC assay detection protocols on instruments or manual procedures (will be variable, depending on biomarker). 11. Wash completed slides in water and dawn dish detergent mix to remove any residual reagents 12. Dehydrate slides and coverslip on Sakura coverslipper or manual dehydrate rinse with xylene > acetone > 80% EtOH > 90% EtOH > 100% EtOH > xylene and coverslip using mounting medium.

PAUSE STEP Slides may remain in xylene and reaction buffer steps for extended periods without damaging the tissue.

4. Expected Results The method described in this study utilized forty-nine sample H&E-stained resection and CNB slides that were analyzed and commented on by a board-certified pathologist for normal or neoplastic status, Gleason grade, preservation status, and any distinguishing features for the categorization of potential aggressiveness. The initial testing was accomplished using DAB IHC detection kits to determine retention of marker stain intensity. During the initial stages of this study, multiple test samples demonstrated lower intensities as a result of utilizing an un-optimized protocol (data not shown). However, continued editing and updates to the initial procedure on re-utilized H&E index slides resulted in viable stain intensity, demonstrating the feasibility of the procedure and potential Methodsfor optimization Protoc. 2019, 2, 86to culminate in stain intensity comparable to that of sequential slides utilizing7 of 15 standard procedures (Figure 2).

Figure 2. Sequential slides of human prostate tissue exhibiting cancer (Ca) invading into normal glands Figure 2. Sequential slides of human prostate tissue exhibiting cancer (Ca) invading into normal and high-grade prostatic intraepithelial neoplasia (HGPIN). There is an aggressive carcinoma invading glands and high-grade prostatic intraepithelial neoplasia (HGPIN). There is an aggressive carcinoma glandular structures that have retained normal basal cells. The retained basal cells are positive for invading glandular structures that have retained normal basal cells. The retained basal cells are A B HMWCKpositive for+ p63 HM (WCK) (grey + p63 arrowheads), (A) (grey arrowheads positive) for, positive phosphatase for phosphatase tensin homolog tensin homolog (PTEN) ( (PTEN) (black) arrows),(B) (black and arrows negative), and for negative ETS-Related for ETS Gene-Related (ERG) Gene (C ()ERG (green) (C arrowheads).) (green arrowheads The cancer). The cancer is negative is fornegative HMWCK for HMWCK+ p63 (A )+ (redp63 ( arrowheads),A) (red arrowheads PTEN), (PTENB) (red (B arrows),) (red arrows but) positive, but posi fortive ERG for ERG (C) ( (blackC) arrowheads). [10 magniﬁcation]. (black arrowheads× ). [10× magnification].

TheThe initial initial testing procedures procedures resulted resulted in moderate in moderate but visible but visiblestain intensity stain intensity providing providing proof- proof-of-concept.of-concept. At this At stage, this stage, further further optimiz optimizationation and repeat and repeattesting testingwas warranted was warranted to increase to increasethe stain the stainintensity intensity to comparable to comparable levels levels of those of those occurring occurring using usingthe standard the standard antibody antibody staining staining methods methods and andto toensure ensure reproducibility. reproducibility. The The procedure procedure methods methods were were improved improved by byfour four steps: steps: 1. Applying 1. Applying timed timed reagent rinse procedures at the xylene, ethanol (EtOH), and Ventana/RTD proprietary reaction buffer steps (1-min rinse times between each hold); 2. Increasing EtOH and reaction buffer reagent rinses from 1 rinse to 5–6 and 3–4 manual rinses respectively for optimal efficiency; 3. Including an approximate 5-min drying step after the reaction buffer rinse to limit residual excess reagent interference in the online application of biomarkers; and 4. Editing online cell conditioning steps (for heat induced antigen retrieval) to reduce potential epitope destruction. This procedure optimization was considered the standard when applied to any H&E-stained slide stored for up to 2 years but needed further optimization for tissues stored for periods 2 years or longer. The subsequent experimentation steps employed the use of antibodies targeting PTEN and ERG biomarkers, VENTANA anti-PTEN (SP218) mouse monoclonal antibody, and anti-ERG (EPR3864) rabbit monoclonal antibody. These validated markers were used since they demonstrate 1. the heterogeneous variability of aggressive prostate cancer and 2. the comparative expression of PTEN loss and ERG expression in aggressive tumors. In this procedure, steps (1–8) represent the optimized H&E de-staining procedures. However, during the testing, unforeseen scheduling resulted in slight deviations (extended reagent HOLD times) in steps 4 and 7 that lead to the determination that certain steps, which were the xylene and reaction buffer HOLD times, could be amended without incurring damage to samples. The updated procedure, which only involved an extended xylene hold time and is essentially the same optimized procedure, resulted in comparable stain intensity to the standard protocol and allowed the ability of distinct determination of aggressive tumor areas (Figure3). After the successful completion of a sequential round of experimentation using IHC DAB, we tested whether antibodies targeting multiple biomarkers could be applied for detection with the use of chromogenic detection reagents. Again, prostate adenocarcinoma CNBs were used as experimental specimens for the de-stain and re-stain procedure. The antibodies chosen were specific for the integrin α6 (CD49f) laminin-binding domain and HMWCK + p63. These markers were chosen due to known membranous expression levels (CD49f) and cytoplasmic and nuclear (HMWCK + p63 respectively) positive expression levels in non-neoplastic basal cells of prostate tissue. In PCa, CD49f expression is membranous and aggressive and invasive disease exhibits an intracellular expression pattern [14–18]. It is also associated with poor patient prognosis, reduced survival, and increased metastasis [19–21]. These markers were not expected to colocalize but to demonstrate the expression pattern of both Methods Protoc. 2019, 2, 86 8 of 15 non-neoplastic and neoplastic structures and focal areas in tissue after marker application, following Methodsthe de-stain Protoc. 2019 procedure., 2, x FOR PEER REVIEW 8 of 15

Figure 3. Prostate cancer core needle biopsies (CNBs) sample H&E slides with PTEN and ERG IHC Figure 3. Prostate cancer core needle biopsies (CNBs) sample H&E slides with PTEN and ERG IHC DAB stained slides. The first slide for each sample was H&E-stained (A). Each sample H&E was DAB stained slides. The first slide for each sample was H&E-stained (A). Each sample H&E was de- de-stained and re-stained with either PTEN or ERG antibody depending on the pathologist analysis stained and re-stained with either PTEN or ERG antibody depending on the pathologist analysis for for biomarker loss or positivity to demonstrate tumor heterogeneity (B). The additional sequential biomarker loss or positivity to demonstrate tumor heterogeneity (B). The additional sequential slides slides for each sample were stained with anti-PTEN antibody or anti-ERG antibody (C). Each sample for each sample were stained with anti-PTEN antibody or anti-ERG antibody (C). Each sample H&E H&E was de-stained and re-stained with either PTEN or ERG antibody depending on the pathologist was de-stained and re-stained with either PTEN or ERG antibody depending on the pathologist analysis for biomarker loss or positivity to demonstrate tumor heterogeneity (10 magnification). analysis for biomarker loss or positivity to demonstrate tumor heterogeneity (10×× magnification). The expected outcome was to demonstrate definitive areas of non-neoplastic vs. tumor regions The expected outcome was to demonstrate definitive areas of non-neoplastic vs. tumor regions with the application of antibodies and chromogen detection. This would allow the simultaneous with the application of antibodies and chromogen detection. This would allow the simultaneous detection of normal and aggressive structures in one tissue sample after pathologist analysis of the detection of normal and aggressive structures in one tissue sample after pathologist analysis of the H&E-stained slide, allowing for the potential utilization of one slide. The results demonstrate strong H&E-stained slide, allowing for the potential utilization of one slide. The results demonstrate strong stain intensities for both targets and well-defined areas of demarcation of non-neoplastic vs. tumor stain intensities for both targets and well-defined areas of demarcation of non-neoplastic vs. tumor structures. As expected, both markers are visible in normal basal cells of normal prostate glands structures. As expected, both markers are visible in normal basal cells of normal prostate glands (although the HMWCK + p63 stain intensity primarily masks the CD49f signal in those areas), but (although the HMWCK + p63 stain intensity primarily masks the CD49f signal in those areas), but CD49f antibody clearly displays an intracellular and cytoplasmic expression in the areas of budding tumor (Figure 4).

Methods Protoc. 2019, 2, 86 9 of 15

CD49f antibody clearly displays an intracellular and cytoplasmic expression in the areas of budding tumorMethods (FigureProtoc. 20194). , 2, x FOR PEER REVIEW 9 of 15

Figure 4. Prostate adenocarcinoma sample CNB H&E and Chromogen IHC. The initial H&E-stained slideFigure with 4. Prostate the dotted adenocarcinoma line indicating prostaticsample CNB intraepithelial H&E and neoplasiaChromogen (PIN) IHC. lesion The andinitial tumor H&E area-stained (left panel)slide with (A). Thethe dotted de-stained line H&Eindicating that wasprostatic stained intraepithelial with HMWCK neoplasia+ p63 mouse(PIN) lesion monoclonal and tumor antibody area cocktail(left panel) and anti-CD49f (A). The de rabbit-stained polyclonal H&E that antibodies was stained using Dual with Chromogen HMWCK +detection p63 mouse (right monoclonal panel) (B). Theantibody anti-HMWCK cocktail and+p63 anti antibody-CD49f cocktailrabbit polyclonal (purple) stains antibodies the basal using cells Dual of normal Chromogen prostatic detection glands, (right and thepanel) anti-CD49f (B). The (teal) anti-HMWCK antibody stains+p63 antibody normal basal cocktail cell (purple) membranes stains (masked the basal by thecells HMWCK of normal+ p63)prostatic but demonstratesglands, and the an intracellular anti-CD49f and(teal) cytoplasmic antibody stains expression normal in aggressive basal cell membranestumors (area (masked demonstrating by the buddingHMWCK tumor +p63) outlined but demonstrates in H&E-stained an intracellular slide in the and left panelcytoplasmic and the expression right panel) in (20aggressivemagnification tumors × with(area 60 demonstratinginstep]). budding tumor outlined in H&E-stained slide in the left panel and the right × panel) (20× magnification with 60× instep]). These positive results from testing samples archived up to 2 years warranted the evaluation of the potentialThese positive ability results of this from procedure testing to samples be utilized archived with up other to 2 tissueyears warranted types, for H&Ethe evaluation stained slides of the archivedpotential 2abi yearslity of or this more, procedure and samples to be utilized archived with utilizing other tissue glass types, coverslips. for H&E Therefore, stained slides five archived archived PCa2 years CNBs or more, (2 years and 11 samples months), archived a normal utilizing colon glass (2 years coverslips. 1 month), Therefore, liver and five lung archived samples PCa (4 CNBs years) (2 alongyears with11 months), 4-plus yeara normal (4+) PCacolon resection (2 years (41 month), years 11 liver months) and lung H&E samples stained (4 sample years) slides along sealed with 4 with-plus thinyear film(4+) coverslipPCa resection were (4 tested. years 11 For months) the testing H&E ofsta H&Eined stainedsample slides slides sealed sealed with with thin glass film coverslips, coverslip archivedwere tested. PCa For CNBs the (2 testing years 11of H&E months), stained skin slides samples sealed (5 years) with andglass a coverslips, PCa TMA samplearchived (12 PCa years) CNBs were (2 tested.years 1 During1 months), the skin execution samples of this(5 years) procedure, and a thePCa removal TMA sample of the (12 coverslip years) waswere determined tested. During to be the a limitingexecution factor. of this Therefore, procedure the, parametersthe removal involved of the coverslip with the removal was determined was tracked to andbe a recorded limiting in factor this. reportTherefore (Table, the2). parameters After the removalinvolved of with the the H&E removal stain, thewas sample tracked slides and recorded were re-stained in this report with selected(Table 2). antibodiesAfter the removal utilizing of optimized the H&E stain, protocols the sample adapted slides for IHCwere onre- de-stainedstained with H&E selected slides antibodies (Table3). utilizing optimized protocols adapted for IHC on de-stained H&E slides (Table 3).

Table 2. Time tracking and coverslip parameters.

Sample Archive time Coverslip Removal time Type of Coverslip 1 month 10 min Thin film 1 year 10 min Thin film 2 year 10 min-60min Thin film 4 year ~38 hrs Thin film 4+ year** ~47 hrs Thin film 2+ year# 1-2 days Glass 5 year 4-5 days Glass 5 year 4-5 days Glass 12 year 4-5 days Glass Abbreviations: hrs, hours; ** Sample Archived 4 years 11 months; # Samples Archived 2 years 11 months ~Approximation due to time at removal.

Methods Protoc. 2019, 2, 86 10 of 15

Table 2. Time tracking and coverslip parameters.

Sample Archive Time Coverslip Removal Time Type of Coverslip 1 month 10 min Thin film 1 year 10 min Thin film 2 year 10 min–60 min Thin film 4 year ~38 h Thin film 4+ year ** ~47 h Thin film 2+ year # 1–2 days Glass 5 year 4–5 days Glass 5 year 4–5 days Glass 12 year 4–5 days Glass Abbreviations: h, hours; ** Sample Archived 4 years 11 months; # Samples Archived 2 years 11 months ~Approximation due to time at removal.

Table 3. Immunohistochemistry (IHC) Antibodies and Adapted Staining Protocols.

HMWCK CK 8 &18 Antibody p504s PTEN E-cadherin ERG + p63 (B22.1 CD49f (clone) (SP116) (SP218) (36) (EPR3864) (34βE12) &B23.1) mouse rabbit mouse rabbit mouse rabbit rabbit Species monoclonal monoclonal monoclonal monoclonal monoclonal polyclonal monoclonal Ventana Ventana Medical Medical Cell-Marque, Antibody Systems, Ventana Medical Systems, Inc., Tucson, Systems, Rocklin, N/A Vendor Inc., Arizona Inc., California Tucson, Tucson, Arizona Arizona De-stained H&E slide Immunohistochemistry adapted protocol IHC Ventana Benchmark ULTRA platform Ventana Detection Ventana OptiView DAB ultraView Ventana OptiView DAB IHC Kit IHC DAB IHC Deparaﬃn none 64 min 32 min 36 min 56 min 63 min 64 min 32 min HIER CC1 (pH CC1 (pH CC1 (pH CC1 (pH CC1 (pH CC1 (pH CC1 (pH 8.5) 8.5) 8.5) 8.5) 8.5) 8.5) 8.5) Blocking Peroxidase block Ab 36 C, 16 36 C, 32 37 C, 16 37 C, 16 36 C, 24 36 C, 24 36 C, 32 incubation ◦ ◦ ◦ ◦ ◦ ◦ ◦ min min min min min min min parameters Abbreviations: Ab, antibody; CC, cell conditioning; DAB, 3, 3’- diaminobenzidine HIER, heat-induced epitope retrieval; min, minutes.

The testing of the H&E stained slide samples archived 2 years or more involved extended coverslip removal and reagent rinse times (2+, 4, 4+, 5, and 12-year archived samples) which indicated that archival time, storage condition and coverslip type may play a factor in slide processing with the procedure. The processing of 2-year archived H&E stained slides sealed with thin film (all PCa CNBs) only required minimal extension time of coverslip removal (to ~60 min) but resulted in H&E stain removal and comparable antibody immunostaining intensities compared to the corresponding sequential slides (Figure5A–6E). The reused H&E and corresponding sequential slides were evaluated by a board-certified pathologist in a side by side comparison for immunostaining intensity (Table4). The histopathologic analysis focused on any present tumor or normal regions for intensity. The data Methods Protoc. 2019, 2, 86 11 of 15 analysis indicates that there was a significant matching in the immunostaining intensities between the reused H&E stained slides and sequential comparator slides immunostained with the various antibodiesMethods (Figure Protoc.5 2019F)., 2, x FOR PEER REVIEW 11 of 15

Figure 5.FigureH&E 5. stained H&E stained slide slide image image and and reused reused H&E slide slide selected selected antibody antibody immunostain immunostain comparison comparison with sequential slides in PCa CNBs. (A through 6E). H&E stained Slides (A–E). Antibodies: with sequential slides in PCa CNBs. (A through 6E). H&E stained Slides (A–E). Antibodies: HMWCK + HMWCK+p63 (1A–5A) (Note: uneven data points for Sequential slide 5 due to lack of immunostain p63 (1A–for5A HMWCK+p63) (Note: uneven); CK 8 data &18 points(1B–6B); for CD49f Sequential (1C–6C); E slide-cadherin 5 due (1D to–6D lack); ERG ofimmunostain (1E–6E). Scatter plot for HMWCK + p63); CKassessment 8 &18 ( 1Bof –side6B );by CD49f side comparison (1C–6C); of E-cadherin pathologist (analysis1D–6D scores); ERG and (1E comments–6E). Scatter for sequential plot assessment of side byslide side and comparison reused H&Eof antibody pathologist immunostaining. analysis scoresNote: the and red comments inverted triangles for sequential represent the slide ERG and reused H&E antibodyinternal immunostaining. controls and the open Note: inverted the triangred invertedles with black triangles outline represent represent the CD49f ERG internal internal controls controls (F). Data presented as SEM with Chi square, df (18.12, 1). The results determined by RM one- and the open inverted triangles with black outline represent the CD49f internal controls (F). Data presented as SEM with Chi square, df (18.12, 1). The results determined by RM one-way ANOVA matching across rows (see Table4) showing signiﬁcant matching with p value (< 0.0001). The analysis was performed using GraphPad Prism version 8.2.1. [Images 4 magniﬁcation]. × Methods Protoc. 2019, 2, 86 12 of 15

Table 4. Comparison of H&E re-used and sequential comparator immunostaining intensity scores.

Re-used H&E Initial H&E Marker Sequential Stain Intensity: 0–3 (int ctrl) Stain Intensity Assessment HMWCK + p63 3 2 2 2 0 3 atrophy inflammation Atrophy + CK8&18 3 3 3 3 3 3 inflammation: whole glands CD49f 3 (2) 3 (1) 3 (1) 3 (1) 3 (1) 3 (2) no cancer, int ctrls not much cancer but E-cadherin 3 3 3 3 3 3 weird well Methods Protoc. 2019, 2, x FOR PEER REVIEW differentiated 12 of 15 lots of infiltrating lymphocytes, grade 3 + wayERG ANOVA matching 3 (3) across rows 0 (3) (see Table 0 (3) 4) showing 0 (0) significant 1 (3) 0 (3)matching with p value (< 3 fragmented tumor 0.0001). The analysis was performed using GraphPad Prism version 8.2.1. [Imageslost 4× basal magnification]. cells, int ctrls Abbreviation: int ctrls, internal controls. The processing of the 4 and 4+ year archived H&E stained slides (PCa resections) sealed with thin Thefilm processingcoverslip required of the 4 andextended 4+ year time archived of coverslip H&E removal stained slides(~38 an (PCad 47 resections)h) but resulted sealed in withH&E thinstain film removal. coverslip The requiredarchived extendedH&E stained time slides of coverslip sealed with removal glass (~38 coverslip and 47 for h) 2+, but 5 resultedand 12-year in H&E (PCa stainresection, removal. two skin The and archived PCa TMA) H&E required stained slides1–2 days sealed and with4–5 days glass for coverslip coverslip for removal. 2+, 5 and The 12-year 4+ year (PCaarchived resection, sample two resulted skin and in PCa comparable TMA) required intensity 1–2 (CK days 8 and &18) 4–5 to days the sequential for coverslip comparator removal. slide The 4(+Figureyear archived6). The reused sample 12 resulted-year archived in comparable PCa TMA intensity H&E stained (CK 8 slide &18) immunostained to the sequential with comparator Ventana slideanti-ERG (Figure resulted6). The in reused immuno 12-yearstaining archived but PCaintensity TMA was H&E variable stained slide across immunostained different cores with but Ventanademonstrated anti-ERG feasibility resulted (data in immunostaining not shown). The but reused intensity 5-year was archived variable across H&E stained different slides cores (skin but demonstratedresections) sealed feasibility with glass(data coverslipsnot shown). also The required reused extended5-year archived time for H&E coverslip stained removal slides (skin and resections)reagent rinses. sealed The with resulting glass coverslips H&E stain also removal required exhibited extended residual time for H&E coverslip stain on removal the slides and resulting reagent rinses.in incomplete The resulting immunostaining H&E stain (ERG) removal (data exhibited not shown). residual The resulting H&E stain retention on the of slides the hematoxylin resulting in incompleteand eosin on immunostaining the slides may (ERG)have potentially (data not shown). impacted The the resulting results retentionand will need of the further hematoxylin inquiry and on eosinstorage on conditions the slides may to ascertain have potentially steps to impactedmitigate any the resultsissue. We and found will need that further the storage inquiry conditions on storage of conditionsolder H&E to stained ascertain slides steps to(particularly mitigate any with issue. glass We foundcoverslips) that the cause storaged extensive conditions adhesion of older of H&E the stainedcoverslip slides to the (particularly tissue slide with due glass to the coverslips) extended caused time in extensive storage, adhesion requiring of a the slight coverslip extension to the of tissueextraction slide procedures. due to the extended In addition time, inwe storage, observed requiring that pre a- slightanalytics extension impact ofed extraction H&E removal procedures., resulting In addition,in some we residual observed retention. that pre-analytics Unfortunately, impacted due H&E to age removal, of the resulting slide, pre in-analytica some residuall data retention. was not Unfortunately,available. due to age of the slide, pre-analytical data was not available.

FigureFigure 6. 6. TheThe H&EH&E stained stained slide slide archived archived 4 4++ years years subjected subjected to to de-stain de-stain and and re-stain re-stain procedure procedure comparedcompared to the the sequential sequential sample sample slide. slide. Initial Initial H&E H&E stained stained slide slidecontaining containing regionregion of tumor of ( tumorA). CK (A8 &18). CK antibody 8 &18 re antibody-stained re-stainedslide retaining slide region retaining of interest region and of exact interest architecture and exact (B). architecture Sequential slide (B). Sequentialcomparison slide immunostain comparisoned immunostained with CK 8 with&18 CK exhibi 8 &18ting exhibiting comparable comparable stain intensity stain intensity (C). [10(C).× [10 magniﬁcation]. magnification]× .

Methods Protoc. 2019, 2, 86 13 of 15

5. Discussion When patients are suspected of having PCa, a tissue sample is required for diagnosis. The sampling of the potentially neoplastic area may be assisted through means of ultrasound (US) or multiparametric magnetic resonance imaging (mpMRI) guided techniques. Sample resections and needle biopsies are routinely formalin-fixed and processed and embedded for histological sampling then stained for H&E and IHC, allowing pathologists to analyze an excised patient tissue sample from the affected area after diagnosis to differentiate between cancer and non-neoplastic events, such as benign prostatic hyperplasia. The H&E-stained slide plays a critical role in assisting the diagnosis of the pathologist in corroborating the initial findings with MRI and US procedures. Traditionally, after pathologist analysis and diagnosis, the samples can then be processed with biomarkers targeting detection of epitopes that are overexpressed in aggressive tumors. Currently this is the standard procedure deployed in companion diagnostics that allows for the stratification of patients who may benefit from a specific therapeutic intervention. The accurate evaluation of biomarkers with these samples is critical for patient diagnosis, particularly with smaller samples, such as CNBs, fine needle aspirates, and potentially transurethral resection of prostate samples (TURPS). The smaller size of these tissue samples limits tissue availability and requires precise testing for important results. Loss of available tissue slides is a risk that could be mitigated with the use of H&E slide de-stain and re-stain procedures. The potential to detect multiple markers using chromogenic multiplexing on a single indexed tissue slide that had been analyzed and diagnosed by a pathologist to definitively contain aggressive tumor, leaves open the possibility of predictive companion diagnostics with minimal sampling. This may provide the opportunity for a one sample/one result diagnosis limiting the invasive nature of tissue specimen collection, which benefits the patient greatly. There are few reports that provide instructions for removal of the H&E staining that leaves the target epitopes intact for potential reuse of the slide for selective biomarkers. Current existing protocols (and forums) only discuss de-stain procedures for slides that have stained inadequately, or have been stained with excessive hematoxylin and have lengthy protocol steps that may extend the procedure hours to days. Others may require the use of more corrosive reagents (% HCL solutions). Procedures utilizing either beta-mercaptoethanol/sodium dodecyl sulfate (2ME/SDS), 6 guanidinium hydrochloride (GnHCL) or 6 M Urea have been demonstrated to elute antibodies from immunostained tissues on positively charged glass slides (or glass coverslips) for sequential antibody re-stain [22,23]. However, these methods focused on the removal of the bound primary antibody and the reagents used were not intended to remove the H&E stain. For this report, an innovative method utilizing non-corrosive reagents was created and applied in a particular procedure using these reagents in sequence that optimized the H&E slide de-stain. This procedure removed the majority of the visible stain while retaining tissue integrity and morphology and allowed the preparation of specified IHC protocol to re-stain the sample sides. The primary tissue sample used for initial testing was prostate adenocarcinoma, however, this will translate to other tissues. This study utilized liver, colon, skin and PCa resections and CNB samples for procedure testing. The study included the addition of antibodies detecting clinically relevant biomarkers such as PTEN, ERG, E-cadherin, Racemase (p504s), cytokeratin 8 and 18 and the CD49f protein for potential indication of aggressiveness and antibodies against HMWCK cocktailed with a p63 marker (a p53 homologue containing the N-terminal transactivation domain) as well as the variant p40 marker (lacking the N-terminal domain), that will detect the presence of normal basal cells of prostatic glands. These antibodies were critical in detection of differentiating prostatic adenocarcinomas vs. the detection of non-neoplastic prostatic tissue, as well as determination of intracellular marker activity and basal cell attenuation, respectively. Moreover, during this study the positive outcome from testing various tissue samples archived beyond 4 years utilizing thin film and 12 years with glass coverslips yielded promising results indicating tissue epitopes remain stable on H&E stained slides archived at a minimum of 4 years. This indicates that the procedure may be useful for the interrogation of other clinically Methods Protoc. 2019, 2, 86 14 of 15 relevant proteins in tissues other than prostate and for H&E-stained slides stored for longer periods of time. However, the conditions of the slide storage and the type of adhesives applied to seal the slide may have an impact on results. Another factor may be the specific antibody selected for each specific study. The antibodies used for this study yielded promising results but each antibody demonstrates various qualities, therefore, continued optimization may be warranted for this procedure. Further experimentation will be repeated involving archived specimen slides utilizing film coverslips that have been stored for 4 years and more, as well as continued interrogation of samples sealed with glass coverslips. This will determine the robustness of the procedure to encompass reproducible testing of samples from decades past to incorporate newly discovered targets to test protein expression that may offer answers to questions that may have remained unsolved. Moreover, with the development of newer chromogen dyes, the possibility of utilizing one slide for multiple markers may now become a distinct possibility saving valuable time and resources. The results demonstrated in this report can be considered the first step towards a more extensive study incorporating much larger cohorts that may ultimately utilize this procedure as a viable tool in cancer diagnosis and treatments.

Author Contributions: Conceptualization, design, J.P.H., A.E.C., and R.B.N.; development of methodology, J.P.H.; experiments, J.P.H.; acquisition of sample tissue, J.P.H., R.B.N., H.A.T.; data analysis and interpretation, J.P.H., J.C.G., R.B.N.; writing-original draft preparation, J.P.H.; writing-review and/or manuscript editing, J.P.H., K.D., E.R., R.B.N.; administrative, technical, or material support, W.J.F., J.C.G., A.E.C., R.B.N.; study supervision, A.E.C.; R.B.N. Funding: Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under award number P30CA023074 and supported by Institutional Research Grant number IRG-16-124-37- IRG from the American Cancer Society. Acknowledgments: We thank the TACMASR core service of the University of Arizona Cancer Center for providing core needle biopsies samples. We would like to acknowledge the Ventana/RTD Integrated Core (iCore) services and Scott Gill for providing tissue sample resections for testing as well as Angela Schumacher and Adrian Murillo for providing reagents and chromogen detection kits. We would also like to thank William Harryman for editorial assistance. Conﬂicts of Interest: J.P.H., K.D., E.R., W.J.F. and J.C.G. are paid employees of Ventana/Roche Tissue Diagnostics. A.E.C, H.A.T. and R.B.N. have no conﬂicts to declare.

References

1. Bancroft, J.D.; Layton, C. The Hematoxylin and Eosin. In Theory Practice of Histological Techniques, 7th ed.; Suvarna, S.K., Layton, C., Bancroft, J.D., Eds.; Elsevier: Oxford, UK, 2013; pp. 173–186. [CrossRef] 2. Chan, J.K.C. The Wonderful Colors of the Hematoxylin–Eosin Stain in Diagnostic Surgical Pathology. Int. J. Surg. Pathol. 2014, 22, 12–32. [CrossRef] 3. Fischer, A.H.; Jacobson, K.A.; Rose, J.; Zeller, R. Hematoxylin and Eosin Staining of Tissue and Cell Sections. Cold Spring Harb. Protoc. 2008.[CrossRef] 4. Titford, M. The Long History of Hematoxylin. Biotech. Histochem. 2005, 80, 73–78. [CrossRef][PubMed] 5. Feldman, A.T.; Wolfe, D. Tissue Processing and Hematoxylin and Eosin Staining. In Histopathology: Methods and Protocols; Day, C.E., Ed.; Springer: New York, NY, USA, 2014; Volume 1180, pp. 31–43. [CrossRef] 6. Mukhopadhyay, S.; Feldman, M.D.; Abels, E.; Ashfaq, R.; Beltaifa, S.; Cacciabeve, N.G.; Cathro, H.P.; Cheng, L.; Cooper, K.; Dickey, G.E.; et al. Whole Slide Imaging Versus Microscopy for Primary Diagnosis in Surgical Pathology: A Multicenter Blinded Randomized Noninferiority Study of 1992 Cases (Pivotal Study). Am. J. Surg. Pathol. 2018, 42.[CrossRef][PubMed] 7. Zarella, M.D.; Bowman, D.; Aeﬀner, F.; Farahani, N.; Xthona, A.; Absar, S.F.; Parwani, A.; Bui, M.; Hartman, D.J. A Practical Guide to Whole Slide Imaging: A White Paper from the Digital Pathology Association. Arch. Pathol. Lab. Med. 2018, 143, 222–234. [CrossRef] 8. Ayala, G.; Frolov, A.; Chatterjee, D.; He, D.; Hilsenbeck, S.; Ittmann, M. Expression of ERG Protein in Prostate Cancer: Variability and Biological Correlates. Endocr. Relat. Cancer 2015, 22, 277–287. [CrossRef][PubMed] 9. Han, B.; Mehra, R.; Lonigro, R.J.; Wang, L.; Suleman, K.; Menon, A.; Palanisamy, N.; Tomlins, S.A.; Chinnaiyan, A.M.; Shah, R.B. Fluorescence in Situ Hybridization Study Shows Association of PTEN Deletion with ERG Rearrangement during Prostate Cancer Progression. Mod. Pathol. 2009, 22, 1083. [CrossRef] Methods Protoc. 2019, 2, 86 15 of 15

10. Jamaspishvili, T.; Berman, D.M.; Ross, A.E.; Scher, H.I.; De Marzo, A.M.; Squire, J.A.; Lotan, T.L. Clinical Implications of PTEN Loss in prostate Cancer. Nat. Rev. Urol. 2018, 15, 222–234. [CrossRef] 11. Leinonen, K.A.; Saramaki, O.R.; Furusato, B.; Kimura, T.; Takahashi, H.; Egawa, S.; Suzuki, H.; Keiger, K.; Ho Hahm, S.; Isaacs, W.B.; et al. Loss of PTEN is associated with aggressive behavior in ERG-Positive Prostate Cancer. Cancer Epidemiol. Biomark. Prev. 2013, 22, 2333–2344. [CrossRef] 12. Nagle, R.B.; Algotar, A.M.; Cortez, C.C.; Smith, K.; Jones, C.; Sathyanarayana, U.G.; Yun, S.; Riley, J.; Nagy, D.; Dittamore, R.; et al. ERG overexpression and PTEN status predict capsular penetration in prostate carcinoma. Prostate 2013, 73, 1233–1240. [CrossRef] 13. Yoshimoto, M.; Joshua, A.M.; Cunha, I.W.; Coudry, R.A.; Fonseca, F.P.; Ludkovski, O.; Zielenska, M.; Soares, F.A.; Squire, J.A. Absence of TMPRSS2: ERG Fusions and PTEN Losses in Prostate Cancer Is Associated with a Favorable Outcome. Mod. Pathol. 2008, 21, 1451–1460. [CrossRef][PubMed] 14. Ports, M.O.; Nagle, R.B.; Pond, G.D.; Cress, A.E. Extracellular Engagement of 6 Integrin Inhibited Urokinase-Type Plasminogen Activator-Mediated Cleavage and Delayed Human Prostate Bone Metastasis. Cancer Res. 2009, 69, 5007–5014. [CrossRef][PubMed] 15. Das, L.; Anderson, T.A.; Gard, J.M.C.; Sroka, I.C.; Strautman, S.R.; Nagle, R.B.; Morrissey, C.; Knudsen, B.S.; Cress, A.E. Characterization of Laminin Binding Integrin Internalization in Prostate Cancer Cells. J. Cell. Biochem. 2017, 118, 1038–1049. [CrossRef][PubMed] 16. Harryman, W.L.; Hinton, J.P.; Rubenstein, C.P.; Singh, P.; Nagle, R.B.; Parker, S.J.; Knudsen, B.S.; Cress, A.E. The Cohesive Metastasis Phenotype in Human Prostate Cancer. Biochim. Biophys. Acta 2016, 1866, 221–231. [CrossRef][PubMed] 17. Sroka, I.C.; Anderson, T.A.; McDaniel, K.M.; Nagle, R.B.; Gretzer, M.B.; Cress, A.E. The laminin binding integrin alpha6beta1 in prostate cancer perineural invasion. J. Cell. Physiol. 2010, 224, 283–288. [CrossRef] 18. Sroka, I.C.; Chopra, H.; Das, L.; Gard, J.M.; Nagle, R.B.; Cress, A.E. Schwann Cells Increase Prostate and Pancreatic Tumor Cell Invasion Using Laminin Binding A6 Integrin. J. Cell. Biochem. 2016, 117, 491–499. [CrossRef] 19. Friedrichs, K.; Ruiz, P.; Franke, F.; Gille, I.; Terpe, H.-J.; Imhof, B.A. High Expression Level of α6 Integrin in Human Breast Carcinoma Is Correlated with Reduced Survival. Cancer Res. 1995, 55, 901. 20. Landowski, T.H.; Gard, J.; Pond, E.; Pond, G.D.; Nagle, R.B.; Geﬀre, C.P.; Cress, A.E. Targeting Integrin 6 Stimulates Curative-Type Bone Metastasis Lesions in a Xenograft Model. Mol. Cancer Ther. 2014, 13, 1558–1566. [CrossRef] 21. Stewart, R.L.; West, D.; Wang, C.; Weiss, H.L.; Gal, T.; Durbin, E.B.; O’Connor, W.; Chen, M.; O’Connor, K.L. Elevated integrin α6β4 expression is associated with venous invasion and decreased overall survival in non-small cell lung cancer. Hum. Pathol. 2016, 54, 174–183. [CrossRef] 22. Bolognesi, M.M.; Manzoni, M.; Scalia, C.R.; Zannella, S.; Bosisio, F.M.; Faretta, M.; Cattoretti, G. Multiplex Staining by Sequential Immunostaining and Antibody Removal on Routine Tissue Sections. J. Histochem. Cytochem. 2017, 65, 431–444. [CrossRef] 9 23. Gendusa, R.; Scalia, C.R.; Buscone, S.; Cattoretti, G. Elution of High-aﬃnity (>10− KD) Antibodies from Tissue Sections: Clues to the Molecular Mechanism and Use in Sequential Immunostaining. J. Histochem. Cytochem. 2014, 62, 519–531. [CrossRef][PubMed]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).