Chapter 1 Introduction
1.7 Tyrosylprotein sulfotransferase in nervous system
So far, cholecystokinin (CCK) is the only TPST substrate found in the nervous system, and it functions as hormonal regulators of various digestive processes and feeding behaviors.
However, no neuronal function has been elucidated for CCK (Nichols et al., 1988), except it
7
has been suggested that CCK administration causes nausea and anxiety, and induces a satiating effect (Greenough et al., 1998). More importantly, the question remains in what the function of TPST is in the nervous system due to its highly expressed mRNA. In the present study, neuron-specific TPST knockout of transgenic D. melanogaster is used as an animal model to systematically identify the possible TPST-regulated protein based on proteomic system, and to further investigate the physiological roles of these proteins. The results may help to obtain a greater understanding about the biological roles of TPST in tyrosine sulfation post-translational modification.
8
Chapter 2 Experimental Procedures
2.1 Materials 2.1.1 Fly Strains
UAS-CG32632RNAi lines (transformant ID: 22915, 23777, 41596, 41597, and 41925) were purchased from Vienna Drosophila RNAi Center (VDRC), and a neuron-specific Gal4 driver line, Appl-Gal4 line, was kindly provided by Prof. Horng-Dar Wang’s lab (National Tsing Hua University, Taiwan). Appl-Gal4 was used to cross with UAS-CG32632RNAi to express double-stranded RNA interference (Figure 2), or each of them crossed with wild type fly w1118 as controls. All flies were incubated in 12 hr day/night cycle incubator at 25℃and 60-70% humidity, raised and maintained with standard fly food.
2.1.2 Oligonucleotide primers
Oligonucleotide primers were synthesized by Mission Biotech Co., Ltd. (Taiwan). The primers for CG32632 are 5’-AATGGCAGCTGCTTTATCGT-3’ (forward) and 5’-CATGCTGTCCGTGCTCG-3’ (reverse); the primers for Superoxide dismutase [Cu-Zn]
(SOD1) are 5’-TTGACTTGCTCAGCTCGTGT-3’ (forward) and 5’-CACGGTTTTCTTCGAACAGG-3’ (reverse).
2.1.3 Proteomics
The 13-cm Immobiline DryStrip (pH 4-7), IPG strip buffer (pH4-7), 87% glycerol, and DryStrip cover fluid were purchased from GE Healthcare Bio-Science (NJ, USA). The protein maker kit was purchased from Fermentas (Harrinton, Canada). The RC DC protein assay kit and 40% acrylamide/bis solution 29:1 were obtained from BioRad (Richmond, CA, USA). Sequencing-grade modified trypsin was purchased from Promega (Mannheim, Germany). Other chemicals: agarose was purchased from Amresco; acetic acid was
9
purchased from Fluka; 37% formaldehyde, glycine, trifluoracetic acid (TFA), and urea were purchased from J. T. Baker; methanol was purchased from Mallinckrodt; isopropanol was purchased from Merck; potassium hexaxyano ferrate (III) and Zinc sulfate were purchased from Panreac; sodium carbonate was purchased from Riedel-de Haën; ammonium bicarbonate, bromophenol blue, ethylenediaminetetraacetic acid (EDTA), imidazole, iodoacetamide (IAA), silver nitrate, sodium acetate, sodium thiosulphate, tetramethylenediamine (TEMED), triton X-100, and trypsin were purchased from Sigma-Aldrich; ammonium persulfate (APS), 3-[(3-cholamidopropyl) dimethylammonio]
-1-propanesulfonate (CHAPS), dithiotreitol (DTT), sodium dodecyl sulfate (SDS), and tris were purchased from USB.
2.2 Methods 2.2.1 RT-PCR
1- to 3-days old male flies were homogenized by pestle, total RNA was extracted with REzole, and then 5 μg of the total RNA was reverse transcribed with random hexamer primer.
For each cDNA preparation, a control synthesis reaction was treated with DNase on 37°C for 25 mins to ensure that there was no contaminating genomic DNA. The resulting cDNA library was subjected to PCR with primers. The products of PCR were analyzed by electrophoresis on 1% agarose gels, and then visualized using ethidium bromide (EtBr).
rp49 gene was used as an internal standard among different lines.
2.2.2 Sample Preparation
Whole flies were washed 3 times with PBS, and then the total protein of the whole fly were extracted using ultrasonication with lysis buffer containing 8 M urea, 2% CHAPS, 1%
v/v Triton X-100, 0.1% w/v SDS, and 5 mM DTT. The extract was boiled at 95℃ for 5 minutes following centrifugation for 10 mins at 15000¯g for several times until the
10
supernatant was clear.
2.2.3 Protein Quantitation
Using BSA as a standard, protein quantitation of the homogeneous from D. melanogaster was estimated by a colorimetric assay (RC DC protein assay, Bio-Rad). Reagent A’ (2.5μl DC Reagent S and 125μl DC Reagent A) was prepared for each sample. Mix 25μl of sample (25X dilution with extraction buffer) with 125μl RC Reagent I. After gently vortex, 125μl of RC Reagent II was added, and centrifuge at 15000Xg for 10minutes. The protein pellet was mixed with 127μl Reagent A’, and then after the incubation for 5 minutes until the pellet was dissolved, added 1ml of DC Reagent B into the solution. Incubate the solution at room temperature for 15 minutes, and finally absorbance can be read at 750nm with UV/VIS spectrophotometer.
2.2.4 Two Dimensional Electrophoresis
IEF: 200μg or 500μg of the total proteins were loaded depending on the detection methods (silver/SYPRO Ruby stain or negative stain). The proteins were mixed with the rehydration buffer (8 M urea, 2% w/v CHAPS, 0.5% v/v IPG buffer pH 4-7, 0.002%
bromophenol blue, and 18.2 mM DTT), and then eletrofocused in DryStrip using the following protocol with 50μA per strip at 20℃: (1) rehydration for 12 hours; (2) 500V for 500VHr (step and hold); (3) 1000V for 1000VHr (step and hold); (4) 3000V for 3000VHr (gradient); (5) 5000V for 5000VHr (gradient); (6) 8000V for 8000VHr (gradient); (7) 8000V for 40000VHr (step and hold). For silver stain and SYPRO Ruby stain, the strips were shaken for 15 minutes with first equilibration buffer (6 M urea, 29.3% v/v glycerol, 2% w/v SDS, and 75 mM Tris-HCl at pH 8.8) which contain 1% DTT, and then were shaken for 15 minutes with second equilibration buffer which contain 2.5% IAA. For reverse stain, the strips were shaken for 2X15 minutes with equilibration buffer.
11
SDS-PAGE: the strips were transferred onto 12.5% second-dimensional SDS-PAGE (1.5mm for silver stain or 1.0mm for negative stain) with the following protocol: (1) 15mA/strip for 30minutes at 4℃; (2) 30mA/strip for 5.5 hours at 4℃.
2.2.5 Protein Staining and analysis of 2-DE gels
Silver stain: fixation (40% methanol, 10% acetic acid) for overnight; sensitization (30%
methanol, 0.2% w/v sodium thiosulphate, 0.5M sodium acetate) for 30 minutes; washing (ddH2O) for 3X5 minutes; silver reaction (0.25% w/v silver nitrate) for 20 minutes; washing (ddH2O) for 2X1 minute; developing (2.5% w/v sodium carbonate, 0.02% formaldehyde) for 3-5 minutes; stopping (0.05M EDTA) for 10 minutes; washing (ddH2O) for 3X5 minutes;
preserving (8.7% glycerol, 30% methanol) for 1 hour.
Reverse stain: fixation (40% methanol, 10% acetic acid) for overnight; neutralization (tris, glycine, SDS) for 2X30 minutes; solution I (200mM imidazole, 0.1% SDS) for 1 hour;
washing (ddH2O) for 2X 1 minute; solution II (300mM zinc sulfate) for 1-2 minutes; stopping (ddH2O).
SYPRO Ruby staining: fixation (40% methanol, 10% acetic acid) for overnight; the gel was staining with SYPRO Ruby protein gel stain (Bio-Rad) for 16-18 hours; Rinse the gel in 10% methanol and 7% acetic acid for 1 hour, which to decrease the background fluorescence;
finally was the gel before imaging. In SYPRO Ruby staining gel, it is readily visualized using a UV or blue light source box.
Digital images of the gels were scanned by ImageScanner, and analyzed using ImageMaster 2D Platinum software V5.0 (GE Healthcare Bio-Science). The spots were detected and the background was subtracted (mode: average on boundary), and the gels were aligned and matched. A quantitative determination of the spots volumes was performed (mode: total spot volume normalization). Specific spots, either upregulated or downregulated, were excised for further identifying by MS analysis.
12
2.2.6 In-gel Digestion of Protein Spots in 2-DE gels
Specific spots, either upregulated or downregulated, were excised from the gels. The gel particles were washed with 50mM NH4HCO3/acetonitrile (1:1) for 15 minutes, and then silver destain solution (15mM K3Fe(CN)6, 50mM Na2S2O3) was used to destain the gel particles within 2X10 minutes. After removing the solution, we washed the gel particle in 20mM ammonium bicarbonate until it became colorless. The remaining liquid was removed, and then the gel particles were shrunken by adding just enough acetonitrile to cover the gel.
Finally, the gel particles were dried down for 15-20 minutes at room temperature. (In reverse stain, the gel particles were swelled in 10mM DTT/25mM NH4HCO3 for 1 hour at 56℃, and then we replaced the solution by 55mM IAA/25mM NH4HCO3 for 30minutes at room temperature. The gel particles were washed with 50mM NH4HCO3/acetonitrile (1:1) for 15 minutes, then shrunk and dried down following the steps described before.) The proteins in the gel particles were digested in the enzyme solution (25mM NH4HCO3 with 5ng/μl of trypsin) at 4℃ for 1 hour, and then incubated at 37℃ overnight by adding 3μl of 25mM NH4HCO3. The digests were sonicated in a water bath for 10 minutes, and then 50%
acetonitrile with 1% TFA was added. Finally, the supernatants were collected for analysis of MALDI-TOF-MS.
2.2.7 Protein Identification by MALI-TOF Mass Spectrometry
The MS raw data were processed by searching the protein databases (Swiss-Prot, MSDB, NCBInr) using MASCOT (http://www.matrixscience.com). To denote a protein as unambiguously identified, the Mowse scoring algorithms were used. Only proteins whose score exceeded the significance threshold (P<0.05) were concerned.
2.2.8 Oxidation Stress Assay of Flies from APPL-GAL4>UAS-TPST
RNAi, APPL-GAL4/+, and UAS-TPST
RNAi/+
13
The young male flies about 1- to 3-days old were collected for oxidative stress test.
The flies were fed with 10mM paraquat in 5% sucrose water once a day (16:30). The dead fly number was counted every 4 hours (00:30, 8:30, 12:30, 16:30, 20:30) till all flies were dead. Each tube contained about 20 flies, and at least 100 flies were included for each line.
The statistical significance of the observed change in the stress test was evaluated by using Student’s t test.
14
Chapter 3 Results
3.1 TPST mRNA distribution of wild type D. melanogaster
The expression of TPST mRNA of TPST varied in different tissues from the analysis of Affymatrix Dros2 expression arrays whose results deposited in Flyatlas (http://flyatlas.org/).
Flyatlas showed that TPST gene was highly abundant in several tissues, such as spermatheca, male accessory glands, salivary gland, and in the nervous system, including head, brain, and thoracicoabdominal ganglion as shown in Table 1. The mRNA signal of over 100 indicated as being abundant and over 1000 as remarkable. The present call showed how many of the four arrays for each sample actually gave a detectable expression. The enrichment displayed how much higher the signal is in a particular tissue than in the whole fly, which indicated whether the gene is tissue-specific. The biological rules of the higher mRNA expression on these tissues will need to be further clarified.
3.2 TPST mRNA expression in neuron-specific knockdown D. melanogaster
To determine whether TPSTRNAi can sufficiently knockdown TPST in the nervous system of flies, five different RNAi lines were crossed with Appl-Gal4 line. By using rp49 as an internal control, total mRNA of APPL-GAL4>UAS-TPSTRNAi, UAS-TPSTRNAi alone, and APPL-GAL4 alone were extracted from male flies and proceeded with RT-PCR, respectively (Figure 3). APPL-GAL4>UAS-TPSTRNAi representing the TPST gene was knocked down in the nervous system of flies; UAS-TPSTRNAi/+ and APPL-GAL4/+ represents the control groups. The relative quantities of TPST mRNA were analyzed and showed that only one (41596) of the five transgenic flies had a statistically significant decrease in TPST mRNA content compared to Appl-Gal4 and UAS-TPSTRNAi lines alone (Figure 4). This decrease of TPST mRNA content might correspond to the relative quantity of TPST in the pan-neuron showed in Table 1. Therefore, this RNAi transgenic line (41596) was chosen for further
15
experiments.
3.3 Proteome analysis of neuron-specific TPST knockdown D. melanogaster
The two-dimensional electrophoreses was utilized to spread the protein spots of D.
melanogaster proteome, and 13-cm drystrip with pH 4-7 as well as 12.5% SDS-PAGE were
optimized for analytical condition. The gels were repeated at least three times in each line of flies, and combined all together for further spots analysis. Approximately 1200 protein spots were detected under silver stain for each gel. However, only the protein spots that up and down-expressed in APPL-GAL4>UAS-TPSTRNAi fly compared to both of the control groups were considered as shown in Table 2. The protein expression of 22 proteins spots were considered to be significantly changed, including 4 proteins were up-regulated (spot 1, 5, 7, 17) and 18 proteins were down-regulated (spot 2-4, 6, 8-16, 18-22). Arrows represented that the protein spots were up-expression (Figure 5a, b, c).
3.4 Identification of proteins changes in neuron-specific TPST knockdown of D.
melanogaster
According to the analysis of two-dimensional SDS PAGE, 22 protein spots were excised from the gel and in-gel digested by trypsin after SYPRO Ruby staining, only 9 protein spots were identified by MALDI-TOF mass spectrometer as shown in Table 3. The up-regulated proteins were retinoblastoma-associated protein B (GE16567, Appendix 7a) which may involve in retinal cancer; phosphoenolpyruvate carboxykinase (GA10647, Appendix 7g) which may involve in gluconeogenesis; and CG9062 which may involve in pre-mRNA processing and cytoskeleton assembly (Appendix 7e). The down-regulated identified proteins were cytochrome P450 4e3 (CG4105, Appendix 7b) and superoxide dismutase [Cu-Zn] (CG11793, Appendix 7i) that were responsible for oxidative stress;
fructose-bisphosphate aldolase (CG6058, Appendix 7h) that is involved in glycolysis; alanine
16
dehydrogenase (GA20134, Appendix 7f) that is involved in metabolic process; CG31779, which has serine-type endopeptidase inhibitor activity (Appendix 7c); and GM05777p with an unknown function (Appendix 7d). From these identified proteins, the TPST somehow regulated the stress tolerance by down-regulated cytochrome P450 4e3 and superoxide dismutase [Cu-Zn].
3.5 Oxidative stress assay of neuron-specific TPST knockdown in D. melanogaster
In order to identify the possible physiological pathway of TPST in the nervous system of
D. melanogaster, stress tolerance was chosen since proteins had been identified to involve in
oxidative stress. The oxidative stress induced by 10mM paraquat revealed that the neuron-specific TPST knockdowned flies whose mean survival time (103.6 ± 13.3 hrs) were significant long compared to the UAS-TPSTRNAi alone (37.8 ± 7.3 hrs) and APPL-GAL4 alone (78.6 ± 8.3 hrs) in Figure 6. The neuron-specific TPST knockdowned flies displayed an increasing in survival time of 2.74-fold compared to UAS-TPSTRNAi flies and of 1.32-fold compared to APPL-GAL4 flies.
3.6 mRNA expression of superoxide dismutase [Cu-Zn] in neuron-specific knockdown D.
melanogaster
Although the protein spots of superoxide dismutase [Cu-Zu] (SOD1) disappeared on the two-dimensional electrophoresis gel, however Figure 7 shown that the mRNA expression of SOD1 in neuron-specific TPST knockdowned flies remained to have no difference compared to the two controls. Therefore, the down-regulation of SOD1 did not result from the mRNA expression in gene level. Moreover, Flyatlas performed that SOD1 was highly abundant in a majority of D. melanogaster tissues as shown in Table 4, which also indicates that the SOD1 was expressed ubiquitously in D. melanogaster.
17
Chapter 4 Discussions
Tyrosine sulfation was discovered in 1950s by the first sulfated protein, bovine fibrinogen (Bettelheim, 1954), and then tyrosylprotein sulfotransferase was denoted as the enzyme that catalyzed this reaction in 1983 (Moore, 2003). Protein sulfation has been largely researched over the past 50 years, however, a number of bottlenecks serve as challenges in previous studies, such as the difficulty of sourcing the homogeneous enzyme, limited information of enzyme characteristics (kinetics), unstable sulfated groups on the substrate, and lack of sensitive detecting methods for the sulfate group. Tyrosine-sulfated proteins have been indentified in physiological processes, including coagulation, leukocyte adhesion, chemokine signaling, and HIV entry (Seibert and Sakmar, 2008). At present, the understandings of TPST function are confined on the specific substrate involved in biological functions as described above. In order to systematically analyze the roles of protein sulfation in physiological regulation, we designed a proteome-wide screening tool that basically relied on proteomic techniques. Drosophila melanogaster was chosen as the source of study animal, due to it can grow easily, short generation span, well-established genomic database, commercial transgenic lines, and more importantly, D. melanogaster only has a single TPST gene (Moore, 2003). The amino acid sequence of TPST in D.
melanogaster shares 58% and 56% with human TPST1, and TPST2, respectively (Figure 8).
Approximately 75% of known human disease genes have a recognizable match in the genetic code of D. melanogaster, and 50% of D. melanogaster protein sequences have mammalian analogues (Reiter et al., 2001) which makes D. melanogaster an appropriate animal model for pathological studies on TPST.
To study the biological function of protein sulfation, knockdown endogenous TPST by RNAi technique in D. melanogaster is the most direct way. At present, TPST knockout mice was developed by Moore group, and the dysfunction of TPST has been indicated to
18
cause a severe defection of sperm motility which reveals that protein sulfation may be involved in the reproductive system (Hoffhines et al., 2008). Moreover, there is no information available about the relationship between protein sulfation and neurons, we are interested in focusing on neuron-specific knockdown due to the highly mRNA expression of TPST in the D. melanogaster nervous system (Table 1). The neuron-specific driver, APPL-GAL4, was used to cross with five different UAS-TPSTRNAi transgenic flies as our animal sources; only one had accomplished the neuron-specific knockdown in statistical significance compared to the controls, APPL-GAL4 alone and UAS-TPSTRNAi alone (Figure
3, 4). The mRNA was extracted from whole flies, therefore, the result of RT-PCR does not
directly indicate the knockdown in the nervous system, and instead, represents the knockdown of TPST in the whole fly. Theoretically, the two controls, APPL-GAL4 alone and UAS-TPSTRNAi alone, should have similar quantity of TPST mRNA expression. Figure 3 shows the two controls have different relative quantity of TPST, which may contribute to the distinct genetic background by different maintenance and growing conditions between the two controls.Total proteins were extracted from 1-7 day-old male flies with ultrasonication, and 200μg protein were loading onto each drystrip. pH4-7 Drystrips were chosen because the protein spots mostly tended to be distribute in the acidic region. Approximately 1200 proteins spots were visualized for each SDS-PAGE after silver staining, and different patterns of protein spots varied by different extracting methods. It has been estimated that only about 8% of the protein encoded by the genome could be analyzed in a two-dimensional SDS-PAGE of a total protein extract of D. melanogaster (Ericsson, 1999). Twenty-two protein spots showed a significant difference when compared with the controls, with 4 proteins up-regulated and 18 down-regulated. Among those protein spots, a number of spots seemed to have a pI shifting on the gel. Only 9 proteins were identified from MALDI-TOF mass spectrometry after preceding an in-gel trypsin digestion with trypsin. The protein loss
19
during the process of in-gel digestion could be a reason to decrease the protein identification of mass spectrometry. More importantly, however, the visualizing method, silver staining in this study, was the major cause for the low efficiency of protein identification by mass.
Although silver staining methods have raised the detection limit to the nanogram range, the protein identification of excised spots was often an obstacle that cannot be overcome easily (Poland et al., 2005). In order to solve the problem, the silver staining gels are usually treated as an analytical gel; other staining methods are then used as preparative gel for mass identification, such as coomassie staining, reverse staining, and SYPRO Ruby staining (Sasse and Gallagher, 2004). We chose SYPRO Ruby and reverse staining for the preparative gel, and the mass analysis results were shown in Table 3 and Appendix 7a-i.
The neuron-specific knockdown flies down-regulated two oxidative stress proteins:
cytochrome P450 4e3 (Cyp4e3) and superoxide dismutase [Cu-Zn] (SOD1). Superoxide dismutases is an ubiquitous enzymes that functions to efficiently catalyze the dimutation of superoxide anions (Zelko et al., 2002), which is known to protect organisms from reactive oxygen metabolites (Goulielmos et al., 2003). SOD1 is widely distributed and comprised 90% of the total SOD (Noor et al., 2002), the mRNA expression of D. melanogaster is shown in Table 4. It is obvious to notice, however, that the protein spot of SOD1 (spot 20) in the gel was completely disappeared in TPST knockdown flies (Figure 5). Interestingly, TPST is only knocked down in nervous system, but it caused SOD1, expressed ubiquitously, to completely vanish in silver staining vision. Besides the actual down-regulation of the protein, the disappearance of protein spot could be contributed to the change of either isoelectric focusing point or the molecular weight of the protein, which caused a spot shift.
Further confirmation is needed for the clarification.
Based on the down-regulated Cyp4e3 and SOD1, implications can be made that the neuron-specific knockdown flies tended to suffer stress more easily, especially oxidative stress. In the oxidative stress assay, we used paraquat to increase the quantity of free radical
20
in D. melanogaster. An unexpected result revealed that the mean survival time of the TPST knockdown flies was much longer than UAS-TPSTRNAi alone for 66 hours and APPL-GAL4 alone for 25 hours, (Figure 6). The longevity of the APPL-GAL4 alone might need to be clarified in advance. The incredible longevity of neuron-specific TPST knockdown flies seems to be an opposite result as we expected. Oxidative stress has been reported to be a common underlying mechanism in the pathogenesis of many neurodegenerative disorders such as Alzheimer, Huntington, and Parkinson disease (Gruenewald C et al., 2009).
Previous researches on SOD1 and neuron have indicated that overexpression of SOD1 in D.
melanogaster can reduce oxidative damage (Landis and Tower, 2005), extend lifespan (Parkes et al., 1998) (Sampayo et al., 2003), and neuron protection (Botella et al., 2008). The
overall evidence reveals that down-regulated SOD1 should decrease the survival rate of TPST knockdown flies in oxidative stress assay. Nevertheless, our result is in conflict with previous findings. The disappearance of the SOD1 (spot 20) on the two-dimensional SDS-PAGE did not result from the mRNA depletion on the gene level, which was proved by RT-PCR (Figure 7).
Moreover, there is only a single TPST gene in D. melanogaster by the analysis from BLAST. The TPST gene, however, might express two isoforms, Tango-PB and Tango-PC, with different length of amino acids. The difference between these two isoforms is that Tango-PB possessed extended C-terminal 150 amino acid residues with polyglutamine (polyQ) and polyasparagine (polyN) (Figure 9). A number of neurodegenerative diseases are characterized by the formation of intracellular protein aggregates and neurodegeneration.
The polyQ sequence can easily cause protein misfolding and the formation of inclusion body
The polyQ sequence can easily cause protein misfolding and the formation of inclusion body