Systematic Evaluation of Sample Preparation Methods for Gel-Based

Nov 25, 2005 - Saliva analysis by surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF/MS): from sample collection...
73 downloads 15 Views 798KB Size
Systematic Evaluation of Sample Preparation Methods for Gel-Based Human Urinary Proteomics: Quantity, Quality, and Variability Visith Thongboonkerd,* Somchai Chutipongtanate, and Rattiyaporn Kanlaya Siriraj Proteomics Facility, Medical Molecular Biology Unit, Office for Research and Development, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand Received August 2, 2005

Abstract: We performed systematic evaluation of 38 protocols to concentrate normal human urinary proteins prior to 2D-PAGE analysis. Recovery yield and pattern of resolved protein spots were compared among different methods and intra-/inter-individual variabilities were examined. Precipitation with 90% ethanol provided the greatest protein recovery yield (92.99%), whereas precipitation with 10% acetic acid had the least protein recovery (1.91%). In most of precipitation protocols, the higher percentage of applied organic compounds provided the greater recovery yield. With a fixed concentration at 75%, the urine precipitated with acetonitrile had the greatest number of protein spots visualized in 2D gel, whereas the acetic-precipitated sample had the smallest number of spots. For the intra-individual variability, the first morning urine had the greatest amount of total protein but provided the smallest number of protein spots visualized. Excessive water drinking, not caffeine ingestion, caused alterations in the urinary proteome profile with newly presenting spots and also proteins with decreased excretion levels. As expected, there was a considerable degree of inter-individual variability. Coefficients of variation for albumin and transferrin expression were greatest by inter-individual variables. Male urine had greater amount of total protein but provided smaller number of protein spots compared to female urine. These data offer a wealth of useful information for designing a high-quality, large-scale human urine proteome project. Keywords: urine • urinary • proteomics • proteome • yield • sample preparation • precipitation • lyophilization • ultracentrifugation • filtration

Introduction Human urine is one of the most interesting and useful biofluids for clinical proteomics study, as it is available in almost all of patients and because of noninvasiveness and simplicity of specimen collection. Several recent and ongoing * To whom correspondence should be addressed. Siriraj Proteomics Facility, Medical Molecular Biology Unit, Office for Research and Development, 12th Floor Adulyadej Vikrom Building, Siriraj Hospital, 2 Prannok Road, Bangkoknoi, Bangkok 10700, Thailand. Phone/Fax: +66-2-4184793. E-mail: [email protected] (or) [email protected]. 10.1021/pr0502525 CCC: $33.50

 2006 American Chemical Society

studies have applied urinary proteomics to biomarker discovery for kidney diseases as well as other disorders that may have systemic alterations in metabolic and biochemical profiles that can affect urinary protein excretion. Identification of urinary biomarkers may lead to the development of a simple diagnostic test to be used in clinical practice for earlier disease detection and/or better therapeutic outcome.1,2 However, recent urinary proteomics studies have not reached this ideal goal because of the difficulty to examine the entire urinary proteome.3 Additionally, normal human urine has very diluted protein concentration with high-salt contents that interfere with proteomic analysis, especially during the isoelectric focusing (IEF) in gel-based study. Moreover, yield of protein recovery remains unsatisfactory by protein loss during sample preparation. It is, therefore, crucial to use an effective protocol to isolate/ concentrate urinary proteins and to eliminate those interfering salts. One of the top priorities in the urinary proteomics field during the coming years is to optimize sample preparation methods to examine the urinary proteome.1-3 Indeed, urinary proteome can be examined using either gel-based or gel-free techniques. While sample preparation to concentrate the urine may be unnecessary for surface-enhanced laser desorption/ ionization (SELDI)4,5 and capillary electrophoresis coupled to mass spectrometry (CE-MS),6,7 it is a crucial step in gel-based proteomics study using two-dimensional polyacrylamide gel electrophoresis (2D-PAGE).8 In the present study, we have focused on sample preparation methods used for gel-based human urinary proteomics. There are several protocols that can be employed to isolate/ concentrate urinary proteins; e.g., precipitation,9-13 lyophilization,12,14-17 ultracentrifugation 10 and centrifugal filtration.8,11,12,18 However, the information on relative efficacy of these different methods in term of their quality and recovery yield is still lacking and such comparison is urgently required. The present study was conducted to systematically examine protein recovery yield (quantity) and 2D spot pattern (quality) of urinary proteins isolated with various protocols, including precipitation with acetic acid, acetone, acetonitrile, ammonium sulfate, chloroform, ethanol and methanol (with varying final concentrations of 10, 25, 50, 75, and 90%); centrifugal filtration (with molecular mass cutoff at 10 kDa); lyophilization; and ultracentrifugation (at 200 000 × g for 2 h). Most of these methods have been already applied to the human urine in other previous studies,8-18 while some protocols have been first developed and/or proven to be effective for the human urine Journal of Proteome Research 2006, 5, 183-191

183

Published on Web 11/25/2005

technical notes

Sample Preparation for Urinary Proteomics

in our present study (i.e., precipitation with ethanol and acetic acid). Another issue that needs to be concerned in urinary proteomics study is timing of sample collection. We, thus, evaluated the intra-individual variability by comparing the amounts of total protein and 2D spot patterns of urine samples collected from the same subject at different time-points during the day. Additionally, effects of caffeine and excessive water drinking on urinary protein quantity and 2D spot pattern were examined. Finally, inter-individual variability was evaluated.

Materials and Methods Urine Collection. Mid-stream urine specimens were collected from a total of 8 healthy individuals: 4 males (age 28.8 ( 3.9 years) and 4 females (age 25.8 ( 1.0 years), who had no recent medication. All females had no menstrual cycle at the time of collection. The urine was collected with 1 mM sodium azide and pool urine was utilized to examine the difference among various sample preparation protocols to avoid intraand inter-individual variabilities. To evaluate the intraindividual variability, the urine was sequentially collected from a 34-year-old healthy male subject at variable time-points during the day: (i) first morning void; (ii) afternoon void; (iii) after drinking a cup of coffee (urinary bladder was empty before caffeine intake and the urine was collected when the bladder was full); and (iv) after 1 L water loading within 20 min (the bladder was empty before water loading and the urine was collected when the bladder was full). To evaluate the interindividual variability, urine samples from individual subjects were compared within group of the same gender, and between groups of the two genders. After the urine was collected, cell debris and particulate matter were removed with 1000 × g centrifugation (at 4 °C for 5 min) and the urinary supernatant was subjected to further isolation/concentration procedures. Equal volume (10 mL) was employed for each sample preparation protocol. Precipitation. Precipitation of urinary proteins with different organic compounds was performed using similar protocol. Acetic acid, acetone, acetonitrile, ammonium sulfate, chloroform, ethanol and methanol were added to the urine with various final concentrations of 10%, 25%, 50%, 75%, and 90%. The solution was mixed and incubated at 4 °C for 10 min and the precipitant was isolated using 12 000 × g (at 4 °C for 5 min). The supernatant was discarded and the pellet was allowed to air-dry and then resuspended with a solubilizing buffer containing 7 M urea, 2 M thiourea, 4% CHAPS, 2% (v/v) ampholytes (pH 3-10), 120 mM DTT, and 40 mM Tris-base. The solution was dialyzed against 18 MΩ‚cm (dI) water (at 4 °C overnight), lyophilized and then resuspended with the mentioned solubilizing buffer. Protein concentration was measured using Bradford’s method. Centrifugal Filtration. The urine was spun at 12 000 g using 10-kDa cutoff centrifugal column (Vivaspin; Vivascience AG, Hannover, Germany) at 4 °C until approximately 1/30 of initial volume remained. The concentrated urine was collected, dialyzed against dI water (at 4 °C overnight), lyophilized and then resuspended with the solubilizing buffer as described above. Protein concentration was measured using Bradford’s method. Lyophilization. The urine was lyophilized until dry and then resuspended in the solubilizing buffer. The resuspension was dialyzed against dI water (at 4 °C overnight), lyophilized again, 184

Journal of Proteome Research • Vol. 5, No. 1, 2006

and then resuspended with the solubilizing buffer. Protein concentration was measured using Bradford’s method. Ultracentrifugation. Urinary proteins were isolated using ultracentrifugation (200 000 × g) at 4 °C for 2 h (Optima LE80K Ultracentrifuge; Beckman Coulter, Inc., Fullerton, CA). The supernatant was discarded and the remaining thin film of proteins was resuspended with the solubilizing buffer. The resuspension was dialyzed against dI water (at 4 °C overnight), lyophilized, and then resuspended with the solubilizing buffer. Protein concentration was measured using Bradford’s method. 2D-PAGE and Staining. Immobiline DryStrip, linear pH 3-10, 7 cm long (Amersham Biosciences, Uppsala, Sweden) was rehydrated overnight with 200 µg total protein (equal loading for all samples) that was premixed with rehydration buffer containing 7 M urea, 2 M thiourea, 2% CHAPS, 2% (v/v) ampholytes (pH 3-10), 120 mM DTT, 40 mM Tris-base and bromophenol blue (to make the final volume of 150 µL per strip). The first dimensional separation (IEF) was performed in Ettan IPGphor II IEF System (Amersham Biosciences) at 20 °C, using stepwise mode to reach 9000 V‚hours. After completion of the IEF, proteins on the strip were equilibrated with a buffer containing 6 M urea, 130 mM DTT, 30% glycerol, 112 mM Tris base, 4% SDS and 0.002% bromophenol blue for 10 min, and then with another buffer containing 6 M urea, 135 mM iodoacetamide, 30% glycerol, 112 mM Tris base, 4% SDS and 0.002% bromophenol blue for 10 min. The IPG strip was then transferred onto 12% acrylamide slab gel (8 × 9.5 cm) and the second dimensional separation was performed in SE260 Mini-Vertical Electrophoresis Unit (Amersham Biosciences) with the current 20 µA/gel for 1.5 h. Separated protein spots were then visualized using Coomassie Brilliant Blue R-250 stain. Spot Analysis and Matching. Image Master 2D Platinum (Amersham Biosciences) software was used for matching and analysis of protein spots on 2D gels. Parameters used for spot detection were (i) minimal area was 10 pixels; (ii) smooth factor was 2.0; and (iii) saliency was 2.0. A reference gel was created from an artificial gel combining all of the spots presenting in different gels into one image. The reference gel was then used for matching of corresponding protein spots between gels. Percentage of matching between two gels was calculated using the following formula % match ) 2 × no. of matched spots × 100/total number of spots in gel1 + gel2

Results and Discussion Dialyzed versus Non-Dialyzed Urine Sample. Frequently, the IEF of urinary proteins isolated with various protocols is problematic because of the high-salt contents as well as lipids and glycosaminoglycans in the urine. Precipitation is one of the protocols that are expected to remove these contaminants or interfering substances. However, the poor IEF is occasionally observed in precipitation protocols using organic solvents or other organic compounds, probable because the interfering substances remain in the protein pellet. Figure 1 shows the results of 2D separation of urinary proteins isolated from the pool urine using 50% acetone and 50% ethanol precipitation, with or without dialysis to remove salts and other interfering substances. The dialyzed samples (Figures 1B and 1D) had better spot resolution, whereas the IEF of the nondialyzed samples was problematic (Figure 1, parts A and C). Our results were consistent with those reported in a previous study, in which the nondialyzed urine fractionated with am-

technical notes

Thongboonkerd et al. Table 1. Protein Recovery Yield of Various Sample Preparation Protocols Utilized to Isolate Human Urinary Proteins

isolation methods

Figure 1. Dialyzed versus nondialyzed urine. Poor IEF was occasionally observed in precipitated samples, most likely due to the interfering substances (salts, lipids and glycosaminoglycans) that remained in the protein pellets. (A) and (C) show unsatisfactory 2D gels of nondialyzed samples precipitated with 50% acetone and 50% ethanol, respectively. (B) and (D) illustrate the improvement of protein spot resolution after dialysis of the precipitated samples with dI water prior to 2D-PAGE.

monium sulfate and precipitated with trichloroacetic acid (TCA) showed severe smearing on a 2-D gel, whereas the dialyzed sample separated well on the 2-D gel.12 The same study also demonstrated that the urine precipitated with ethanol following a single dialysis against dI water had the poor IEF, whereas the sample treated with two cycles of dialysis had the better IEF but could not be precipitated with 75% ethanol. In our present study, ethanol precipitation (10-90%) provided satisfactory results on both the IEF (Figure 1D) and protein recovery yield (see next section). These disparate results between the two studies probably because of the sequence of dialysis and precipitation; we precipitated the urine with ethanol prior to dialysis and lyophilization, while a vice versa order was done in the previous study.12 Although the unsatisfactory IEF of urinary proteins does not always occur, we recommend using routine dialysis to remove salts and interfering substances during sample preparation of urinary proteins for gel-based proteomic analysis to avoid unnecessary repetition of experimental procedures that may cause sample depletion, particularly in cases of rare diseases. Another technique that can be applied to efficiently eliminate contaminants is solid-phase extraction.19 However, the issue to be concerned is that “the more steps of sample preparation, the greater protein loss”. Protein Recovery Yield and Quality of Resolved Protein Spots Using Different Protocols. We evaluated protein recovery yield of 38 different protocols for isolating urinary proteins including precipitation with acetic acid, acetone, acetonitrile, ammonium sulfate, chloroform, ethanol and methanol (with various final concentrations of 10%, 25%, 50%, 75% and 90%); centrifugal filtration (with molecular mass cutoff at 10 kDa); lyophilization; and ultracentrifugation (200 000 × g for 2 h at 4 °C). The pool urine was used in this experiment to avoid the biases from intra- and inter-individual variabilities. Table 1 illustrates that the greatest protein recovery yield was obtained

a. acetic acid 10% 25% 50% 75% 90% b. acetone 10% 25% 50% 75% 90% c. acetonitrile 10% 25% 50% 75% 90% d. ammonium sulfate 10% 25% 50% 75% 90% e. chloroform 10% 25% 50% 75% 90% f. ethanol 10% 25% 50% 75% 90% g. methanol 10% 25% 50% 75% 90% h. centrifugal filtration (10 kDa cutoff) i. lyophilization j. ultracentrifugation

urine volume (mL)

total proteina (mean ( SD) (µg)

recovery yieldb (average) (%)

10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10

26.76 ( 0.77 40.69 ( 0.91 66.93 ( 2.98 138.40 ( 2.55 171.85 ( 3.89 83.11 ( 0.27 160.48 ( 2.58 351.87 ( 4.23 460.14 ( 1.78 570.72 ( 1.98 77.19 ( 7.41 104.96 ( 11.43 197.91 ( 2.54 212.02 ( 27.96 265.35 ( 35.41 76.93 ( 5.84 86.94 ( 1.02 109.29 ( 1.06 235.83 ( 9.59 161.36 ( 2.37 111.78 ( 1.03 182.75 ( 22.87 189.78 ( 18.06 174.21 ( 22.59 152.47 ( 15.31 78.34 ( 0.64 245.97 ( 4.49 409.57 ( 6.45 908.11 ( 54.74 1301.81 ( 75.80 80.28 ( 0.26 242.66 ( 0.78 430.57 ( 7.19 884.56 ( 26.04 1017.25 ( 141.04 254.00 ( 5.25

1.91 2.91 4.78 9.89 12.28 5.94 11.46 25.13 32.87 40.77 5.51 7.50 14.14 15.14 18.95 5.49 6.21 7.81 16.85 11.53 7.98 13.05 13.56 12.44 10.89 5.60 17.57 29.26 64.87 92.99 5.73 17.33 30.76 63.18 72.66 18.14

10 10

1017.10 ( 26.64 292.00 ( 25.88

72.64 20.86

a n ) 3 for each sample and the pool urine was used. b % recovery yield ) total protein isolated by each method (µg) × 100/total protein measured in untreated urine (µg).

from the sample precipitated with 90% ethanol (92.99%), whereas precipitation with 10% acetic acid had the least protein recovery yield (1.91%). Figure 2 illustrates differential protein spot patterns obtained from 10 different methods with an equal amount of total protein (200 µg) loaded in each 2D gel (for precipitation technique, only a final concentration at 75% of organic compounds was used for the comparison). The data in Table 1 and Figure 2 clearly demonstrated that not only the quantity but also the quality of recovered proteins were affected by differential sample preparation protocols. Image Master 2D Platinum software was employed to analyze protein spots across different gels. Spots in the 2D gel were carefully detected as ‘true protein spots’ using the highstringency criteria: (i) minimal area was 10 (pixel unit); (ii) smooth factor was 2.0; and (iii) saliency was 2.0. With these criteria of spot detection, the maximal number of spots visualized was 206 from 75% acetonitrile-precipitated sample, whereas lyophilized and 75% acetic-precipitated samples had the smallest number (72 spots) of visualized protein spots (Table 2). It should be noted that precipitation methods may create protein variants as a result of chemically induced Journal of Proteome Research • Vol. 5, No. 1, 2006 185

technical notes

Sample Preparation for Urinary Proteomics

Table 2. Total Number of Protein Spots Obtained from Different Isolation Protocols group ID

isolation methods

total no. of spotsa

A B C D E F G H I J

acetic acid (75%) acetone (75%) acetonitrile (75%) ammonium sulfate (75%) chloroform (75%) ethanol (75%) methanol (75%) centrifugal filtration (10 kDa cutoff) lyophilization ultracentrifugation

72 159 206 142 146 187 189 116 72 76

a Equal amount (200 µg) of total protein was loaded and the pool urine was used.

Figure 2. Differential 2D proteome profiles of the pool urine concentrated with various protocols. The pool urine was utilized to avoid intra- and inter-individual variabilities of samples, and equal amount (200 µg) of total protein was loaded in each gel. The 2D spot patterns of urinary proteins isolated with various protocols obviously differed. % Match between gels varied from 11.65% to 65.52% (Table 3).

modifications; thus, more spots could be observed as compared to nonprecipitation methods. This should be kept in mind when posttranslationally modified proteins are examined. The present study reported smaller total number of protein spots visualized in a 2D gel as compared to the number reported in previous studies, most likely due to the higher stringency criteria that we applied to detect ‘true protein spots’. 186

Journal of Proteome Research • Vol. 5, No. 1, 2006

In addition, we used Coomassie Brilliant Blue R-250 stain for all gels to be compared. Much more spots can be visualized by the more sensitive stains; e.g. silver, SYPRO Ruby, or Deep Purple (Figure S1). Note that different stains may provide some minor differences on 2D spot pattern as the mechanisms of staining differ; thus, some proteins favor a particular stain whereas other proteins may favor another one. Although precipitation with acetonitrile provided the largest number of protein spots that could be detected during 2D spot analysis, it had a low protein recovery yield. We, therefore, recommend precipitation with ethanol, methanol, or acetone as the method of choice for routine, gel-based, urinary proteome analysis, because these protocols provided the high protein recovery yield and the large number of protein spots visualized in a 2D gel, thus a wide variety of proteins can be examined in one experiment. Acetonitrile precipitation is recommended for the urine obtained from patients with proteinuria (much higher protein concentration than the normal urine) or when a large volume of urine is used for an analytical gel. Results from spot matching across different gels using various sample preparation protocols to isolate urinary proteins showed that. % match (% match ) 2 × no. of matched spots × 100/total no. of spots in gel1 + gel2) between gels varied from 11.65% to 65.52%; indicating a high degree of variations among different protocols (Table 3). The best matched pair was the acetone-precipitated sample vs methanolprecipitated urine. The second and third best matched pairs were ammonium-precipitated urine vs chloroform-precipitated sample and lyophilized urine vs ultracentrifuged sample, respectively. These findings may indicate the similarity in physicochemical properties of proteins isolated by methods within the best matched pair that would be expected to have similar mechanisms to isolate urinary proteins. Because the degree of variations among different protocols was high, it is expected that combining more than one protocol would be beneficial to visualize and/or examine the larger number of resolved proteins on 2D gels. We evaluated the number of protein spots that could be detected on 2-D gels using single or more than one protocol (Figure S2). The results demonstrated that combining two protocols could increase the number of resolved protein spots up to 254% (average 93%) of the number obtained from each single protocol. Combination of 7 precipitation techniques increased the total spot number to 604 spots and combination of 7 precipitation techniques with centrifugal filtration, lyophilization and ultracentrifugation provided a total of 685 spots (Figure S2). Therefore, more than one isolation technique is required to study the entire human

technical notes

Thongboonkerd et al.

Table 3. Matching of Protein Spots between Gels of Various Isolation Protocols gel no. 1

gel no. 2

A B A C A D A E A F A G A H A I A J B C B D B E B F B G B H B I B J C D C E C F C G C H C I C J D E D F D G D H D I D J E F E G E H E I E J F G F H F I F J G H G I G J H I H J I J (Average % match)

no. of matched spots

18 31 30 21 24 21 39 24 22 27 28 32 31 114 25 38 35 52 35 56 23 38 23 24 83 62 43 33 17 30 48 46 16 26 30 46 28 26 27 25 33 46 12 17 34

% matcha

15.58 22.30 28.04 19.27 18.53 16.09 41.49 33.33 29.73 14.79 18.60 20.98 17.92 65.52b 18.18 32.90 29.79 29.89 19.89 28.50 11.65 23.60 16.55 17.02 57.64b 37.69 25.98 25.58 15.89 27.52 28.83 27.46 12.21 23.85 27.03 24.47 18.48 20.08 20.53 16.39 25.29 34.72 12.77 17.71 45.95b (25.25)

a % Match ) 2 × Number of matched spots × 100/Total number of spots in gel1 + gel2. A - acetic acid (75%); B - acetone (75%); C - acetonitrile (75%); D - ammonium sulfate (75%); E - chloroform (75%); F - ethanol (75%); G methanol (75%); H - centrifugal filtration (10 kDa cutoff); I - lyophilization; J - ultracentrifugation. b Three best matched pairs. (See gel images in Figure 2).

urine proteome as a single method is not sufficient to recover all of the urinary proteins to be examined. Applying additional strategies, such as fractionation, zoom-in IEF, removal of major abundant proteins, immunoaffinity, affinity column, etc., would be expected to further increase the number of resolved proteins. Effect of Differential Concentrations of Organic Compounds Used in Precipitation Protocols on Protein Recovery Yield and Quality of Resolved Protein Spots. We also evaluated the effect of differential concentrations of organic compounds used in precipitation protocols on protein recovery and protein spot pattern. The pool urine was employed in this experiment to avoid intra- and inter-individual variabilities. The results in Table 1 show that the higher concentration provided the greater recovery yield for precipitation with acetic acid, acetone,

Figure 3. Effect of differential concentrations of ethanol on 2D spot patterns. Various concentrations of ethanol were utilized to precipitate proteins from the pool urine (equal amount of 200 µg total protein was loaded in each gel). The data clearly demonstrate that using different concentrations of organic compound to precipitate proteins had some effects on the 2D spot patterns and the maximal % match between gels was only 66.49% (Table 5).

acetonitrile, ethanol and methanol. However, this rule could not be applied to precipitation with ammonium sulfate and chloroform as the maximal recovery yield for precipitation with ammonium sulfate and chloroform was observed at the concentrations of 75% and 50%, respectively. To examine the effect of differential concentrations of organic compounds on protein spot pattern, ethanol precipitation protocols were evaluated. Figure 3 illustrates that protein spot pattern of 10%, 25%, 50%, 75%, and 90% ethanolprecipitated urinary proteins were somewhat different. Theoretically, a protein in the solution tended to precipitate when pH of the solution is at or closed to the isoelectric point (pI) of that protein. The pH values of the urine samples precipitated with 10-90% ethanol varied from 7.16 to 7.80. The difference in protein spot pattern observed when different concentrations of ethanol were used might be partly due to differential pH levels of the mixture solutions. Table 4 demonstrates that even though the lowest concentration of ethanol had the least protein recovery yield it provided the largest number of detected protein spots. Table 5 shows that the two best matched pairs were from 75% vs 90% and 25% vs 50%. These findings underscore the equal significance of quantity and quality of resolved proteins, both of which should be concerned in a proteomic analysis. Intra-Individual Variability. The kidney is a major vital organ that maintains the normal homeostasis by adjusting Journal of Proteome Research • Vol. 5, No. 1, 2006 187

technical notes

Sample Preparation for Urinary Proteomics Table 4. Total Number of Protein Spots Obtained from Differential Concentrations of Ethanol Used for Precipitating Urinary Proteins group ID

isolation methods

total no. of spotsa

A B C D E

10% ethanol 25% ethanol 50% ethanol 75% ethanol 90% ethanol

311 295 259 173 209

a Equal amount (200 µg) of total protein was loaded and the pool urine was used. (See gel images in Figure 3).

Table 5. Matching of Protein Spots between Gels Using Various Concentrations of Ethanol to Precipitate Urinary Proteins gel no. 1

gel no. 2

A B A C A D A E B C B D B E C D C E D E (average % match)

no. of matched spots

% matcha

159 171 109 122 172 110 122 129 127 127

52.48 60.00 45.04 46.92 62.09b 47.01 48.41 59.72 54.27 66.49b (54.24)

a % Match ) 2 × no. of matched spots × 100/total no. of spots in gel1 + gel2. A - 10% ethanol; B - 25% ethanol; C - 50% ethanol; D - 75% ethanol; E - 90% ethanol. b Two best matched pairs (See gel images in Figure 3).

Table 6. Quantitative Analysis of Total Protein Obtained from Different Sets of Samples to Demonstrate Intra- and Inter-individual Variabilities sample variables

urine volume (mL)

intra-individual variables first morning (n ) 3) 10 afternoon (n ) 3) 10 effect of caffeine (n ) 3) 10 effect of water loading (n ) 3) 10 inter-individual variables male (n ) 4) 10 female (n ) 4) 10 a

total proteina (mean ( SD) (µg)

530.37 ( 40.76 486.47 ( 10.69 319.88 ( 0.86 127.89 ( 11.65 512.96 ( 65.66 355.01 ( 105.23

75% Ethanol precipitation was used for all samples in these two sets.

excretion and reabsorption of urinary constituents and by other regulatory mechanisms. Urinary constituents would, therefore, be expected to alter dynamically in association with diurnal variations and physiological changes. In the present study, intra-individual variability was evaluated in a 34-year-old healthy male. Four urine samples were sequentially collected at 4 different time-points during the same day, including first morning void, afternoon void, after drinking a cup of coffee and after water loading (1L within 20 min). As expected, the results in Table 6 show that the sample that had the greatest amount of total protein (or the most concentrated one) was the first morning urine, whereas the urine after water loading had the smallest amount of total protein (or the most diluted one). However, the first morning urine that had the greatest protein amount provided the smallest number of resolved protein spots in a 2D gel (Figure 4 and Table 7). There was a mild degree of variability in protein spot pattern; the % match between gels varied from 54.00% to 73.09% (Figure 4 and Table 8). The average % match of this set of intra188

Journal of Proteome Research • Vol. 5, No. 1, 2006

Figure 4. Intra-individual variability. Four urine samples were sequentially collected from a 34-year-old healthy male at 4 different time-points during the same day, including first morning void, afternoon void, after drinking a cup of coffee and after water loading (1L within 20 min). Equal amount (200 µg) of total protein was loaded and proteins were isolated with 75% ethanol precipitation for all samples. As expected, there was a considerable degree of variability in protein spot patterns as shown in this figure and % match between gels varied from 54.00% to 73.09% (Table 8). Interestingly, there were a few spots present only after water loading (circle area). Additionally, excretion levels (as determined by spot intensity) of the other few spots were decreased to almost absent after water loading (rectangular area). Table 7. Total Number of Protein Spots Recovered from the Urine Obtained from the Same Subject at Various Time-Points During the Day Using an Identical Sample Preparation Method group ID

intra-individual variables

total no. of spotsa

A B C D

first morning urine afternoon urine effect of caffeine effect of water loading

135 144 166 156

a Equal amount (200 µg) of total protein was loaded and proteins were isolated with 75% ethanol precipitation for all samples. (See gel images in Figure 4).

Table 8. Matching of Protein Spots between Gels of the Urine Obtained from the Same Subject at Various Time-points during the day Using an Identical Preparation Method gel no. 1

gel no. 2

A B A C A D B C B D C D (average % match)

no. of matched spots

90 110 83 95 81 114

% matcha

64.52 73.09 57.04 61.29 54.00 70.81 (61.99)

a % match ) 2 × no. of matched spots × 100/total no. of spots in gel1 + gel2. (See gel images in Figure 4). A - first morning urine; B - afternoon urine; C - effect of caffeine; D - effect of water loading.

individual variables using the identical sample preparation method was greater than those reported in Tables 3 and 5 for the sets of pool urine isolated with different protocols and with differential concentrations of ethanol precipitation, respectively

technical notes

Thongboonkerd et al. Table 9. Total Number of Protein Spots Recovered from the Urine Obtained from Different Subjects Using an Identical Sample Preparation Method group ID

interindividual variables

total no. of spotsa

M1 M2 M3 M4 F1 F2 F3 F4

male #1 male #2 male #3 male #4 female #1 female #2 female #3 female #4

128 115 120 153 246 157 229 173

a Equal amount (200 µg) of total protein was loaded and proteins were isolated with 75% ethanol precipitation for all samples. (See gel images in Figure 5).

Table 10. Matching of Protein Spots between Gels of the Urine Obtained from Different Subjects Using an Identical Sample Preparation Method gel no. 1

Figure 5. Inter-individual variability. The results clearly illustrate the difference of 2D protein spot patterns among individuals of both males (A-D) and females (E-H). Equal amount (200 µg) of total protein was loaded and proteins were isolated with 75% ethanol precipitation for all samples. There was a considerable degree of inter-individual variability as demonstrated by varying % match between gels (ranging from 36.83% to 82.52%; Table 10).

(61.99% vs 25.25% and 54.24%, respectively). Interestingly, water loading had more effects on the urinary proteome profile than caffeine ingestion, as there were a few spots that were present only after water loading (in a circular area labeled in Figure 4). In contrast, the intensity of other few protein spots was decreased to almost absent after water loading (rectangular area in Figure 4), indicating the decrease of its excretion into the urine after water loading. These may reflect regulatory mechanisms of the kidney to maintain the normal homeostasis during the physiological changes. The intra-individual variability was also evaluated in a previous study,12 in which the significant difference of 2D spot pattern was observed when the urine samples were collected on different days from the same subject. Other factors that may have influence on the urinary proteome pattern of an individual are foods, drugs, daily activities, exercise, stress, menstrual cycle, other physi-

gel no. 2

M1 M2 M1 M3 M1 M4 M1 synthetic M M1 synthetic F M2 M3 M2 M4 M2 synthetic M M2 synthetic F M3 M4 M3 synthetic M M3 synthetic F M4 synthetic M M4 synthetic F synthetic M F1 synthetic M F2 synthetic M F3 synthetic M F4 synthetic M synthetic F F1 F2 F1 F3 F1 F4 F1 synthetic F F2 F3 F2 F4 F2 synthetic F F3 F4 F3 synthetic F F4 synthetic F (average % match)

no. of matched spots

72 92 99 84 62 68 83 72 58 95 85 62 85 63 72 72 58 57 57 123 132 107 92 90 89 86 112 91 87

% matcha

59.26 74.19 70.46 78.50 56.36 57.87 61.94 71.64 56.04 69.60 82.52 58.49 71.13 51.43 43.37 59.26 36.83 44.02 64.04 61.04 55.58 51.07 54.44 46.63 53.94 69.08 55.72 56.70 65.66 (59.68)

a % match ) 2 × no. of matched spots × 100/total no. of spots in gel1 + gel2. (See gel images in Figure 5).

ological conditions, temperature, humidity, and other environmental factors. Inter-Individual Variability. One of the most concerned issues in the urinary proteomics field (as well as in other subdisciplines of proteomics) is the inter-individual variability. We initially observed that the amount of recovered proteins obtained from male urine was not the same as that of female urine; the male urine tended to have greater amount of the recovered proteins (Table 6). On the other hand, the total number of resolved protein spots in 2D gels of male urine, however, tended to be smaller (Figure 5 and Table 9). There was a considerable degree of inter-individual variability as demonstrated by the lowest % match of 36.83% (Table 10). The average % match of this set of samples was less than that reported in the set of intra-individual variables (Table 8). Journal of Proteome Research • Vol. 5, No. 1, 2006 189

technical notes

Sample Preparation for Urinary Proteomics

Figure 6. Intra- and inter-individual variabilities of urinary excretion of albumin and transferrin. The results illustrate mean intensity volumes (central tendency), mean squared deviations (dispersion) and coefficients of variation of albumin (A) and transferrin (B) spots in different sets of samples, including the set of pool urine precipitated with varying % ethanol, the set of intra-individual variables; and the set of inter-individual variables. The data clearly confirmed that the coefficients of variability for both albumin and transferrin were greatest in the set of urine obtained from different subjects (inter-individual variability).

Intra- and inter-Individual Variabilities of Urinary Excretion of Albumin and Transferrin. To further evaluate intraand inter-individual variabilities, we compared the normalized spot intensity volumes of albumin and transferrin, which are the major abundant proteins in normal human urine (the proteome map of these and other proteins can be found in Figure S3). Quantitative intensity analysis of protein spots was performed using the intensity volume of each spot normalized with integral or total intensity volume of all spots within the same gel. In addition, background subtraction was also performed to further ensure the comparability of quantitative intensity analysis across different gels. Because albumin and transferrin normally have multiple isoforms in the urine, we therefore used the summation of normalized intensity volumes of all isoforms for each protein for the comparisons across different gels. The parameters used for comparison were as follows: (i) mean intensity volume (which represents the central tendency); (ii) mean squared deviation (which represents the dispersion of the datasthe squared root of the average squared difference of each sample value to the center location was calculated); and (iii) coefficient of variation (which is the dispersion divided by the central tendencysit measures the relative variability of a group by correcting for the magnitude of the data values, thus giving a measure that has no unit). Figure 6 illustrates these parameters obtained from albumin and transferrin spots in different sets of samples: the pool urine precipitated with varying % ethanol; the set of intra-individual variables; and the set of inter-individual variables. The data clearly demonstrated that the coefficients of variation for both albumin and transferrin were greatest in the set of urine obtained from different subjects (inter-individual variability). Additionally, female urine tended to have higher concentration of albumin and transferrin as compared to male urine. 190

Journal of Proteome Research • Vol. 5, No. 1, 2006

Conclusions Our data clearly demonstrated that sample preparation methods are crucial to determine the quantity (protein recovery yield) and quality (2D spot patterns or proteome profiles) of recovered urinary proteins. Therefore, selection of sample preparation methods is very important for a high-quality, largescale urinary proteomics study. There is no single perfect protocol that can be used to examine the entire urinary proteome as each method has both advantages and disadvantages compared to the others; all of them are, indeed, complementary. Combination of various sample preparation methods is, therefore, required to obtain the most quantitative and qualitative information. Finally, intra- and inter-individual variabilities are other crucial factors to be concerned.

Acknowledgment. We thank Dr. Prida Malasit (Head of Medical Molecular Biology Unit and a Senior Research Scholar of the Thailand Research Fund [RTA4680017]) for his advice and support, and Ms. Wipawanee Kasemworaphoom and Mr. Napat Songtawee for their technical assistance. V.T. is supported by Siriraj Preclinic Grant and Siriraj Foundation (D 2362: Surapongchai Fund). S.C. is a Ph.D. candidate in the Doctor of Philosophy Program in Immunology at Faculty of Medicine Siriraj Hospital and in the Medical Scholars Program (Ph.D./M.D.) supported by Mahidol University. Supporting Information Available: (Figure S1) 2D gel images of the pool urine precipitated with 50% methanol using different stains, including Coomassie Brilliant Blue R-250, SYPRO Ruby, and Deep Purple stains. (Figure S2) Number of unique protein spots obtained from each isolation protocol or from combination of more than one method. (Figure S3) The partial proteome map of normal human urine. This material is available free of charge via the Internet at http://pubs.acs.org.

technical notes References (1) Thongboonkerd, V. Proteomic analysis of renal diseases: Unraveling the pathophysiology and biomarker discovery. Expert Rev. Proteomics 2005, 2, 349-366. (2) Thongboonkerd, V.; Malasit, P. Renal and urinary proteomics: Current applications and challenges. Proteomics 2005, 5, 10331042. (3) Thongboonkerd, V. Proteomics in Nephrology: Current Status and Future Directions. Am. J. Nephrol. 2004, 24, 360-378. (4) Hampel, D. J.; Sansome, C.; Sha, M.; Brodsky, S.; Lawson, W. E.; Goligorsky, M. S. Toward proteomics in uroscopy: urinary protein profiles after radiocontrast medium administration. J. Am. Soc. Nephrol. 2001, 12, 1026-1035. (5) Schaub, S.; Rush, D.; Wilkins, J.; Gibson, I. W.; Weiler, T.; Sangster, K.; Nicolle, L.; Karpinski, M.; Jeffery, J.; Nickerson, P. Proteomicbased detection of urine proteins associated with acute renal allograft rejection. J. Am. Soc. Nephrol. 2004, 15, 219-227. (6) Weissinger, E. M.; Wittke, S.; Kaiser, T.; Haller, H.; Bartel, S.; Krebs, R.; Golovko, I.; Rupprecht, H. D.; Haubitz, M.; Hecker, H.; Mischak, H.; Fliser, D. Proteomic patterns established with capillary electrophoresis and mass spectrometry for diagnostic purposes. Kidney Int. 2004, 65, 2426-2434. (7) Haubitz, M.; Wittke, S.; Weissinger, E. M.; Walden, M.; Rupprecht, H. D.; Floege, J.; Haller, H.; Mischak, H. Urine protein patterns can serve as diagnostic tools in patients with IgA nephropathy. Kidney Int. 2005, 66. (8) Thongboonkerd, V.; Klein, E.; Klein, J. B. Sample preparation for 2-D proteomic analysis. Contrib. Nephrol. 2004, 141, 11-24. (9) Marshall, T.; Williams, K. Two-dimensional electrophoresis of human urinary proteins following concentration by dye precipitation. Electrophoresis 1996, 17, 1265-1272. (10) Thongboonkerd, V.; McLeish, K. R.; Arthur, J. M.; Klein, J. B. Proteomic analysis of normal human urinary proteins isolated by acetone precipitation or ultracentrifugation. Kidney Int. 2002, 62, 1461-1469. (11) Lafitte, D.; Dussol, B.; Andersen, S.; Vazi, A.; Dupuy, P.; Jensen, O. N.; Berland, Y.; Verdier, J. M. Optimized preparation of urine samples for two-dimensional electrophoresis and initial application to patient samples. Clin. Biochem. 2002, 35, 581-589.

Thongboonkerd et al. (12) Oh, J.; Pyo, J. H.; Jo, E. H.; Hwang, S. I.; Kang, S. C.; Jung, J. H.; Park, E. K.; Kim, S. Y.; Choi, J. Y.; Lim, J. Establishment of a nearstandard two-dimensional human urine proteomic map. Proteomics 2004, 4, 3485-3497. (13) Tantipaiboonwong, P.; Sinchaikul, S.; Sriyam, S.; Phutrakul, S.; Chen, S. T. Different techniques for urinary protein analysis of normal and lung cancer patients. Proteomics 2005, 5, 11401149. (14) Spahr, C. S.; Davis, M. T.; McGinley, M. D.; Robinson, J. H.; Bures, E. J.; Beierle, J.; Mort, J.; Courchesne, P. L.; Chen, K.; Wahl, R. C.; Yu, W.; Luethy, R.; Patterson, S. D. Towards defining the urinary proteome using liquid chromatography- tandem mass spectrometry. I. Profiling an unfractionated tryptic digest. Proteomics 2001, 1, 93-107. (15) Davis, M. T.; Spahr, C. S.; McGinley, M. D.; Robinson, J. H.; Bures, E. J.; Beierle, J.; Mort, J.; Yu, W.; Luethy, R.; Patterson, S. D. Towards defining the urinary proteome using liquid chromatography- tandem mass spectrometry. II. Limitations of complex mixture analyses. Proteomics 2001, 1, 108-117. (16) Celis, J. E.; Rasmussen, H. H.; Vorum, H.; Madsen, P.; Honore, B.; Wolf, H.; Orntoft, T. F. Bladder squamous cell carcinomas express psoriasin and externalize it to the urine. J. Urol. 1996, 155, 2105-2112. (17) Rasmussen, H. H.; Orntoft, T. F.; Wolf, H.; Celis, J. E. Towards a comprehensive database of proteins from the urine of patients with bladder cancer. J. Urol. 1996, 155, 2113-2119. (18) Pieper, R.; Gatlin, C. L.; McGrath, A. M.; Makusky, A. J.; Mondal, M.; Seonarain, M.; Field, E.; Schatz, C. R.; Estock, M. A.; Ahmed, N.; Anderson, N. G.; Steiner, S. Characterization of the human urinary proteome: a method for high-resolution display of urinary proteins on two-dimensional electrophoresis gels with a yield of nearly 1400 distinct protein spots. Proteomics 2004, 4, 1159-1174. (19) Smith, G.; Barratt, D.; Rowlinson, R.; Nickson, J.; Tonge, R. Development of a high-throughput method for preparing human urine for two-dimensional electrophoresis. Proteomics 2005, 5, 2315-2318.

PR0502525

Journal of Proteome Research • Vol. 5, No. 1, 2006 191