Refined Microdialysis Method for Protein Biomarker Sampling in

Article pubs.acs.org/ac

Refined Microdialysis Method for Protein Biomarker Sampling in Acute Brain Injury in the Neurointensive Care Setting Andreas P. Dahlin,*,† Karlis Purins,‡ Fredrik Clausen,‡ Jiangtao Chu,† Amir Sedigh,§ Tomas Lorant,§ Per Enblad,‡ Anders Lewén,‡ and Lars Hillered‡ †

Department of Engineering Sciences, Uppsala University, PO Box 534, SE-751 21 Uppsala, Sweden Department of Neuroscience, Division of Neurosurgery, Uppsala University Hospital, Uppsala University, SE-751 85 Uppsala, Sweden § Department of Surgical Sciences, Section of Transplantation Surgery, Uppsala University Hospital, Uppsala University, SE-751 85 Uppsala, Sweden ‡

S Supporting Information *

ABSTRACT: There is growing interest in cerebral microdialysis (MD) for sampling of protein biomarkers in neurointensive care (NIC) patients. Published data point to inherent problems with this methodology including protein interaction and biofouling leading to unstable catheter performance. This study tested the in vivo performance of a refined MD method including catheter surface modification, for protein biomarker sampling in a clinically relevant porcine brain injury model. Seven pigs of both sexes (10−12 weeks old; 22.2−27.3 kg) were included. Mean arterial blood pressure, heart rate, intracranial pressure (ICP) and cerebral perfusion pressure was recorded during the stepwise elevation of intracranial pressure by inflation of an epidural balloon catheter with saline (1 mL/20 min) until ̈ MD catheter and one surface modified with Pluronic F-127 (10 mm membrane, 100 kDa molecular brain death. One naive weight cutoff MD catheter) were inserted into the right frontal cortex and perfused with mock CSF with 3% Dextran 500 at a ̈ catheters showed unstable fluid recovery, sensitive to ICP changes, flow rate of 1.0 μL/min and 20 min sample collection. Naive ̈ catheters failed to deliver a stable fluid recovery. which was significantly stabilized by surface modification. Three of seven naive MD levels of glucose, lactate, pyruvate, glutamate, glycerol and urea measured enzymatically showed an expected gradual ̈ and surface modified catheters. The ischemic and cellular distress response to the intervention without differences between naive 17 most common proteins quantified by iTRAQ and nanoflow LC-MS/MS were used as biomarker models. These proteins ̈ MD catheters showed a significantly more homogeneous response to the ICP intervention in surface modified compared to naive with improved extraction efficiency for most of the proteins. The refined MD method appears to improve the accuracy and precision of protein biomarker sampling in the NIC setting.

C

after traumatic brain injury (TBI), making MD a preferred sampling method for biomarkers, such as cytokines, chemokines, and neurotrophic factors.5,10 Numerous in vitro studies have revealed that MD protein biomarker sampling is more complex than traditional small molecular biomarker sampling, involving protein−protein interaction, protein−surface interaction and biofouling (for references see.11,12 By using nano liquid chromatography (nanoLC) in combination with tandem mass spectrometry (MS/MS) we showed that the proteins adsorbed onto the MD membrane may be lost to biomarker analysis in the dialysate because they are prevented from crossing the microdialysis membrane.13 Additionally, there is concern that changes in ICP, a common phenomenon in acute brain injury patients, may influence MD

erebral microdialysis (MD) is well established for sampling of low molecular weight biomarkers of energy metabolic perturbation and cellular distress in the neurointensive care (NIC) setting.1 There is growing interest in MD for sampling of protein biomarkers of secondary injury mechanisms in acute traumatic and neurovascular brain injury in NIC patients.2−4 Evolving analytical methodology allowing for multiplex biomarker analysis in 1−25 μL individual samples opens a new possibility for temporal mapping of complex secondary injury cascades, such as inflammation and cell-specific injury components. Such data may serve as an important basis for improved characterization of complex injuries (e.g., traumatic brain injury, neurovascular brain injury) and help defining novel targets and treatment windows for novel neuroprotective drug development. In this context, recent studies in human NIC patients have appeared studying temporal patterns of inflammatory biomarkers.5−9 The study by Helmy et al. on multiple (n = 42) inflammatory biomarkers also supports the emerging notion that the innate immune system of the brain is activated early © 2014 American Chemical Society

Received: April 30, 2014 Accepted: July 30, 2014 Published: July 30, 2014 8671

dx.doi.org/10.1021/ac501880u | Anal. Chem. 2014, 86, 8671−8679

Analytical Chemistry

Article

catheter performance in vivo. Thus, Helmy et al.5 (Supporting Information Figure 2) found a significant correlation between ICP and fluid recovery (FR); the percentage of perfusate collected after passage through the catheter, with crystalloid perfusion medium in TBI patients that was abolished by the addition of 3% human albumin, suggesting that the colloid osmotic pressure of the perfusate is important for optimal MD catheter performance. These results have inspired research of the mechanisms and challenges involved with MD protein biomarker sampling. Our published in vitro studies in this area showed that, using large dextran colloids in the MD perfusate, we were able to stabilize the pressures within the MD system and achieve FRs close to 100%, which is the target for comparative studies. Additionally, by dynamically modifying the surfaces of the membrane and the inner tubing of the MD catheter by the triblock polymer coating (Pluronic F-127) we were able to decrease the protein adsorption and increase precision in FR, improving extraction efficiency for some proteins in human ventricular cerebrospinal fluid.11 By using nanoLC MS/MS analysis we showed that surface-modified MD membranes adsorbed 33% less proteins than control membranes.14 Our hypothesis is that the combination of large dextran colloids in the MD perfusate and lowering of protein adsorption to the MD membrane and tubing will reduce biofouling and improve FR and protein biomarker extraction efficiency, increasing the overall robustness of MD catheter performance. The present study aimed at testing this hypothesis by applying our refined MD methodology in a clinically relevant model of acute brain injury caused by a gradual elevation of ICP leading to brain death.15 We sought to evaluate the in vivo performance of surface modified clinical MD catheters regarding FR and extraction efficiency (also named relative recovery; RR) for traditional low molecular weight (LMW) biomarkers of energy crisis and cellular distress, as well as protein biomarkers. Additionally, the influence of ICP on FR and biomarker sampling performance was studied.

front of the coronal suture. In a separate burr hole a balloon catheter (Foley catheter, 14 Fr, Willy Rusch AG, Kernen, Germany) was placed epidurally 10 mm from the midline and 10 mm in front of the coronal suture above the left frontal lobe (Supporting Information Figure S-1). Arterial blood pressure and ICP was recorded continuously during stepwise elevation of ICP by inflation of the epidural Foley catheter. Cerebral perfusion pressure was calculated (mean arterial blood pressure, ICP). The experiment was started by an initial recording of 30 min baseline values for MD and physiological data. Thereafter, the intracranial volume elevation phase followed (Figure 1).

Figure 1. Data and sample collection scheme before and during the intervention. Gray bars illustrate the stepwise 1 mL fluid injections (1−11) into the implanted Foley catheter. CPP, cerebral perfusion pressure; ICP, intracranial pressure. Supporting Information Table 1 gives ICP and CPP data in detail.

■

EXPERIMENTAL SECTION Chemicals and Reagents. Acetonitrile (ACN) and acetic acid (HAc) were obtained from Merck (Darmstadt, Germany). Urea, trifluoroacetic acid (TFA), and Pluronic F-127 were obtained from Sigma-Aldrich (St.Louis, MO, USA). Dextran T500 (Molecular weight 500 kDa was purchased from Pharmacosmos (Holbaek, Denmark). Perfusion Fluid CNS composed of aqueous solution of 147 mM NaCl, 2.7 mM KCl, 1.2 mM CaCl2, and 0.85 mM MgCl2, was obtained from M Dialysis AB (Stockholm, Sweden). Ultrapure water was produced by a Milli-Q+ system, Millipore Corp (Marlborough, MA, USA). Animals, Anesthesia, and Brain Death Model. Seven pigs of triple breed and both sexes were included in the study; age 10 to 12 weeks weighing 22.2−27.3 kg. All studies were performed under general anesthesia of the animals. Details of the anesthesia procedure are described in the Supporting Information. A sagittal midline skin incision was performed and the skin was retracted. Three drill holes were made for placement of an intracranial pressure (ICP) monitor (Camino, Integra Neurosciences, Plainsboro, New Jersey, USA) and the two MD catheters (71 High Cut-Off Brain Microdialysis Catheter, M Dialysis AB, Stockholm, Sweden). The ICP probe was placed 5 mm behind the coronal suture and 7 mm from the midline in the right hemisphere and the MD catheters were placed in close vicinity to each other approximately 5 mm in

The volume of the epidural balloon was increased by 1 mL every 20 min up to 9−11 mL.15 Microdialysis Methodology. High molecular weight cutoff (100 kDa) brain microdialysis catheters, kindly provided by M Dialysis AB, Stockholm, Sweden (polyarylethersulphone membrane, 10 mm in length and 0.6 mm outer diameter) were surface modified according to our previously published work.11 Briefly, the surface modification was performed at room temperature by soaking the catheters in a 5% w/v solution of Pluronic F-127 for 24 h. The catheters were also continuously perfused with 5% w/v Pluronic F-127 solution during this time with a flow rate of 0.5 μL/min in order to cover the membrane and tubing. After the surface modification, the catheters were washed with water for 12 h. All MD catheters were perfused with modified Perfusion fluid CNS (M Dialysis AB) with 3% w/v Dextran 500 kDa added to stabilize fluid recovery.11 A CMA 402 MD Pump (CMA microdialysis, Kista, Sweden) was used and set to deliver a flow rate of 1.0 μL/min and samples were collected every 20 min in 0.5 mL of protein low binding vials (Eppendorf, Hamburg, Germany). Figure 1 illustrates the MD sampling in relation to the stepwise ICP intervention. MD vials were weighed before and after sampling to determine fluid recovery (FR). FR is defined as the volume of fluid recovered 8672



Article

in the vial divided by the volume of fluid pumped into the MD catheter. The latter was calculated from the flow rate of the MD pump, that is, 1 and 20 μL/20 min. The difference in weight of the vial before and after sampling was used to calculate the volume of fluid recovered. Chemical Analyses of Low Molecular Weight (LMW) Analytes. MD samples were analyzed using an ISCUSflex Microdialysis Analyzer (M Dialysis) for concentrations of the energy metabolic biomarkers glucose, lactate and pyruvate. The lactate/pyruvate ratio was calculated. The cellular distress biomarkers glutamate and glycerol were also analyzed.1 Urea was monitored to control the MD catheter performance.16 The remaining samples were stored at −70 °C until further analyzed for proteins. The ISCUSflex was automatically calibrated when started as well as every sixth hour using standard calibration solutions from the manufacturer (M Dialysis) and quality control samples at two different concentrations for each substance were run every weekday. Total imprecision coefficient of variation was 120%. A linear regression line is plotted with 95% confidence interval. Upper panel ̈ shows a significant positive correlation between FR and ICP in naive MD catheters with a high degree of variability (r = 0.30, p = 0.02, std error of estimate = 32.3). Lower panel shows that this correlation switched to a weak negative relation in modified catheters with markedly reduced scatter (r = −0.04, p = 0.04, std error of estimate = 4.8), suggesting a more stable MD performance as a result of surface modification. 8674



Article

named U-protein are uncharacterized proteins according to the uniprot database (www.uniprot.org, 07-OCHT-2013). Uncharacterized proteins are known to be present based on expressed sequence tag evidence, but their functions are not yet characterized. Vectors pointing straight up (green) in Figure 5 displays >50% increase in the second fraction compared to the first fraction. Vectors pointing straight down (red) display a >50% decrease when the second MD fraction is compared with the first fraction. A change of >50% is considered to be significant.20 Vectors pointing diagonally display either an increase (light green) or a decrease (orange) of 25−50%. These changes are not significant but they show interesting tendencies. Finally, vectors pointing to the right (yellow) range between 0.75 and 1.25 and show no difference between the two fractions. The vector diagram analysis showed a significantly more homogeneous protein pattern in response to the intervention in ̈ MD catheters. This was supported modified compared to naive by the principal component analysis; (Supporting Information Figure S-3 and S-4) indicating that the performance of the surface modified MD catheters did show a difference as ̈ catheters. Naive ̈ catheter from animals 1, compared to the naive 5, and 6 that failed to deliver stable FR also showed large variation in protein response. Catheters that fail in FR is not the only reason for high variability in protein response which is apparent in animal 7. The distinctive behaviors of the modified catheters are dual: first, they presented a tighter clustering which means more consistency on their performance; second, they promoted the sampling recovery for most of the proteins found in the dialysate. Both features are advantageous, if more consistent sampling performance and higher protein recovery are considered as goals for improving MD in vivo performance. Individual Protein Changes in Response to the Intervention. According to the uniprot database (www. uniprot.org, 07-OCHT-2013), 11 of the proteins have been characterized and data on their physical properties and functions is given in Table 2. Notably, as previously observed,11,21 we also detected proteins with significantly larger molecular mass than the designated 100 kDa MWCO of the membrane. For example, complement C3 that has a molecular weight of 184 kDa. One reason for this could be that complement C3 is cleaved into 11 subunits when activated which all are smaller than the MWCO of the membrane. Furthermore, MWCO of a membrane is equal to the molecular weight at which 80% of the molecules are prevented from flux through the dialysis membrane22 meaning that the MWCO is not an absolute measure of the pore size of the membranes. Numerical protein data was only used to estimate the individual response of the 11 characterized proteins (Table 2) to the intervention. A nonparametric Wilcoxon’s matched pair test on the difference in protein content between Protein sample 2 and Protein sample 1 were performed. We found that ̈ MD catheters 2 out of 11 proteins in samples from the naive showed a significant increase in response to the intervention whereas 9 out of 11 proteins appeared to increase significantly in samples from coated MD catheters (Table 2). This finding prompted a literature search for published data on these proteins as potential biomarkers showing that most of them have been studied in conjunction with acute brain injury. The results are briefly summarized in Table 2. Even though this study focused on the response patterns of the most abundant proteins identified by the proteomic analysis to compare the sampling performance of the MD catheters, some of the proteins listed in

Table 1. Number of Identified and Quantified Proteins in Each Individual Animal animal identified identified and quantified

1 26 14

2 41 28

3 36 20

4 22 11

5 33 21

6 52 37

7 42 35

Figure 2B (animal numbers are presented to the right to the data point for FR > 120%) and also in Supporting Information Figure S-2A. This is further demonstrated in Figure 4where the ̈ are presented for the FR and log2 of the ratio coated/naive LMW biomarkers. A value of zero means no difference between ̈ A negative value indicates better extraction coated and naive. ̈ catheter (white area in Figure 4) and performance of the naive a positive value indicates better extraction performance of the coated catheter (yellow area in Figure 4). The FR-failing catheters 1,5 and 6 also gave largest variance in Urea. Cellular distress biomarkers are more promoted by the coating and the ̈ energy metabolic biomarkers are slightly promoted in the naive ̈ catheters. The largest variation between the coated and naive catheters for the LMW biomarkers where found in the order of animal 1,3, 5, and 6. It was also found that the largest difference where in the six last sample fractions (7−12) except for glutamate, where the largest difference were found in the first six sample fractions, all originated from animal 3. Assuming faulty glutamate determination in animal 3 and ignoring those results leaves the largest LMW-biomarker variation in animals 1, 5, and ̈ catheters that failed in the fluid recovery 6, which were the naive test. The variances from the log2 ratios for LMW biomarkers are much larger than for the FR meaning that an adjustment of the concentrations with the FR would not affect the variances of the LMW-biomarkers and could not be used to correct the LMWconcentrations for a FR-failing catheter. But a coated catheter significantly reduces the failing frequency and would therefore give more reliable LMW-biomarker concentrations. Microdialysis Protein Biomarker Patterns. All identified and quantified proteins are presented in Supporting Information Table 2. The numbers of proteins found and quantified in each animal are presented in Table 1. In total, 92 unique proteins were identified in the seven animals. Of these 92 proteins, 66 proteins were also relatively quantified in one or more animals using the “require bold red criteria”, that is, requiring at least one tryptic peptide from each protein to be the most likely assignment for that particular protein sequence. However, only proteins that were quantified in ≥4 animals were considered for further evaluation (n = 17). These 17 proteins, presented in Figure 5, were used as protein biomarker models to study the extraction performance of the MD catheters. Because of the normalization of the data, ̈ proteins in MD samples from surface modified versus naive MD catheters were analyzed by vector diagrams to illustrate the protein pattern changes in response to the intervention. The proteins (Figure 5) are presented with their name, uniprot-database number (www.uniprot.org) and their individual response to the intervention The proteins are ranked according to their Mascot score, that is, the probability that the experimental data set matches the database data.19 The vectors represent the difference in relative amount between the second and the first fraction. Proteins that were identified but not relatively quantified in the dialysate fractions are indicated as Id. A white box indicates that the protein was not found in the sample. This is a reflection of the sample composition rather than related to MD catheter performance. Proteins that are 8675



Article

Figure 3. Microdialysate concentrations of low molecular-weight biomarkers in modified and control MD catheters in response to the intracranial pressure intervention. (a) Energy metabolism biomarkers, (b) cellular distress biomarkers, and (c) urea. The data shows a significant change in response to the intervention for both catheters for the energy metabolism biomarkers and cellular distress biomarkers but not for the urea.

fractions 6 by 6 (Figure 1) with the limitation that the earliest part of the ICP intervention could have influenced the protein content in protein fraction 1. However, since CPP at the end of the second hour (MD sample 6) was still above approximately 70 mmHg (Supporting Information Table 1), we consider this

Table 2 may turn out to be biomarkers of relevance for the NIC setting based on the present results and published biomarker data. Limitations of the Study. To obtain sufficient MD sample volume for the protein analysis, we had to pool the MD sample 8676



Article

Figure 4. Box plot showing the individual LMW-biomarker response for each animal. On the y-axis each animal is numbered and the x-axis shows ̈ catheter. A value of zero on the x-axis means no difference between coated and naive. ̈ A negative value indicates the log2 of the ratio coated/naive ̈ catheter (white area) and a positive value indicates better extraction performance of the coated catheter better extraction performance of the naive (yellow area). The line within the box is the median value and the box itself frames the first and third percentile. The whiskers mark the maximum and minimum value.

Figure 5. Vector diagram showing the response of the 17 most abundant proteins according to the Mascot score, quantified in ≥4 animals. Proteins with vectors are relatively quantified. Proteins tagged with “Id.” are identified but not quantified, that is, fulfilling the identification criteria only. Proteins are presented with names, their Uniprot database number (www.uniprot.org) and their response to intervention for surface modified and ̈ catheters. The relative protein amount after intervention (protein fraction 2, MD fractions 7−12) is compared with the relative protein amount naive from before intervention (protein fraction 1, MD fractions 1−6). An arrow pointing straight up (green) means that the particular protein has increased in concentration by >50% due to the intervention. An arrow pointing straight down (red) means that the concentration of the protein has decreased by >50% because of intervention, etc (see text). U-protein means uncharacterized proteins. White boxes means that the protein was not identified or quantified in the sample. Thus, a white box is a reflection of the sample composition, not related to MD catheter performance.

The study design only allowed for testing our refined MD method in the short term. It is our working hypothesis that the method in forth coming human studies will show a stable MD

potential confounder to be insignificant and, if anything, would imply an underestimation of the difference between Protein sample 1 and 2 (Figure 4). 8677



Article

Table 2. Data on the 11 Characterized Proteins Used in the Current Studya response to intervention coated probes

response to intervention naiv̈ e probes

potential biomarker for

osmotic pressure of blood oxygen metabolism iron transport oxygen metabolism energy metabolism inhibits lysosomal proteinases 6.71 lipid transport 5.40 ion transport

+ + + + NS NS

NS + NS + NS NS

blood-brain-barrier breakdown24 red blood cell degradation23 free iron in the brain tissue25 red blood cell degradation23 enhancing predictive power of S100B as a biomarker26 increased autophagy28

+ +

NS NS

membrane remodeling due to cellular trauma29 systemic response to cerebral injury30

5.92 lipid transport 6.06 innate immunity

+ +

NS NS

membrane remodeling due to cellular trauma29 activation of the innate immune response to injury27

protein

MW

pI

serum albumin hemoglobin subunit beta serotransferrin hemoglobin subunit alpha creatine kinase B-type cystatin C

69670 16166 78919 15170 20693 15734

5.98 6.70 7.34 8.74 7.19 8.98

apolipoprotein A-1 alpha-2-HS-glycoprotein, a.k.a fetuin-A apolipoprotein E complement C3

9178 38430 36665 184601

function

a MW, molecular weight (Da); pI, isoelectric point; protein function; response to intervention (+ statistically significant increase; NS, no significant increase in response to the intervention); potential role as a biomarker of acute brain injury according to published data.

■

performance also in the longer term. The reasoning is based on lowered protein adsorption and protein interaction on the membrane surface13,14 because of the surface modification. This in turn will delay the biofouling process and thereby prolong the lifetime of the MD catheter membrane in terms of in vivo biomarker sampling performance. Even though surface modification improved the MD performance with a more stable FR close to 100% with a more homogeneous protein response to the intervention, the coating did not improve protein extraction efficiency for all proteins. Future research will determine if protein recovery can be further improved by optimizing the coating method.

ASSOCIATED CONTENT

S Supporting Information *

Additional material as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org/

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Telephone: +46 18 4716805. Fax: +46 18 471 35 72. Notes

The authors declare no competing financial interest.

■

■

ACKNOWLEDGMENTS This study was supported by Swedish Research Council, VINNOVA Foundation, Uppsala Berzelii Technology Centre for Neurodiagnostics, Uppsala University Hospital, Selander Foundation, Åhlén Foundation. We would like to acknowledge Rolf Danielsson for statistical assistance, Margareta Ramström for mass spectrometric assistance, Marcus Sjödin for help with protein database searches, Inger Ståhl-Myllyaho for help with analysis of LMW biomarkers, Anders Nordgren and Monica Hall for their help during the animal phase and Visualize your Science for graphical assistance.

CONCLUSION ̈ Naive catheters showed unstable fluid recovery, sensitive to ICP changes, which was significantly stabilized by surface modification. MD levels of glucose, lactate, pyruvate, glutamate, glycerol and urea measured enzymatically showed an expected gradual ischemic and cellular distress response to the inter̈ and surface modivention without differences between naive fied catheters. Our interpretation is that the unstable FR in ̈ catheters affect the recovery of LMW biomarkers by naive ̈ increased variability. Conversely, the unstable FR in naive catheters had a significantly negative influence on the protein recovery pattern. The vector diagram (Figure 5) shows a more robust MD performance with a more homogeneous response pattern during the intervention in membrane coated catheters. Additionally, the coated MD catheters did promote the recovery for most of the proteins found in the dialysate according to the PCA evaluation (Supporting Information Figures S-3 and S-4). We submit that our refined MD method is a promising tool for more accurate and stable biomarker sampling in the NIC setting. In combination with emerging analytical methodology allowing for multiplexed biomarker analysis in small (1−25 μL) MD samples, it may provide a novel opportunity for temporal mapping of complex secondary injury cascades, such as inflammation and cell-specific injury components, directly in the injured human brain. Such data may serve as an important basis for improved characterization of complex injuries (e.g., traumatic brain injury, neurovascular brain injury) and identification of novel targets for intervention.

■

REFERENCES

(1) Hillered, L.; Vespa, P. M.; Hovda, D. A. J. Neurotrauma 2005, 22, 3−41. (2) Hillman, J.; Aneman, O.; Anderson, C.; Sjogren, F.; Saberg, C.; Mellergard, P. Neurosurgery 2005, 56, 1264−1268 discussion 1268− 1270.. (3) Hillman, J.; Milos, P.; Yu, Z. Q.; Sjogren, F.; Anderson, C.; Mellergard, P. Acta Neurochir. (Wien) 2006, 148, 319−324 discussion 324.. (4) Hutchinson, P. J.; O’Connell, M. T.; Nortje, J.; Smith, P.; AlRawi, P. G.; Gupta, A. K.; Menon, D. K.; Pickard, J. D. Physiol. Meas. 2005, 26, 423−428. (5) Helmy, A.; Carpenter, K. L.; Menon, D. K.; Pickard, J. D.; Hutchinson, P. J. J. Cereb. Blood Flow Metab. 2011, Feb; 31, 658−670. (6) Hillman, J.; Aneman, O.; Persson, M.; Andersson, C.; Dabrosin, C.; Mellergard, P. J. Neurosurg. 2007, 106, 820−825. (7) Hutchinson, P. J.; O’Connell, M. T.; Rothwell, N. J.; Hopkins, S. J.; Nortje, J.; Carpenter, K. L.; Timofeev, I.; Al-Rawi, P. G.; Menon, D. K.; Pickard, J. D. J. Neurotrauma 2007, 24, 1545−1557. (8) Mellergard, P.; Aneman, O.; Sjogren, F.; Pettersson, P.; Hillman, J. Neurosurgery 2008, 62, 151−157 discussion 157−158..

8678



Article

(9) Mellergard, P.; Aneman, O.; Sjogren, F.; Saberg, C.; Hillman, J. Neurosurgery 2011, 68, 12−19 discussion 19.. (10) Helmy, A.; Antoniades, C. A.; Guilfoyle, M. R.; Carpenter, K. L.; Hutchinson, P. J. PLoS One 2012, 7, e39677. (11) Dahlin, A. P.; Wetterhall, M.; Caldwell, K. D.; Larsson, A.; Bergquist, J.; Hillered, L.; Hjort, K. Anal. Chem. 2010, 82, 4376−4385. (12) Helmy, A.; Carpenter, K. L.; Skepper, J. N.; Kirkpatrick, P. J.; Pickard, J. D.; Hutchinson, P. J. J. Neurotrauma 2009, 26, 549−561. (13) Wetterhall, M.; Bergquist, J.; Hillered, L.; Hjort, K.; Dahlin, A. P. Eur. J. Pharm. Sci. 2014, 57, 34−40. (14) Dahlin, A. P.; Hjort, K.; Hillered, L.; Sjodin, M. O.; Bergquist, J.; Wetterhall, M. Anal. Bioanal. Chem. 2012, 402, 2057−2067. (15) Purins, K.; Sedigh, A.; Molnar, C.; Jansson, L.; Korsgren, O.; Lorant, T.; Tufveson, G.; Wennberg, L.; Wiklund, L.; Lewen, A.; Enblad, P. Crit. Care Med. 2011, 39, 512−517. (16) Ronne-Engstrom, E.; Cesarini, K. G.; Enblad, P.; Hesselager, G.; Marklund, N.; Nilsson, P.; Salci, K.; Persson, L.; Hillered, L. J. Neurosurg. 2001, 94, 397−402. (17) Chu, J.; Hjort, K.; Larsson, A.; Dahlin, A. P. Biomed. Microdevices 2014, 16 (2), 301−310. (18) Keeler, G. D.; Durdik, J. M.; Stenken, J. A. Eur. J. Pharm. Sci. 2014, 57, 60−67. (19) Perkins, D. N.; Pappin, D. J.; Creasy, D. M.; Cottrell, J. S. Electrophoresis 1999, 20, 3551−3567. (20) Abdi, F.; Quinn, J. F.; Jankovic, J.; McIntosh, M.; Leverenz, J. B.; Peskind, E.; Nixon, R.; Nutt, J.; Chung, K.; Zabetian, C.; Samii, A.; Lin, M.; Hattan, S.; Pan, C.; Wang, Y.; Jin, J.; Zhu, D.; Li, G. J.; Liu, Y.; Waichunas, D.; Montine, T. J.; Zhang, J. J. Alzheimer’s Dis 2006, 9, 293−348. (21) Schutte, R. J.; Oshodi, S. A.; Reichert, W. M. Anal. Chem. 2004, 76, 6058−6063. (22) Maurer, M. H.; Berger, C.; Wolf, M.; Futterer, C. D.; Feldmann, R. E., Jr.; Schwab, S.; Kuschinsky, W. Proteome Sci. 2003, 1, 7. (23) Babu, R.; Bagley, J. H.; Di, C.; Friedm, A. H.; Adamson, C. Neurosurg. Focus 2012, 32, E8. (24) Blyth, B. J.; Farhavar, A.; Gee, C.; Hawthorn, B.; He, H.; Nayak, A.; Stocklein, V.; Bazarian, J. J. J. Neurotrauma 2009, 26, 1497−1507. (25) Park, U. J.; Lee, Y. A.; Won, S. M.; Lee, J. H.; Kang, S. H.; Springer, J. E.; Lee, Y. B.; Gwag, B. J. Acta Neuropathol. 2011, 121, 459−473. (26) Bazarian, J. J.; Beck, C.; Blyth, B.; von Ahsen, N.; Hasselblatt, M. Restor. Neurol. Neurosci. 2006, 24, 163−172. (27) Ducruet, A. F.; Zacharia, B. E.; Hickman, Z. L.; Grobelny, B. T.; Yeh, M. L.; Sosunov, S. A.; Connolly, E. S. Exp. Neurol. 2009, 219, 398−403. (28) Liu, Y.; Cai, H.; Wang, Z.; Li, J. K.; Wang, K. Y.; Yu, Z. Q.; Chen, G. Eur. J. Med. Res. 2013, 18, 21. (29) Mahley, R. W.; Huang, Y. Neuron 2012, 76, 871−885. (30) Wang, H.; Sama, A. E. Curr. Mol. Med. 2012, 12, 625−633.

8679


Refined Microdialysis Method for Protein Biomarker Sampling in

Recommend Documents