HOPE-BAL

Abstract

The bronchoalveolar lavage (BAL) and its cells have been widely used as a support for clinical diagnosis and as a versatile tool for research questions since many years. Because there are no sufficient possibilities of long-term storage, the authors explore in this study the utility of a new fixative for fixation and paraffin embedding of human lavage cells with the possibility of implementing standard molecular biology techniques. HOPE-fixed, paraffin-embedded BAL cells of patients with different lung diseases (asthma, chronic obstructive pulmonary diseases, tuberculosis, sarcoidosis, emphysema, and fibrosis) were subjected to immunohistochemistry, in situ hybridization, quantitative polymerase chain reaction, and transcription microarray analysis. Furthermore, two-dimensional gel electrophoresis was conducted to evaluate the range of possible applications for research, diagnostics, and further implementing in biobanks. The authors show, by targeting some exemplary molecules, the power of screening and validating HOPE-BAL for new biomarkers. The transforming growth factor β signaling pathway may play a central role in immunomodulation upon infection as well as asthma. Furthermore, haptoglobin was overexpressed in asthma and sarcoidosis. Because of the excellent preservation of nucleic acids, protein, and morphologic structures, HOPE-BAL is a step forward into enhanced molecular diagnostics and biobanking in pulmonary medicine.

Starting as a treatment of alveolar proteinosis or other respiratory diseases featured by extensive accumulation of secretions found with cystic fibrosis, chronic asthmatic bronchitis, or bacterial pneumonia (Ramirez et al. 1965; Rogers et al. 1972), the application of bronchoalveolar lavage (BAL) has emerged as having both curative and diagnostic utility. With the use of smaller volumes of instilled saline, clinicians began in 1960s to study cellular functions of alveolar macrophages and surfactant material (Finley and Ladman 1972; Finley et al. 1967 and 1972) as well as immunoglobulins (Keimowitz 1964). First attempts using Carlen’s tube (Carlens 1949) or the Metras catheter (Metras 1953) were soon replaced by the more advanced technique of bronchofiberscope (Ikeda et al. 1968), which permitted routine sampling of cells and secretions accompanied by safe handling. In general, the right middle lobe (RML) and lingula are the most used sites for lavaging, which allow good retrieval of lavage fluid, and protocols range in employed aliquots from 20 to 60 ml in a single lavage site with a total of 120 to 250 ml (Meyer 2007). Cases in which the retrieved volume is less than 5% to 10% of instilled total volume are generally regarded as inadequate sampling and do not represent secretions from distal bronchoalveolar airspaces (Baughman et al. 2000). Common recommendations for BAL procedure and analysis have been published by the European Respiratory Society and the American Thoracic Society (Klech and Pohl 1989; American Thoracic Society 1990) to standardize the course of action as well as results.

The aspirated BAL fluid (BALF) and BAL cells are appropriate for a plethora of techniques, ranging from differential cell counts for diagnostics (Welker et al. 2004; Hauber and Zabel 2009), detection of mycobacteria (Chen et al. 2002), and other viral or bacterial pathogens (Meduri et al. 1992; Ioannas et al. 2001; Lee et al. 2006) as well as carcinoma cells (Springmeyer et al. 1983). Beside detection and counting of cells or pathogens, multiple endeavours aimed at discovering the proteome and cytokine milieu of BALF (Rottoli et al. 2005; Magi et al. 2006) and establishing databases of BALF proteins (Wattiez et al. 2000; Noel-Georis et al. 2002) have been pursued. The long-term storage of BAL cells for research and retrospective analysis as well as diagnosis is not satisfying and has not received much attention to date. We present in this study a novel technique for archiving BAL cells with implementation of modern molecular biological techniques and furthermore standardized analysis. Instead of historically popular formalin, we used the Hepes-glutamic acid buffer–mediated organic solvent protection effect (HOPE) as a new fixative (Olert et al. 2001) on BAL cells. The well-established HOPE technique, which does not cross-link proteins or degrade nucleic acids, has been shown to allow versatile analyses such as in situ hybridization (Goldmann et al. 2002), immunohistochemistry without antigen retrieval (Goldmann et al. 2003), Northern blotting, transcription microarrays (Goldmann et al. 2004), and Western blotting (Uhlig et al. 2004). Furthermore, HOPE-fixed tissues can be used for tissue microarrays (Goldmann et al. 2005) and laser-capture micro-dissection (Goldmann et al. 2006). Transferred and adapted to cells from human bronchoalveolar lavage, we present in this study a new approach for a diagnostic relevant tool.

Materials and Methods

BAL and Fixation of BAL Cells

The study has been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. All persons gave their informed consent prior to their inclusion in the study.

BAL samples were obtained by flexible bronchoscopy as described elsewhere (Gaede et al. 2004). In brief, 200 ml of sterile saline (0.9% NaCl) was instilled into the right middle lobe in 25-ml aliquots. Each aliquot was immediately aspirated, and the recovered aliquots were pooled. Furthermore, aliquots of each BAL were centrifuged for 10 min at 500 × g and washed three times with phosphate-buffered saline at 4C. Differential cell counts were determined using cytospin preparations of at least 400 cells (Cytospin II; Shandon Instruments, Sewickley, PA), which were stained with Hemacolor quick stain (Merck, Darmstadt, Germany). Viability was determined with Trypan blue solution (Sigma, Deisenhofen, Germany). Samples with viability less than 95% were discarded and not included in further processing. Preparations of each BAL sample were diluted to a standard cell number of 1 × 10⁶ and incubated in micro-centrifuge tubes with either HOPE fixative (DCS, Hamburg, Germany) for 16 hr at 4C or 4% neutral buffered formalin or stored at –80C.

Processing of HOPE-Fixed BAL Cells

After incubation in HOPE solution (Fig. 1A), cells were centrifuged at 8000 × g for 2 min, and the supernatant was discarded. For dehydration, 1 ml of acetone at 4C was added, and after 30 min of incubation, the cells were centrifuged and acetone was removed. This procedure was repeated five times (Fig. 1B). Following final removal of acetone, prewarmed low-melting paraffin (melting point 52–54 C; Polarit, DCS, Hamburg, Germany) was added and incubated overnight at 60C (Fig. 1C). Next day, the paraffin–cell solution was allowed to cool down at room temperature, and the upper part of the micro-centrifuge tube was cut off with a scalpel. The remaining lower micro-centrifuge tube part was cut laterally to extract the paraffin-embedded BAL cells (Fig. 1D). The resulting paraffin cone was submerged into prewarmed and precasted paraffin (Fig. 1E). HOPE-fixed BAL blocks were stored at a constant temperature of 4C. For quality control, samples from 15 patients were fixed with HOPE as well as with neutral buffered formalin (4%), respectively. The formalin-fixed cells were dehydrated after incubation in formalin in increasing concentrations of graded alcohol with final washing in xylene. Instead of low–melting point paraffin, formalin-fixed samples were embedded with normal paraffin (DCS). In addition, for addressing RNA preservation, fresh-frozen samples (−80C) were included.

Figure 1.

Schematic overview of processing and workflow of HOPE-BAL. HOPE, Hepes-glutamic acid buffer–mediated organic solvent protection effect; BAL, bronchoalveolar lavage.

Safety Issues with HOPE-Fixed Materials

HOPE reagents are based on Hepes-glutamic acid buffer–mediated organic solvent protection effect containing up to 0.03% NaN₃ (sodium azide). Sodium azide is not classified as hazardous at the concentration used in this product (guideline 1999/45/EG). However, toxicity information about sodium azide at the product’s concentration has not been thoroughly investigated. In contrast to other fixation methods, HOPE does not completely denature structural proteins, enzymes, and nucleic acids. This means that HOPE-fixed tissues could also include active virus, prions, microorganisms, and so on, although mycobacteria get killed during the acetone series (unpublished results). Therefore, HOPE-fixed materials should be considered potentially infectious, and thus it is critical for technicians to always wear gloves when handling HOPE-fixed materials (Olert et al. 2001).

Production of BAL Microarray

For enhanced/expanded and high-throughput analysis of multiple BAL samples, tissue microarrays of paraffin-embedded HOPE-BAL cells were produced as described elsewhere (Goldmann et al. 2005). In short, punches from 29 paraffin-embedded BAL blocks with a diameter of 2 mm were joined in a recipient block. BAL microarray (BMA) was used for immunohistochemistry (IHC) as well as for in situ hybridization (ISH) to ensure comparable treatment of specimens.

Immunocytochemistry

For immunocytochemical detection of target proteins, 1-µm-thick sections of BAL blocks or BMAs were cut on a microtome (SM 2000R; Leica, Wetzlar, Germany) and mounted on Super Frost+ cover slides (Menzel Gläser, Braunschweig, Germany). The BAL samples were deparaffinized by 10 min of incubation in isopropanol at 60C, followed by a short washing in fresh isopropanol at the same temperature. Deparaffinized sample slides were air-dried at room temperature and dehydrated for 10 min in 70% (v/v) acetone/DEPC-treated water at 4C. Remaining acetone was removed by incubation for 10 min in DEPC-treated water at 4C and transferred into distilled water at room temperature. Heat-induced antigen retrieval for formalin-fixed samples was conducted with citric acid buffer (pH 6) at 90C for 30 min using a steamer. Endogenous peroxidases were blocked for 10 min in 3% H₂O₂ solution. Primary antibodies were diluted as described in Table 1 with antibody diluent (Zytomed Systems, Berlin, Germany) and applied for 60 min in a moist chamber. For blocking and detection, a horseradish peroxidase (HRP)–conjugated polymer kit according to the manufacturer’s instructions (Zytomed Systems) was used. Washing steps were carried out three times for 5 min after each reagent step with washing buffer (50 mM Tris saline buffer with 0.1% [v/v] Tween 20; pH 7.6). Negative controls were included under omission of the secondary antibody. Permanent AEC (Permanent AEC Kit; Zytomed Systems) was used as substrate for HRP-conjugated polymer. Color reaction was stopped with distilled water. Samples were dehydrated in increasing concentrations of ethanol and washed for 20 sec in xylene, and cover slips were mounted using Pertex (Medite, Burgdorf, Germany) as mounting medium.

Table 1.

Overview of Used Antibodies and Dilution for Immunocytochemistry

Antigen	Clone	Producer/Distributor		Clonality	Host	Dilution
BAMBI	4e8	ebioscience	San Diego, CA, USA	Monoclonal	Mouse	1/100
phosphoSmad3	C25A9	Cell Signaling	Beverly, MA, USA	Monoclonal	Rabbit	1/200
TGF-β		abcam	Cambridge, UK	Polyclonal	Mouse	1/100
CD8	C8/144B	DakoCytomation	Glostrup, Denmark	Monoclonal	Mouse	1/50
CD4	MT310	DakoCytomation	Glostrup, Denmark	Monoclonal	Mouse	1/50
CD68	PG-M1	DakoCytomation	Glostrup, Denmark	Monoclonal	Mouse	1/100
Ki-67	MIB-1	Research Center Borstel	Borstel, Germany	Monoclonal	Mouse	1/3
LCA	PD7/26/16 + 2B11	NeoMarkers	Fremont, CA, USA	Monoclonal	Mouse	1/100
Pan-cytokeratin	MNF116	DakoCytomation	Glostrup, Denmark	Monoclonal	Mouse	1/100

BAMBI, BMP and activin membrane-bound inhibitor; TGF-β, transforming growth factor β1; LCA, leukocyte common antigen.

Extraction of Total RNA

For extraction of total RNA, the Rneasy Mini Kit (Qiagen, Hilden, Germany) was used. Tissues from BAL blocks were cut as for IHC and transferred into a micro-centrifuge tube. Samples were deparaffinized with xylene for 10 min (two times) and ethanol for 10 min (two times). The extract was dried with a vacuum centrifuge (SpeedVac, Savant, Farmingdale, NY, USA) and total RNA extracted according to manufacturer’s instructions with slight modifications: Cell lysate/70% ethanol solution was loaded two times on a mini-spin column. Eluted RNA was quantified and stored at −80C. RNA quality was analyzed on an Agilent Bioanalyzer with RNA 6000 Nano Chip assay (Agilent, Böblingen, Germany) according to the manufacturer’s instructions.

cDNA Synthesis

For cDNA synthesis, 1 µg of total RNA was subjected to DNAse I treatment with 1 µl of 10× DNAse I reaction buffer (Invitrogen, Karlsruhe, Germany) for 15 min at room temperature. All further incubation steps were carried out in a thermocycler (Uno, Biometra, Göttingen, Germany). DNAse I was further inactivated with 1 µl of 25 mM EDTA (Invitrogen) for 10 min at 65C. Following a short incubation on ice, 6.25 pM of Oligo dT15 primers (Eurofins MWG Operon, Ebersberg, Germany) was added to each sample for annealing to the poly-A-tail of mRNAs at 70C for 10 min and 25C for 10 min. Reverse transcription of mRNA to cDNA was conducted with Superscript II (Invitrogen) for 50 min at 42C according to the manufacturer’s instructions. Synthesized cDNA was stored until further analysis at −20C.

Quantitative RT-PCR

Quantitative RT-PCR was performed on the Roche LightCycler 480 system using the LightCycler 480 SYBR Green I Master (Roche Applied Science, Mannheim, Germany). Input cDNA was diluted 1:10 and PCR was performed in a total volume of 10 µl according to the manufacturer’s instruction with a final primer concentration of 0.5 µM for each primer. Primer-amplifying transforming growth factor β1 (TGF-β1) and proteasome subunit beta 2 (PSMB2) sequences were as summarized in Table 2.

Table 2.

Primer Sequences

Primer Designation	Sequence (5′-3′)
TGF-β1_for	CACGTGGAGCTGTACCAGAA
TGF-β1_rev	TGCAGTGTGTTATCCCTGCT
PSMB2_for	GGAGTCGGACCCCATATCA
PSMB2_rev	CTGAGAGTCAGGAAGGCACC
Bambi_for	CAGCTACATCTTCATCTGGC
Bambi_rev	AGAAGTCTAGAGAAGCAGGC

Human BMP and activin membrane-bound inhibitor (BAMBI) was amplified using the QuantiTect Primer Assay QT00091329 according to the manufacturer’s instructions (Qiagen). Cycling conditions were as follows: 40 cycles at 95C for 10 sec, touch-down annealing temperature (63–58C, temperature reduction 0.5C per cycle) for 4 sec, and 72C for 10 sec. All reactions were done in triplicate. Negative controls were also included to detect possible contaminations. Amplification specificity was checked using the melting curve following the manufacturer’s instructions. Relative quantification was done by establishing standard curves for each primer pair by serial dilution of a pooled cDNA sample. Specific mRNA levels were normalized to the level of the housekeeping gene PSMB2 in the same sample. In addition, a calibrator cDNA was included in every run to correct run-to-run differences.

For real-time experiments, cDNA from one case of unaffected lung, three cases of fibrosis, two cases of emphysema, three cases of sarcoidosis, one case of multi-drug-resistant tuberculosis, and one case of asthma was used. Standard deviation of values was included.

In Situ Hybridization

Probes for in situ hybridization were digoxigenin (DIG) labeled and hybridized as recently described (Umland et al. 2003). For targeting, mRNA of the BAMBI primer as summarized in Table 2 was used.

In negative controls, the respective probe was excluded and substituted with an aliquot of distilled water. In addition, non-sense probes were generated using a 527-bp fragment of a pBR322 Msp1 digest (New England Biolabs, Frankfurt am Main, Germany) or Lambda phage DNA (New England Biolabs) as a template for DIG labeling. Hybridization of sense and non-sense probes was conducted with 2 ng each probe overnight at 46C under sealed cover slips.

Transcription Microarray Analysis

Total RNA of 1 × 10⁶ BAL cells from patients with chronic obstructive pulmonary disease (COPD), tuberculosis, sarcoidosis, idiopathic pulmonary fibrosis (IPF), and emphysema was obtained as described above. Patients who underwent bronchoscopy due to chronic cough and retrospectively free of inflammatory or malignant disorders served as controls. All patients and controls gave their informed consent to the study. Information about differential cell counts, mean age, and sex is shown in Table 3. Total RNA was sent to a commercial company (ImaGenes, Berlin, Germany) for quality control, labeling, hybridization, and subsequent analysis on an Agilent Human Whole Genome 44K Array platform. A t-test was applied to quantil-normed signals, and a p-value of <0.05 was set as a threshold for significance. Significantly differential gene expression was determined as the log₂ ratio of each gene according to gene expression of the unaffected lung sample.

Table 3.

Sample Information of Bronchoalveolar Lavage Used for Transcription Microarray Analysis

			Differential Cell Counts (%)
Diagnosis	Age	Sex	Macrophages	Lymphocytes	Neutrophils	Eosinophils
Asthma I	51	F	85	12	2	1
Asthma II	53	F	64	6	27	3
Asthma III	31	F	6	1	93	0
COPD I	66	M	57	3	26	4
COPD II	79	F	36	3	61	0
COPD III	52	M	94	2	3	1
Emphysema	44	M	93	2	2	3
Unaffected lung I	82	F	79	18	2	1
Unaffected lung II	32	F	96	1	2	0
Fibrosis	76	M	89	9	2	0
Sarcoidosis	37	M	85	13	1	1
Tuberculosis	33	M	75	23	2	0

COPD, chronic obstructive pulmonary disease; F, female; M, male.

Two-Dimensional Gel Electrophoresis and Silver Staining

Prior to extraction of proteins, sections of BAL blocks were deparaffinized as for extraction of total RNA. On deparaffinized BAL cells, 1 ml of complex extraction buffer (7 M urea, 2 M thiourea, 100 mM dithiothreitol [DTT], 4% [w/v] Chaps, 2% [w/v] Igepal, 1% [w/v] Triton X-100, 5 mM PMSF, 0.5 mM EDTA, and 40 mM Tris) was added and lysed by the use of an electronic pestle (Kontes, Vineland, NJ) and incubated for 1 hr on a rotating device (Heto Lab Equipment, Allerod, Denmark). Amount of extracted protein was quantified using the Bradford assay (Coomassie Bradford Protein Assay Kit, Pierce, Rockford, IL).

Of each sample, 400 µg was desalted using ReadyPrep 2-D Cleanup Kit (Bio-Rad, Hercules, CA) according to the manufacturer’s instructions. The resulting pellet was dissolved in 155 µl of rehydration buffer (8 M urea, 2% [w/v] Chaps, 0.5% [v/v] Zoom Carrier Ampholyte [Invitrogen], 0.002% [w/v] bromophenol blue, 20 mM DTT, and 100 mM 2-hydroxyethyl disulfide). In-gel rehydration using the Zoom IPG Runner Cassettes and Zoom Strips pH 3-10 (Invitrogen) system was performed overnight. After rehydrating, isoelectric focusing of sample-loaded IPG strips was conducted by 20 min with 200 V, 15 min with 450 V, 15 min with 750 V, and finally 2000 V for 1 hr. Equilibration was applied as following: 2× 15 min in equilibration buffer (50 mM Tris, 6 M urea, 30% [w/v] glycerol and 2% [v/v] SDS) with either 1% (w/v) DTT and 2.5% (w/v) 2-iodoacetamide. Equilibrated IPG strips were placed on NuPage 4-12% Bis-Tris gels (Invitrogen), and 5 µl of molecular weight marker was added (Novex Sharp Pre-Stained Protein Standard; Invitrogen). Second-dimensional separation was achieved by applying 50 V for 10 min and 100 V until bromophenol-blue running front reached end of gel. Separated proteins were visualized by silver staining according to Yan et al. (2000).

Image Analysis and Figure Processing

All sample slides were analyzed on a microscope (DMLB2; Leica), and pictures were taken with a CCD camera (DFC 320; Leica). Contrast and brightness of images were adjusted using FixFoto software (Joachim Koopmann Software, Wrestedt, Germany). All figures and tables were produced by using the Microsoft Office Package 2010 (Microsoft, Redmond, WA), Adobe Creative Suite 5 (Adobe, San Jose, CA), or GraphPad Prism 4 (GraphPad Software, La Jolla, CA).

Results

HOPE-BAL allows application of common immunocytochemistry (ICC) markers and research-relevant molecules

To address the diagnostic value of HOPE-BAL cells, sections of BAL blocks were stained for CD4, CD8, CD68, leukocyte common antigen (LCA), Ki67, and pan-cytokeratin (Fig. 2). The results were compared with formalin-fixed and paraffin-embedded BAL cells from the same patient. All negative controls showed no signal (data not shown). Results revealed a good morphology compared to formalin-fixed material and allowed reliable differentiation of human immune cells in HOPE-fixed, paraffin-embedded lavage cells. Immune detection of the CD4 antigen was not possible in formalin-fixed BAL cells (Fig. 2A), whereas HOPE-fixed BAL cells showed clear staining of membranes from macrophages and lymphocytes (Fig. 2D). Staining for CD8, CD68, Ki67, and pan-cytokeratin displayed no serious differences between both fixation methods. Regarding LCA, signals revealed a slightly more intense signal in HOPE-fixed cells (Fig. 2J) compared to formalin-fixed material (Fig. 2G). During chromogenic reaction, the HOPE-fixed samples developed more early a distinct and robust staining.

Figure 2.

Immunocytochemistry comparing HOPE-fixed and formalin-fixed BAL samples (A–L) from the same patient with sarcoidosis as a clinical diagnosis. Immune staining was conducted to target CD4 (A, D), CD8 (B, E), CD68 (C, F), leukocyte common antigen (G, J), Ki67 (H, K), and pan-cytokeratin (I, L). The last row displays immune staining for Bambi (M), TGF-β (N), and pSmad3 (O) in HOPE-fixed BAL cells from a patient with idiopathic interstitial pneumonia. All images were taken at a magnification of ×400 (scale bar 50 µm) and using AEC as a chromogen (red signals). HOPE, Hepes-glutamic acid buffer–mediated organic solvent protection; BAL, bronchoalveolar lavage; BAMBI, BMP and activin membrane-bound inhibitor; TGF-β, transforming growth factor β1.

Besides standard ICC markers, sections were further analyzed for expression of TGF-β, phosphorylated Smad3 (pSmad3), and BAMBI. As illustrated in Figure 2M–O, mediators of TGF-β signaling can be addressed and targeted for protein-level pathway analysis in different cell types. The human TGF-β pseudoreceptor BAMBI showed clearly membranous staining pattern (Fig. 2M), and TGF-β is displayed on membranes and within the cytoplasm (Fig. 2N). In case of an activated signaling cascade, nuclear signals can be observed within the nucleus (Fig. 2O).

RNA is well preserved and suitable for ISH, qPCR, and transcriptome analysis

Preserving nucleic acids in human tissue or cell specimens is crucial for molecular biology–based techniques. To check the integrity of RNA in HOPE-BAL samples, RNA was extracted and compared to fresh-frozen and formalin-fixed material from the same patient. RNA samples were analyzed on an Agilent Bioanalyzer with RNA 6000 Nano Chip Assay for whole eukaryotic RNA. Comparison of the electropherograms revealed no detectable RNA from formalin-fixed samples and well-preserved RNA from fresh material. RNA from HOPE-fixed samples showed a lower RNA integrity than fresh samples (Fig. 3A). RNA gel-like image obtained by capillary electrophoresis (Fig. 3B) displayed no clear bands in formalin-fixed samples and distinct bands in fresh and HOPE-fixed samples. The yield of total RNA from either fresh (mean value of four different samples: 157 ng/µl) or HOPE-fixed (mean value of 4 different samples: 138 ng/µl) material resulted in comparable amounts. From formalin-fixed samples, only 3.2 ng/µl (mean value of 4 different samples) was extracted.

Figure 3.

Assessing of RNA quality from formalin-fixed (F), fresh-frozen (FM), and HOPE-fixed (H) BAL samples. Samples from four different patients were included; RNA was extracted and submitted to an Agilent Bioanalyzer RNA 6000 Nano whole RNA assay. Electropherograms of capillary chip electrophoresis are depicted in A and gel-like image in B. Corresponding molecular weight marker is shown as [nt] and the respective size. HOPE, Hepes-glutamic acid buffer–mediated organic solvent protection; BAL, bronchoalveolar lavage.

As RNA is preserved in sufficient quality, in situ hybridization targeting BAMBI in a BAL sample from a patient with idiopathic interstitial pneumonia was conducted. Signals for BAMBI mRNA can be mainly detected in the cytoplasm of alveolar macrophages (Fig. 4A). Omission of probe as well as applying non-sense probe (shown only for pBR322 Msp I–derived probe) resulted in no positive cytoplasmatic signals (Fig. 4A,B).

Figure 4.

In situ hybridization targeting transcripts of BAMBI (A) in a bronchoalveolar lavage sample from a patient with idiopathic interstitial pneumonia. Specificity of sense probe was tested with a non-sense probe (B). As a control for detection specificity, omission of probe was undertaken (C). All scale bars 50 µm. BAMBI, BMP and activin membrane-bound inhibitor.

HOPE-BAL cells were further used for establishing quantitative real-time PCR targeting BAMBI and TGF-β1 as well as PSMB2 as the housekeeping gene. Results were normalized to the expression of PSMB2 within the same sample (Fig. 5A,B). Although TGF-β1 mRNA levels are lowest in sarcoidosis, expression was unaltered in other lung diseases investigated. Transcript levels of BAMBI seem to be upregulated in lavages of patients with multi-drug-resistant tuberculosis as well as in cases of sarcoidosis and asthma (Fig. 5B).

Figure 5.

Quantitative real-time PCR analysis for the expression of TGF-β1 (A) and BAMBI (B) from samples of different lung diseases. BAMBI, BMP and activin membrane-bound inhibitor; TGF-β, transforming growth factor β1; MDR-TBC, multi-drug-resistant pulmonary tuberculosis.

To show that quantitative real-time PCR and in situ hybridization are applicable on HOPE-BAL cells, transcription microarray analysis was performed for detection of differential expressed genes. Samples from patients with different lung diseases as summarized in Table 2 were subjected to microarray analysis on an Agilent 44K array. The results are displayed as quantil-normalized signals. Total numbers of significantly regulated genes are shown in Table 4. To address reproducibility and consistency of data, the expression of 32,772 genes from two different cases of unaffected lung was correlated against each other. Correlation resulted in a coefficient of 0.81. Correlation of the transcriptome data from different diseases over 32,772 different genes resulted in coefficients ranging (Table 5) from 0.6 (COPD vs. tuberculosis and COPD vs. fibrosis) to 0.96 (COPD vs. asthma). Comparing and correlating the expression of 32 ribosomal genes from two different patient lavages resulted in homogeneous and almost equal gene expression and a correlation coefficient of 0.92 (Fig. 6). As the TGF-β signaling pathway plays a central role in various lung diseases, we further examined pathway signaling members on the transcriptional level in COPD, asthma, emphysema, fibrosis, tuberculosis, sarcoidosis, and the respective unaffected lung Fig. 7). The cytokine itself was solely expressed in elevated amounts in BAL cells from a patient with tuberculosis. Decreased expression could be found in the lavages of emphysema and fibrosis, respectively. Expression of TGFBR1-2 displayed throughout the diseases a homogeneous pattern comparable to the expression of the cytokine. The same held true for the expression pattern of Smad4 and Smad7. Interestingly, a reduced expression of Smad3 could be found within the transcriptome of sarcoidosis and an elevated expression in asthma. A high amount of variation of BAMBI expression in the analyzed diseases could not be observed.

Table 4.

Total Number of Significantly Regulated Genes

	Significantly Regulated Genes Compared to Unaffected Lung
	Total Number of Downregulated Genes (Log2)				Total Number of Upregulated Genes (Log2)
Diagnosis	<2.1	2.1–4.0	>4.1	Σ	<2.1	2.1–4.0	>4.1	Σ
Asthma	5645	643	21	6309	3989	5051	969	10,009
COPD	5895	945	10	6850	3629	4726	1800	10,155
Emphysema	3899	1158	40	5097	4418	1416	38	5872
Fibrosis	2833	687	39	3559	2644	571	56	3271
Sarcoidosis	2588	535	31	3154	2009	594	46	2649
Tuberculosis	3038	842	29	3909	2227	488	57	2772

COPD, chronic obstructive pulmonary disease.

Table 5.

Summary of Correlation Coefficients of Gene Expression Data (32,772 Individual Genes) from Different Transcriptomes

	Emphysema	Fibrosis	Sarcoidosis	Tuberculosis	Asthma
Fibrosis	0.62
Sarcoidosis	0.70	0.88
Tuberculosis	0.62	0.89	0.92
Asthma	0.91	0.65	0.72	0.64
COPD	0.89	0.60	0.66	0.60	0.96

In cases of asthma and unaffected lung, the mean values of expression data were used (each n=3). COPD, chronic obstructive pulmonary disease.

Figure 6.

Expression data of 30 ribosomal genes from two bronchoalveolar lavage of individual patients designated as having unaffected lungs.

Figure 7.

Relative expression levels of TGF-β pathway members in various lung diseases. Expression values of asthma, chronic obstructive pulmonary disease (COPD), and unaffected lung are displayed as mean values. TGF-β pseudo-receptor BAMBI as well as Smad7 are shown as examples for negative regulators of TGF-β signaling. Smad3/Smad4 display the canonical intracellular signaling members and the cytokine TGF-β1 as well as its corresponding receptors (TGBFR1-3). BAMBI, BMP and activin membrane-bound inhibitor; TGF-β, transforming growth factor β1.

Proteomics with HOPE-BAL

Finally, to address protein-based applications, total protein of BAL cells from patients with sarcoidosis and IPF was extracted. Two-dimensional gel electrophoresis was applied to separate extracted proteins. Subsequent silver staining was performed for visualization of both proteomes (Fig. 8). Although both gels show distinct and comparable signals, a few differences have to be verified in further experiments.

Figure 8.

Two-dimensional gel electrophoresis with 200 µg total protein extracted from bronchoalveolar lavage from cases of fibrosis (A) and sarcoidosis (B). MW, molecular weight marker Novex Sharp Pre-Stained Protein Standard (Invitrogen, Karlsruhe, Germany).

Discussion

Archiving tissues and specimens of different diseases for long-term storage implies the need for a wide range of applicable techniques. Biobanks encounter difficulties in choosing the appropriate fixative for their samples because it is not known what will be needed in the future. The BAL has proven its usefulness in both diagnostics and research. Because the BAL allows a more focused view/glance at the level of cellular responses and characteristics during pathogenesis, it offers also a substantial supplement for tissue-based research. To our knowledge, there is to date no satisfying method for storage of BAL. Most lavage cells are analyzed as cytospin preparations and directly fixed on cover slides. Although there have been studies comparing fixation and handling procedures of these cytospin preparations (Xaubet et al. 1991; Moumouni et al. 1994), there is still no notable method for long-term storage of BAL. Therefore, this study presents a novel and simple method for fixation of BAL cells that allows a plethora of applications besides long-term storage.

TGF-β signaling in transcriptome of HOPE-BAL compared to human lung tissues

The TGF-β pseudoreceptor BAMBI was shown to be overexpressed in lung tissues of COPD patients and is strongly regulated upon infection with non-typeable Haemophilus influenzae (NTHI; Drömann et al. 2010). Comparing the results of transcriptome data obtained from infected lung tissues with ours (Fig. 7b), the mRNA expression levels are generally lower. This might be because BAMBI is expressed not only in macrophages but also in alveolar epithelial cells type II, which more or less account for 60% of alveolar tissue (Fehrenbach 2001). Drömann et al. (2010) showed an upregulation of BAMBI in BAL cells by NTHI. In addition to the effect of NTHI on expression of BAMBI in human lung tissue and BAL, we were able to show upregulation in BAL of one patient with multi-drug-resistant mycobacterium tuberculosis infection (Fig. 5B). Because TGF-β is thought to play a central role in asthma, COPD, and pulmonary fibrosis (Araya and Nishimura 2010) and is a potent immune regulatory cytokine (Li et al. 2006), analysis of BAMBI may lead to novel insights into host–pathogen reactions as well as tissue homeostasis.

The chances of HOPE-BAL

We have shown that fixation of BAL with HOPE allows a preservation of RNA and proteins suitable for molecular-based applications such as in situ hybridization, quantitative real-time PCR, transcription microarray analysis, and two-dimensional gel electrophoresis. A further advantage is the long-term storability of HOPE-BAL, which can preserve biomolecules significantly longer than material that is frozen at −80C. Furthermore, a variety of commonly used ICC markers can be applied for diagnostic analysis. It was previously shown (Kähler et al. 2010) that the HOPE technique allows two-dimensional separation of proteins in paraffin-embedded tissue samples, with further validation by mass spectrometry and protein fingerprints. This enhances the possibilities of paraffin-embedded BAL because all conducted studies used either fresh lavages or only BALF (Wattiez et al. 2000). As a whole, HOPE-BAL combines easy handling in the form of paraffin blocks with almost no limitations in readout techniques. Out of this, one can search for biomarkers on different levels and build collections of BAL. In this study, 1 × 10⁶ cells were used for fixation and paraffin embedding. However, we would recommend using more cells (3–4 × 10⁶). This will increase the yield of nucleic acids and proteins and will also allow more applications from the same block.