Serum proteomic profiling and haptoglobin polymorphisms in patients with GVHD after allogeneic hematopoietic cell transplantation

We studied serum proteomic profiling in patients with graft versus host disease (GVHD) after allogeneic hematopoietic cell transplantation (allo-HCT) by two-dimensional gel electrophoresis (2-DE) and mass spectrometry analysis. The expression of a group of proteins, haptoglobin (Hp), alpha-1-antitrypsin, apolipoprotein A-IV, serum paraoxonase and Zn-alpha-glycoprotein were increased and the proteins, clusterin precursor, alpha-2-macroglobulin, serum amyloid protein precursor, sex hormone-binding globulin, serotransferrin and complement C4 were decreased in patients with extensive chronic GVHD (cGVHD). Serum haptoglobin (Hp) levels in patients with cGVHD were demonstrated to be statistically higher than in patients without cGVHD and normal controls (p < 0.01). We used immunoblotting and PCR in combination with 2-DE gel image analysis to determine Hp polymorphisms in 25 allo-HCT patients and 16 normal donors. The results demonstrate that patients with cGVHD had a higher incidence of HP 2-2 phenotype (43.8%), in comparison to the patients without cGVHD (0%) and normal donors (18.7%), suggesting the possibility that specific Hp polymorphism may play a role in the development of cGVHD after allo-HCT. In this study, quantitative serum Hp levels were shown to be related to cGVHD development. Further, the data suggest the possibility that specific Hp polymorphisms may be associated with cGVHD development and warrant further investigation.


Introduction
Allogeneic hematopoietic cell transplantation (Allo-HCT) has been a potentially curative treatment approach for patients with hematological malignancies, lymphohematopoietic failure, autoimmune diseases as well as genetic disorders. Despite its curative potential, the application of allo-HCT is limited by life-threatening complications, in particular, graft-versus-host disease (GVHD), a highly morbid toxic complication [1,2]. As a clinical syndrome related to the reaction of donor-derived immunocompetent cells against patient tissues, GVHD remains the most frequent transplant-related complication.
GVHD is classified as a clinicopathologic syndrome involving skin, liver, gastrointestinal tract, and/or other organs. Currently, there are no reliable laboratory tests that will confirm or refute its presence. Thus, GVHD is mostly a clinical diagnosis. Diagnosis of GVHD requires an interpretation of clinical and laboratory findings, recognizing that in some patients the differential diagnosis may be difficult to resolve [3]. To predict development and clinical prognosis of GVHD, several in vitro tests have been described. However, results have been difficult to reproduce and no assay has been widely adopted [3][4][5][6][7]. Studies of certain cytokine gene polymorphisms, includ- ing tumor necrosis factor alpha, interferon gamma, interleukin-1 (IL-1), IL-6 and IL-10, as well as polymorphisms of certain adhesive molecules such as CD31 and CD54 have been extensively conducted to explore their potential for GVHD risk prediction and the development of predictable genetic risk indexes. However, these efforts have not yet resulted in reliable models [3,[8][9][10][11][12][13].
Over the past decade, the study of proteomics has rapidly evolved and developed. Proteomics studies can generate protein expression profiles which may predict clinical events, therapeutic response, or probe underlying mechanisms of disease. Proteome analysis is emerging as an important technology for understanding biological processes and discovery of novel biomarkers in diseases such as autoimmune disorders, cardiovascular diseases and cancers [14][15][16][17]. A recent study used an intact-proteinbased quantitative analysis system for determining the plasma proteome profile of patients with acute GVHD after transplant. The proteins, including amyloid A, apolipoproteins A-I/A-IV and complement C3 were found to be quantitatively different between the pre-and post-GVHD samples [18]. In another report, several differentially excreted polypeptides were identified from patient urine samples by a capillary electrophoresis and mass spectrometry (CE-MS) based technique. The peptide profile displayed a pattern of early GVHD markers, allowing discrimination of GVHD from patients without the complication [19]. These reports hinted that GVHD can be monitored by changes in protein expression patterns detectable through proteomic methods.
Few investigations utilizing proteomic profiling in the study of patients with and without GVHD after allo-HCT have been reported to date. Several contributions in this regard have recently been reported to be confirmatory of a clinical diagnosis of acute GVHD (aGVHD) and to provide prognostic information. Paczesny et al have developed a panel consistent of 4 biomarkers which both confirm the diagnosis of aGVHD at onset of clinical symptoms and provide prognostic information independent of aGVHD severity [20]. Weissinger, et al have described an aGVHD-specific model consisting of 31 polypeptides and Hori et al have correlated a member of a large chemokine family, CCL8 to be closely correlated with aGVHD severity through proteomic analysis [21,22].
In this study, we performed serum proteomic profiling in a group of patients with and without cGVHD after allo-HCT by 2-dimensional electrophoreses (2-DE) and mass spectrometry based technology. Differential expression patterns of 11 serum proteins were demonstrated in patients before and after cGVHD development. Serum Hp precursors, one of the 11 differentially expressed serum proteins, were found to be significantly up-regulated during cGVHD development. We also investigated the relationship between serum Hp quantity as well as Hp polymorphisms and cGVHD development in this study. Serum Hp level as well as its polymorphisms were shown to be related to cGVHD development. Thus, Hp might serve as a worthy future target for monitoring cGVHD and understanding cGVHD mechanism.
A paired 2-DE gel images from a patient with cGVHD  Sixteen normal healthy donors, 10 males and 6 females, were included as controls in this study. Median age of the normal donors was 38 years (range 20-55).

Serum processing
Peripheral blood samples were obtained, with informed consent, during routine diagnostic blood studies from  patients before and after allo-HSCT in Saint Luke's Cancer Institute (SLCI). Serum samples were collected and aliquoted into 300 μl per tube from whole blood specimens, allowed to stand over night at 4°C without anti-coagulant and stored at -80°C in a freezer. Mononuclear cells (MNC) were isolated by Ficoll Hypaque density gradient separation and cyropreserved in liquid nitrogen. Serum albumin was removed by Swellgel Blue Albumin Removal kit (Pierce biotechnology, Rockford IL) following the instructions provided by the kit. 150 ul of serum was loaded for each single reaction. After preparing resin disc, binding sample and washing to release albumin-free sample, the albumin-free serum was collected for determination of protein concentration. The protein concentration was determined by the D c Protein Assay kit (Bio-Rad, Hercules CA) and following the instructions provided by the kit.

Silver staining
Silver staining was performed according to a protocol published previously [23]. Briefly, gels were fixed with 50% methanol/10% acetic acid for 30 minutes and 5% methanol/1% acetic acid for 15 minutes respectively, and then the gels were washed by distilled water 3 times for 10 The silver nitrate was then discarded and the gels were rinsed with distilled water 3 times for 1 minute each time and then developed with 0.02% formaldehyde and 0.0004% sodium thiosulphate in 6% sodium carbonate with shaking. The development was terminated with 6% acetic acid.

Image analysis
The silver-stained 2-DE gels were scanned with LabScan software on an UMAX Powerlook III scanner (UMAX Tech Inc, California), and the images were digitalized and analyzed with a α-GelFox 2D 3.1 (Alpha Innotech) software.

In-gel digestion and protein identification by LC-MS/MS
Protein spots were cut out from the silver-stained gels for in-gel digestion. Proteins were reduced and alkylated before digestion with trypsin (Promega, Madison, WI) overnight at 37°C. The peptides were extracted from the gel and concentrated in a vacuum centrifuge. 8 μL of concentrated peptide mixtures was injected to an Agilent LC-MSD ion trap mass spectrometer (Agilent Technologies, USA) for identification. Mass spectra were acquired in positive-ion mode with automated data-dependent MS/ MS on the four most intense ions from precursor MS scans. The mass spectra were extracted and searched against the human database using Mascot software.

Haptoglobin genotype determination by PCR
Genomic DNA was extracted from peripheral blood MNC by the QIAamp DNA Kit as suggested by the supplier (Qiagen). Oligonucleotide primers A (5'-GAG-GGGAGCTTGCCTTTCCATTG-3') and B (5'-GAGATTTTT-GAGCCCTGGCTGGT-3') were used for amplification of a 1757-bp Hp-1 allele-specific sequence and a 3481-bp Hp-2 allele-specific sequence. Primers C (5'-CCTGCCTCG-TATTAACTGCACCAT-3') and D (5'-CCGAGTGCTCCA-CATAGCC ATGT-3') were used to amplify a 349-bp Hp-2 allele-specific sequence [24]. The oligonucleotide primers were synthesized by IDT, Inc (Coralville IA). 20-μL reactions contained 2 U of Taq polymerase (Promega), 1-100 ng of DNA, and 200 μM each of dATP, dCTP, dGTP and dTTP (Promega); PCR buffer was used as suggested by the supplier (Promega) with no supplements added. After initial denaturation at 95°C for 2 min, the two-step thermocycling procedure consisted of denaturation at 95°C for 1 min and annealing and extension at 69°C for 2 min (in the presence of primers A and B or primers A, B, C, and D) or 1 min (in the presence of primers C and D only), repeated for 35 cycles, and followed by a final extension at 72°C for 7 min. The thermocycler used was Perkin Elmer 480 PCR system. For genotype assignments, the PCR products where primers A and B were used were separated in 1% agarose gels and products where primers C and D were used were separated in 8% polyacrylamide gels.
Restriction enzyme analysis was performed to verify the identity of Hp-1-and Hp-2-specific PCR products. The 1757-and 3481-bp products were digested with restriction enzyme MlsI, and the 349-bp product was digested with DraI, as recommended by the supplier (MBI Fermentas). DNA fragments were separated by gel electrophoresis.

Immunoblot
Immunoblot were performed as previously published with modifications [25].  and then incubated at 4°C overnight with a 1:1000 dilution of polyclonal rabbit anti-human Hp antibody (Sigma). After washing the membranes three times in TBST, a second antibody, anti-rabbit IgG horse radish peroxidase conjugate (Santa Crus, California) was used at a dilution 1:2000 in TBST; the membranes were then incubated at room temperature for 1 h. The membranes were washed three times in TBST and finally developed with Luminal Reagent (Santa Crus).

Statistical analysis
Statistical comparison of the serum Hp levels and 2-DE gel Hp protein spots VI among the different study groups was done using the Waller-Duncan K-ratio t test.

Changes of protein expression patterns before and after cGVHD development
Thirty-six serum specimens derived from 16 patients and 9 samples from normal donors were subjected to 2-DE gel and silver staining assays. Out of the 36 patient-derived samples, 13 were collected before transplantation. Out of the 23 patient derived samples collected after allo-HCT, 9 samples were collected prior to the cGVHD occurring and 9 during cGVHD development, 5 were from the patients with no cGVHD development at all after receiving allo-HCT. The median number of protein spots in 2-DE gels was 709 (range 499-1012 spots) for the specimens from patients pre-transplantation; 637 (458-806 spots) in samples from prior to or no cGVHD development patients; 760 (508-1031 spots) in the serum from patients with active cGVHD and 735 (645-783 spots) in the serum of normal donors.
Paired 2-DE gel analyses were performed using 2-DE gel software between the serum collected prior to cGVHD and the samples of active cGVHD phase from the same individual patient. Protein spot patterns were significantly different between the gels of pre-cGVHD and active cGVHD phase in the same patient. The median protein spot numbers were determined as 753 (range 610-1012) and 726 (508-1031) for the specimens pre and post-cGVHD in the 7 patients with cGVHD, respectively. Median of matched spots between paired gels was 501 (405-725, median and range), however, in comparison to the gels pre-cGVHD, the medians for missing spots and newly appeared spots were 248 (190-391) and 276 (117-373), respectively in the gels of cGVHD.
A group of protein spots in the 2-DE gels was found significantly and consistently different between the serum collected prior to and during the cGVHD development from various patients. The protein spots, which were collected and analyzed from a paired representative 2-DE gels by LC-MS/MS are demonstrated in Figure 1. The sum of the spot areas multiple spot density as the volume index (VI) was used to determine the individual serum protein level in 2-DE gel semi-quantitatively. Five proteins, including Hp, apolipoprotein A-IV, α-1-antitrypsin, serum paraoxonase and Zn-α-glycoprotein were found quantitatively Comparison of serum Hp level in patient with and without GVHD after allo-HCT by Elisa assay Figure 4 Comparison of serum Hp level in patient with and without GVHD after allo-HCT by Elisa assay. Serum Hp level in GVHD group is significantly higher than all the other 3 groups (p < 0.01) increased or newly appearing after cGVHD development.
A group of 5 proteins were either down-regulated or absent in the 2-DE gel of patient with cGVHD, including sex hormone-binding globulin, clusterin precursor, serum amyloid A protein precursor, serotransferrin and complement C4 ( Table 2). As a representative of protein spots analyzed, the mass spectrum identification of Hp by LC-MS/MS analysis is shown in Figure 2.

Differential expression patterns of haptoglobin in cGVHD development
As a particular example, Hp expression patterns were demonstrated to be significantly different among the patients with and without cGVHD development in our 2-DE gel analysis. To quantitatively compare the differences in Hp spot volumes between the different study groups by 2-DE gel image analyses, we used VI to determine the serum Hp level semi-quantitatively. Since the complexity of Hp α chain polymorphisms and that the Hp β chain are identical in all haptoglobin phenotypes, we selected the VI of Hp precursor β as the representative of Hp volumes in each study groups. The VIs in the cGVHD group (n = 6), was demonstrated to be significantly higher than the VIs of the patients before transplantation (n = 12; p = 0.027); patients with no cGVHD after transplantation (n = 4; p = 0.045) and normal donor group (n = 9; p = 0.044), respectively ( Table 3).

Serum Hp levels in the patients before and after cGVHD development
We performed Elisa assays to confirm the prior finding and compared serum Hp levels in patients before and after cGVHD. Hp levels were examined in the serum from the same patient before and after cGVHD development, as well as normal donors. The mean of Hp concentration was 1.97 ± 0.99 mg/ml (mean ± SD) in the patients with cGVHD (n = 14); 0.83 ± 0.40 in the patients before transplantation (n = 23); 0.74 ± 0.51 in the patients with no Serum Hp phenotype determination of patients by immuoblot . Statistical analysis demonstrated that the serum Hp level in the cGVHD group was significantly higher than all the other 3 groups (p < 0.01). The differences of the serum Hp level were insignificant between the 3 cGVHD-negative groups ( Figure 4).

Determination of Hp polymorphisms in the patients
In humans, Hp is characterized by a molecular heterogeneity with three main genotypes/phenotypes: Hp 1-1, Hp 2-1, and Hp 2-2. These different proteins have distinctive efficiencies and it has been suggested that the polymorphism may have important biological consequences in several diseases [26]. To probe the possible relationship between cGVHD and Hp polymorphism, we used immunoblot and PCR in combination with 2-DE gel image analysis to determine Hp polymorphisms in the 24 allo-HCT patients of this study group as well as 12 normal donors. Serum Hp phenotype was determined by immunoblot using polyclonal antibody, which recognize Hp β, α-1 and α-2 chains and the Hp phenotypes in the patients are shown in Figure 5. To verify the Hp phenotype in these patients, we examined Hp genotypes of the patients by PCR assay (figure 6), and the Hp PCR products were fur-ther confirmed by restriction enzyme analysis (figure 7 and 8). All the DNA typing of the samples derived from 25 patients were matched with their phenotypes determined by immunoblot.

Discussion
Proteome analysis is now emerging as an important technology for deciphering biological processes and is aiding in the discovery of biomarkers for diseases from tissues and body fluids. In this study, we examined serum proteomic profiles in a group of patients with cGVHD after allo-HCT by two-dimensional gel electrophoresis (2-DE) and mass spectrometry based technology. A panel of proteins, Hp alpha-1-antitrypsin, apolipoprotein A-IV, serum paraoxonase and Zn-alpha-glycoprotein were demonstrated to be up-regulated and clusterin precursor, alpha-2-macroglobulin, serum amyloid protein precursor, sex Analysis of DNA amplification products representing geno-types with restriction enzymes MlsI  hormone-binding globulin, serotransferrin and complement 4 were found down-regulated in the patients with cGVHD.
Medical literature in the area of proteomic profiling in GVHD is scarce [18,19,27]. One study has used an intactprotein-based quantitative analysis combined with protein tagging and immunodepletion of abundant proteins to quantitatively profile the plasma proteome in the patients with acute GVHD after transplant [18]. In this study, plasma samples were subjected to immunodepletion chromatography to remove six of the most-abundant plasma proteins (albumin, transferrin, IgG, IgA, Hp and α-1-antitrypsin) to increase the sensitivity of serum low abundant protein detection. However, it is not clear whether or not serum high abundant proteins such as Hp, transferrin and immunoglobin, est., are involved in the pathophysiology of cGVHD. In our study, high abundant proteins including Hp, alpha-1-antitrypsin and transferrin exhibited quantitative differences between the pre-and post-GVHD samples, which suggest that those proteins might be importantly involved in the pathophysiologic processes of cGVHD. Therefore, the potential role of these high abundant proteins in the development and propagation of cGVHD should be fully assessed before being methodically eliminated in proteomic profiling studies. Increased serum Zn-alpha-glycoprotein and decreased complement C4 in patients with cGVHD in our study were in agreement with this report [18]. In contrast, serum amyloid protein and alpha-2-macroglobulin, which were increased in their study, were down-regulated in our study [18]. One possible explanation might be that the immunodepletion process affected their results.
In our study, Hp was identified as one of the increased proteins after cGVHD onset. Both the results of Hp volume index in 2-DE gel image analysis and serum Elisa assay demonstrated a significant increase of Hp in patients with cGVHD. Hp is an acute-phase response serum protein that has been known to play an important inhibitory role in inflammation and the Hp plasma concentration may increase in response to a variety of stimuli, such as: infection, neoplasia, and other inflammatory and immune reactions [28][29][30]. In this study, we report for the first time that an increase of serum Hp concentration is observed in patients with cGVHD after allo-HCT. The quantitative changes of serum Hp, as well as the well known acute-phase reactants found in this study, such as apolipoproteins A-IV, complement C4 and serum amyloid A thus might reflect changes in these proteins as manifestation of their roles in the pathophysiologic development and propagation of cGVHD or, alternatively, simply a nonspecific manifestation of an inflammatory state. Other investigators have described results that differ from the data reported here in terms of some acute phase reactants such as apolipoproteins in the setting of GVHD [31]. However, these data were derived from patients undergoing cord blood transplantation in contrast to our data set which is derived from patients receiving only adult derived hematopoietic stem cell transplantation. Additionally, the samples were collected within the first 100 days of transplant, before cGVHD could have developed in the report of Harvey, et al. Finally, the subtype of apolipoprotein measured differed from our study. Hp levels may increase in response to various stimuli, a further well designed study with more cases included would be necessary for ruling out the Hp changes secondary to transplantation-related infection, lung injury, and other possible complications post-trans- plant. Additionally, haptoglobin levels may be significantly changed in patients experiencing GVHD associated microangiopathy [32]. Although, no patient in our data set met criteria for transplant associated microangiopathy, this represents a well described GVHD associated clinical syndrome and will need to be closely evaluated in future studies concerning the association of haptoglobin and cGVHD.
Hp is characterized by molecular heterogeneity with three major phenotypes: Hp 1-1, Hp 2-2, and the heterozygous Hp 2-1 [33][34][35]. Hp is synthesized as a single polypeptide chain and is proteolytically cleaved to a short α-chain and a long β-chain that remains connected through a disulfide bond. Although Hp is found in serum of all mammals, this polymorphism exists only in humans [36,37]. These Hp phenotypes have different biologic activities, which include a stronger anti-oxidation, hemoglobin banding and anti-·OH production activities derived from Hp 1-1 and otherwise a stronger activity of macrophage activation for Hp 2-2 [23,27,36,37]. The functional differences between Hp phenotypes may play a role in determining the severity and extent of myocardial damage in the setting of myocardial infarction; Hp 2-2 is considered an independent predictor of myocardial infarction [38,39]. The Hp 1-1 phenotype was reported to be protective in the setting of two critical vascular complications of diabetes mellitus: diabetic nephropathy and restenosis after percutaneous transluminal coronary angioplasty [40]. In addition, Hp 2-2 was reported to be overrepresented in autoimmune diseases, such as rheumatoid arthritis and systemic lupus erythematosus [41,42]. In our small case study, 43.8% of patients with cGVHD had an Hp 2-2 phenotype, higher than the normal donor group (Hp 2-2 18.7%). In addition, no Hp 2-2 type was found in the 9 patients with no clinical cGVHD presentation after allo-HCT (Table 4). Based on our preliminary results and the characteristics of the lower anti-oxidation activity and higher potential of APC cell activation by Hp 2-2, we may suspect that the patients with Hp 2-2 phenotype might have more genetic susceptibility or tendency for cGVHD development than the patients with Hp 1-1 or 2-1 type. To further confirm this hypothesis, a prospective analysis of correlations of Hp phenotype and the subsequent development of cGVHD is now being conducted.
In conclusion, we found that an increase in Hp expression is associated with cGVHD development. Further, the Hp 2-2 phenotype is present in patients who develop cGVHD more commonly than in those who do not develop this immunologic complication after allo-HCT. These findings might establish Hp as a valuable protein candidate for early cGVHD prediction and diagnosis. Several recently reported studies have utilized proteomic profiling in the development of predictive models of aGVHD [20][21][22].
Optimally, further studies utilizing proteomic profiling in patients with cGVHD will eventually lead to predictive models as well. Finally, further studies involving many more patients regarding the possible effects of Hp on T cell function during cGVHD are highly desirable.