Abstract
Objective: To implement OFM-recirculation and OFM-suction capable of direct and absolute in-vivo quantification of albumin in the ISF of pigs. Approach: OFM-recirculation and OFM-suction were used to collect ISF in-vivo in pigs and lymph was collected from the same pigs after OFM sampling. Blood was collected before and after OFM sampling, plasma was isolated and mean albumin plasma concentrations per pig were used to yield albumin ISF-to-plasma ratios. We characterized the quality of the collected undiluted ISF via (1) stable albumin ISF-to-plasma ratio in OFM-recirculation and in OFM-suction samples, (2) comparison of albumin ISF-to-plasma ratios from OFM-recirculation and OFM-suction and (3) comparison of normalized albumin concentrations in the ISF and lymph. Main results: Both advanced OFM methods were successfully implemented and albumin was quantified from the collected ISF samples. OFM-recirculation reached stable albumin ISF-to-plasma ratios after 20 recirculation cycles. Absolute ISF albumin concentrations were 11.2 mg ml−1 (OFM-recirculation) and 14.2 mg ml−1 (OFM-suction). Albumin ISF-to-plasma ratios were 0.39 ± 0.04 (OFM -recirculation) and 0.47 ± 0.1 (OFM-suction). Significance: Knowledge of the ISF protein content is of major importance when assessing PK/PD effects, especially of highly protein bound drugs. Up to now, only blood albumin values have been available to determine the degree of protein binding in several tissues. OFM-recirculation and OFM-suction allow direct, absolute quantification of albumin in ISF for the first time and enable investigation of the degree of protein binding of a drug directly in its target tissue.
Export citation and abstract BibTeX RIS
Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Introduction
Interstitial fluid (ISF) is the site of action of numerous drugs, most of them bind reversibly to tissue proteins. Protein binding reduces the proportion of the free and thus pharmacologically active drugs according to the free drug hypothesis [1], which states that at steady-state and in absence of active transport a cell-permeable drug has the same free drug concentration on both sides of a biomembrane. It also states that this free drug concentration determines the pharmacological effects of the drug (pharmacodynamics, PD) [2]. Hence, protein binding has a major impact on a drug's PD and pharmacokinetics (PK). Knowledge about protein binding of a drug directly in the target tissue's ISF is especially important when assessing PK/PD effects of a drug during drug development processes and in particular for highly protein bound drug products that exert their action in tissue rather than plasma.
The most abundant protein in blood, plasma and ISF is albumin and as a universal binding protein in the body it has an outstanding position compared to other binding proteins. Its major function is to maintain the oncotic pressure in body fluids. Thus, albumin levels remain generally stable, independent of trauma reactions or inflammatory responses.
Accurate measurement of albumin in ISF and consequently accurate determination of the degree of protein binding of highly protein bound drugs is currently hindered by the lack of methods that can precisely assess large substances like proteins in ISF.
Commonly applied methods such as skin biopsies [3, 4], wicks or capsules implanted into tissue [5–8], suction blister [9, 10], capillaries or needles with applied vacuum [11] can collect ISF but their time-consuming nature, their invasiveness, and the small sample volumes limit their general use. Further, some of these methods alter the composition of the collected ISF due to tissue trauma, potential inflammation during implantation [5, 12] or due to contamination with intracellular fluid (i.e. with biopsies) [3, 4]. Methods which apply a vacuum also yield highly varying ISF albumin concentrations [3–8, 11, 13–15], indicating that the applied forces may additionally alter the ISF composition.
One way to avoid method-related limitations for protein quantification in the ISF is the use of lymph as an ISF surrogate. However, this approach is limited because lymph and ISF represent different body compartments [16] with differing protein concentrations due to an active exchange of water and proteins across the lymph vessels and in the lymph nodes [17–19].
ISF can be sampled with microdialysis from several tissues [20, 21] but most of the drug-protein interactions cannot be investigated because the microdialysis membrane excludes the majority of proteins [22, 23].
Open flow microperfusion (OFM) can be used to collect ISF without membrane-related exclusion of substances, and can thus be used for relative and absolute quantification of large ISF compounds like proteins [24–27]. Absolute quantification of compounds from diluted ISF samples is possible by combining standard OFM with several available calibration methods [24–26, 28]. However, direct and absolute albumin quantification in the ISF without additional calibration methods requires ISF samples, which are in equilibrium condition with the surrounding ISF, and the collection of such samples thus requires advanced OFM methods.
Up to now, only albumin concentrations measured in blood are available to determine the degree of protein binding of drug products. This may cause some bias, because although one may assume that the ISF contains dissolved substances in similar concentrations as plasma, the protein content in ISF is 50 to 80% lower than in plasma [10, 29, 30]. Direct, absolute quantification of albumin in ISF, which would enable accurate investigation of the degree of protein binding of especially highly protein bound drug in their target tissues is not possible to date.
Therefore, we aimed to implement OFM-recirculation and OFM-suction, which are both able to directly and absolutely quantify albumin in the ISF in pigs. Quality of the ISF collected with these two advanced OFM methods was assessed via (1) stable albumin ISF-to-plasma ratio in OFM-recirculation and in OFM-suction samples, (2) comparison of albumin ISF-to-plasma ratios from OFM-recirculation and OFM-suction and (3) comparison of normalized albumin concentrations in ISF and lymph.
Methods
Animals
All animal experiments were approved by the Austrian federal government (BMWFW-66.010/0047-WF/V/3b/2015) and performed according to Directive 2010/63/EU on the protection of animals used for scientific purposes. Seven domestic pigs with a weight of approximately 40 kg each were included in the study. Pigs were housed in groups of at least two animals and delivered to the animal facility one week before start of the experiment for acclimatization. Each pig was anesthetized with a premedication mixture of midazolam (0.5 mg kg−1), azaperone (2.5 mg kg−1), ketamine (10 mg kg−1), and butorphanol (0.2 mg kg−1). Anesthesia was induced after sufficient preoxygenation with propofol 1% (3 mg kg−1 bolus). Anesthesia was maintained using propofol 1% (2–5 mg kg−1 h−1), fentanyl (20 μg kg−1 h−1) and isoflurane gas 2%, if necessary. Furthermore, an isotonic electrolyte solution was administered with a rate of approximately 10 ml kg−1 h−1 during the first hour of anesthesia and then with a rate of 3 ml kg−1 h−1 until the end of OFM sampling. Each animal was sacrificed immediately after sampling by intravenous administration of potassium chloride.
Sampling of blood and lymph
Arterial blood was collected before and after OFM sampling, and plasma was isolated. Lymph was collected in living animals from the cisterna chyli after OFM sampling. Lymph samples that were contaminated with blood were excluded from analysis.
Sampling of undiluted ISF
OFM-recirculation
OFM probes were implanted into the subcutaneous tissue of each pig´s back and each implanted OFM probe was connected to a peristaltic pump (OFM pump) per PUSH and PULL tubing. A sample tube was mounted between the OFM probe and the PULL tubing. Standard OFM sampling was performed: the OFM probe was perfused with a physiological perfusate (ELO-MEL isotone, Fresenius Kabi, Austria) for 5 min to clean the inner OFM probe lumen from blood, which may have entered during implantation (figure 1 A). Then, the PUSH tubing was connected to the OFM probe to close the recirculation loop (figure 2, left side).
Figure 1. (A): Standard OFM setup for subcutaneous ISF sampling and start setup for OFM-recirculation to clean the inner OFM probe lumen from blood; (B): OFM-recirculation setup; (C): OFM-suction setup.
Download figure:
Standard image High-resolution image
Figure 2. Left side: Closing the recirculation loop by connecting PUSH tubing and OFM probe; right side: Picture of OFM-recirculation application site.
Download figure:
Standard image High-resolution imageThe length of the PUSH tubing was adjusted in accordance to the desired volume of perfusate for OFM-recirculation, and recirculation of the perfusate was performed with a flow rate of 1 μl min−1(figure 1 B). For small numbers of recirculation cycles, i.e. 5, 10 and 15, we experimentally approached steady state by using a smaller number of samples. When performing 20, 25 and 30 recirculation cycles where steady state was reached, we used a higher number of samples.
Each pig had three to six application sites with three OFM probes each (figure 2, right side). Five, 10, 15, 20, 25, and 30 recirculation cycles were performed in a closed loop, each on a separate application site. This resulted in three samples per defined recirculation cycle number. Samples were collected after the recirculation cycles and frozen at −80 °C. The inner volume of the OFM probe and the attached tubing for 5 to 25 recirculation cycles was 30 μl, resulting in sampling times of 2.5 h, 5 h, 7.5 h, 10 h and 12.5 h. The inner volume of the OFM probe and the attached tubing was reduced to 25 μl for 30 recirculation cycles, which resulted in a sampling time of 12.5 h. Five, ten and fifteen recirculation cycles were performed with three pigs to estimate the change of albumin concentration as a function of recirculation cycles. Twenty, 25 and 30 recirculation cycles were performed with seven pigs to ensure data reliability in a range where stable albumin concentrations were expected.
OFM-suction
The OFM probes were implanted into the subcutaneous tissue.
The PUSH part of the OFM probe was sealed and a sample tube was mounted between the OFM probe and the PULL tubing. The PULL tubing was connected to a vacuum pump (Wound Ex K1, Novuqare, The Netherlands) and a suction force of 38 mm Hg was applied (figure 1 C). Nine OFM probes were implanted next to each other into the subcutaneous tissue of each animal´s back (figure 3). Sampling was performed for one hour to clean the OFM probe from blood. This first sample was discarded. Then, three samples were taken after sampling durations of three hours each. Samples contaminated with blood were excluded from analysis. The following numbers of samples per sampling interval were used for analysis: No sample (sampling interval: 1 to 4 h); six samples (sampling interval: 4 to 7 h) and 12 samples (sampling interval: 7 to 10 h). The collected samples were frozen at −80 °C. OFM-suction was performed with three pigs.
Figure 3. Picture of OFM probes implanted in subcutaneous tissue for OFM-suction.
Download figure:
Standard image High-resolution imageAnalytics
Albumin was analyzed by using a bromocresol green assay (Roche Diagnostics, Mannheim, Germany). The procedure was adapted to be applicable for 96 well photometer plates and small sample volumes. Calibration standards were prepared using the assay calibrator and diluting it to concentrations ranging from 0.5 mg ml−1 to 10 mg ml−1 albumin. Precinorm® (Roche Diagnostics, Mannheim, Germany) was used as quality control and was further diluted 1 + 1 and 1 + 4 (V/V) with ELOMEL isotone (Fresenius Kabi Gemany) to yield three different concentrations. Five μl of calibration standard and quality control or sample were added to 125 μl of 95 mol l−1 citrate buffer at pH 4.1 (reagent 1) into a 96-well photometer half area plate (Corning, Maine, USA). After shaking for 2 min, a blank measurement was performed at 630 nm using an ELISA reader (Synergy HT, BIOTEK, Vermont, USA). Ten μl 0.66 mmol l−1 bromocresol green of 95 mmol l−1 citrate buffer at pH 4.1 (reagent 2) was added to the mixture and measured after 10 min at a wavelength of 630 nm. The method was tested with porcine albumin (1, 5, 10 and 15 mg ml−1), which gave an accuracy from 95% to 110%.
Data analysis
Data were analyzed by means of descriptive statistical methods. The mean plasma albumin concentration for each pig was calculated from the plasma albumin concentrations measured before and after OFM sampling and used to normalize the albumin concentrations in OFM ISF samples to yield ISF-to-plasma ratios. Lymph albumin concentrations were also normalized using the mean plasma albumin concentration for each pig to yield lymph-to-plasma ratios. The mean lymph albumin concentration was calculated from lymph samples of all seven pigs.
Results
Plasma and lymph
The mean plasma albumin concentration for all pigs, calculated from the mean plasma albumin concentrations for each pig was 27.9 ± 4.0 mg ml−1. The albumin lymph-to-plasma ratio was 0.63 ± 0.11. Use of these values yielded a mean calculated albumin lymph concentration of 17.6 mg ml−1.
OFM-recirculation
The albumin ISF-to-plasma ratios from samples after 5 to 20 recirculation cycles showed an increasing trend, whereas the albumin ISF-to-plasma ratios after 20, 25 and 30 recirculation cycles remained constant. Thus, equilibrium albumin concentration between perfusate and ISF was reached after 20 circulation cycles and the albumin ISF-to-plasma ratio at equilibrium was 0.39 ± 0.04 (figure 4). Using the mean plasma albumin concentration of 27.9 ± 4.0 mg ml−1, the calculated absolute albumin concentration in the ISF was 11.2 mg ml−1.
Figure 4. Albumin ISF-to-plasma ratios after 5, 10, 15, 20, 25 and 30 OFM-recirculation cycles. Sample numbers: 2 (5 recirculation cycles), 3 (10 recirculation cycles), 2 (15 recirculation cycles), 7 (20 recirculation cycles), 4 (25 recirculation cycles) and 5 (25 recirculation cycles).
Download figure:
Standard image High-resolution imageOFM-suction
Albumin ISF-to-plasma ratios from OFM-suction collected in two sampling periods were 0.46 ± 0.1 (4 to 7 h) and 0.47 ± 0.1 (7 to 10 h) (figure 5). All samples collected during the sampling period from 1 to 4 h were contaminated with blood and therefore excluded from analysis. The amount of OFM-suction samples contaminated with blood decreased over time, which led to a higher number of analyzed samples in later sampling intervals. The mean plasma albumin concentration for the three pigs used in the OFM-suction experiments was 30.3 ± 2.1 mg ml−1. Using this value yielded a calculated absolute albumin concentration of 14.2 mg ml−1 from OFM suction.
Figure 5. Albumin ISF-to plasma ratio in OFM-suction samples after six hours of sampling in two OFM-suction sampling intervals.
Download figure:
Standard image High-resolution imageComparison of normalized albumin concentrations from OFM-recirculation, OFM-suction and lymph
The normalized ISF albumin concentrations from OFM-recirculation and OFM-suction were 0.39 ± 0.04 and 0.47 ± 0.1 respectively and the albumin lymph-to-plasma ratio was 0.63 ± 0.11 (figure 6).
Figure 6. Normalized albumin concentrations from OFM-recirculation (ISF-to-plasma ratio), OFM-suction (ISF-to-plasma ratio) and lymph (lymph-to-plasma ratio).
Download figure:
Standard image High-resolution image
Discussion
We successfully implemented OFM-recirculation and OFM-suction, which enable direct, absolute quantification of albumin in ISF for the first time, and consequently allow investigation of the degree of protein binding of a drug in its target tissue. Evidence that the collected samples contain undiluted ISF is based on the following results: stable albumin ISF-to-plasma ratio in OFM-recirculation samples was achieved after 20 recirculation cycles and albumin ISF-to-plasma ratios were stable during six hours of sampling with OFM-suction. Albumin ISF-to-plasma ratios from the ISF collected with OFM-recirculation and OFM-suction were comparable in their physiological context and the relation of the normalized albumin concentrations in the ISF to that in lymph was in accordance with previous results [11].
Albumin concentrations derived from the ISF collected with OFM-recirculation and OFM-suction reflect the local tissue concentrations. Albumin concentration in the tissue of interest is highly relevant in drug development processes, as albumin is a universal binding protein in the body and knowledge of its ISF concentration allows accurate assessment of protein binding of a drug of interest directly in the target tissue.
Furthermore, albumin´s major function is maintaining the oncotic pressure in body fluids and its levels remain generally stable independent of trauma reactions that may be caused by OFM probe insertion or inflammatory responses. The albumin concentrations in tissue varied in accordance with plasma concentrations among the individual animals. To ensure comparable results, we normalized the albumin concentrations measured in lymph and the ISF using the plasma albumin concentration.
We measured higher albumin concentrations in lymph than in the undiluted ISF samples from OFM-recirculation and OFM-suction. This is in agreement with previously reported values [17, 19, 31] and underlines the conception that lymph and interstitium represent different fluid compartments with differing compositions. When interpreting the results, one has to keep in mind that the resulting albumin ISF concentrations derived from OFM-recirculation and OFM-suction reflect the local tissue albumin concentration, whereas the resulting albumin lymph concentration depicts a sum parameter across several tissues. The here observed albumin lymph-to-plasma ratio of 0.63 ± 0.11 was in line with previously reported values between 0.42 and 0.71 from studies where different methods and animal models have been used [11, 32–34]. The albumin ISF-to-plasma ratios derived from samples collected with OFM-recirculation and OFM-suction (0.39 and 0.47 respectively) were also in agreement with previously reported values ranging from 0.15 to 0.63 [5–7, 11, 24, 32, 35–40]. Reasons for such a wide range of published values lie in the use of various different ISF sampling methods e.g. wicks implanted in the tissue [5–7], suction blister [36], micro puncture [37] or ISF collected from excised muscle and skin tissue [35], and in the use of different animal models [5, 7, 11, 37–41].
Albumin ISF concentrations have also been measured in humans using a range of different methods [24, 35, 36, 42–44]. Suction blister is a commonly applied method in clinical studies to investigate biomarkers and to perform proteomics and metabolomics analysis of the collected blister fluid [42–44]. The resulting normalized albumin concentrations in the collected blister fluid were ranging from 0.3 to 0.5 [13, 14, 45–47], which are similar to the normalized albumin concentration in the ISF collected with OFM-recirculation and OFM-suction in our pig study. However, the composition of blister fluid generally differs from that of ISF due to inflammatory reactions induced during blister formation [12] and due to depletion of analytes in the vicinity of the probe or increased albumin flux from the capillary system into tissue because of the applied suction forces [9]. The slightly higher ISF albumin concentrations in OFM-suction samples relative to OFM-recirculation samples might also be due to the applied suction forces. This indicates that depletion of albumin in the tissue next to the probe is surpassed by the increasing albumin influx from capillaries into ISF prompted by the suction forces. Another reason for the higher albumin concentrations derived from OFM-suction than those from OFM-recirculation could be small, undetected amounts of blood in the ISF samples collected with OFM-suction.
A previous standard OFM study using No-Net-Flux calibration found albumin ISF-to-serum ratios of 0.15 for human adipose tissue ISF and 0.27 for human skeletal muscle ISF [24], which are both considerably lower compared with present results from OFM-recirculation and OFM-suction. Differences might be due to species-related disparities [47] or due to bias introduced by the calibration method applied in combination with standard OFM sampling. For example, the OFM No-Net-Flux protocol requires equal diffusion into and out of the OFM probe, which may not hold for large molecules (unpublished data).
Schaupp et al performed OFM-recirculation for the very first time [28]. However, the OFM-recirculation method we implemented in the here presented study differs in several aspects from the recirculation method by Schaupp et al, and includes several technological advancements. Schaupp et al performed the experiments in a closed loop and therefore, they did not have to detach the samples. OFM-recirculation we implemented in this study needed to collect the samples for subsequent bioanalytical analyses and thus samples had to be collected from the closed loop recirculation setup which required major adaptations of the setup. Further, Schaupp et al combined the OFM probe with a sensor and measured conductivity. Conductivity represents the sum of all ions in the sample and is much more sensitive to ions than to large molecules such as albumin, i.e. the equilibrium which was extrapolated from the measured data applies mainly to ions, whereas we reached equilibrium conditions with respect to the albumin concentration.
Albumin is far more difficult to detect and monitor than conductivity. The whole methodology needed to be adapted to this issue, e.g. one must ensure collection of sample volumes big enough for subsequent analytics.
OFM-suction is a substantial advancement in OFM technology because to the best of our knowledge it has never been done before. OFM-suction differs from standard OFM in several aspects: OFM-suction is operated in PULL mode only—in contrast to the PUSH-PULL mode used during OFM-recirculation and standard OFM. Also, no perfusate is used when performing OFM-suction. The ISF is drawn directly from the subcutaneous space applying suction force of 38 mm Hg. The suction force was carefully chosen because too high suction force may limit the sample volume of the collected OFM samples because the ISF space may collapse.
The here implemented advanced OFM methods have certain advantages but also limitations: a major advantage of OFM-suction is the absence of any tubings which prevents potential adsorption effects of analytes to the tubing's inner surfaces. This is especially beneficial when sampling lipophilic analytes.
A general limitation of OFM studies is the effect of tissue trauma caused by the implantation of OFM probes, which can affect the protein composition in the ISF. Tissue trauma and also bleeding during OFM experiments (particularly at the beginning of OFM-suction experiments and during the first few hours of the sampling) may lead to higher protein concentrations in the samples. OFM-recirculation and OFM-suction samples were thus only included in the analyses after any bleeding had ceased to keep trauma effects low and to avoid contamination of the ISF samples with intracellular and blood analytes.
Conclusion
The here presented advanced OFM methods have high significance in the context of PD and PK, in particular in protein binding issues. Knowledge of the ISF protein content is of major importance when assessing PK/PD effects during drug development processes, especially of highly protein bound drugs, because according to the free drug hypothesis the size of the bound fraction depends on the protein concentration in the respective tissue, which in turn influences the pharmacological effects of the drug.
Until now, only albumin blood concentration was available to determine the degree of protein binding of drug products in several tissues. Direct, absolute quantification of albumin in ISF is for the first time possible with OFM-recirculation and OFM-suction and allows accurate investigation of the degree of protein binding of a drug in its target tissue.
Results from this study will initiate further experiments to optimize the perfusate for OFM studies, which will allow assessment of the absolute concentrations of a wide spectrum of analytes in the skin. In ISF samples that are collected with OFM-recirculation and OFM-suction any albumin-bound analyte and all analytes smaller than albumin can be quantified. Such analytes represent the majority of therapeutic drugs. Improvement of experimental conditions will also allow the quantification of proteins larger than albumin.
Direct quantification of albumin in the ISF by using OFM-recirculation and OFM-suction will strongly promote basic research e.g. investigation of protein binding of drugs of interest directly in the target tissue and support drug development processes, e.g. by stability assessments of an active pharmaceutical ingredient at the site of its action in-vivo.
Acknowledgments
The authors thank Sonja Kainz, Peter Reisenegger, Jürgen Lancaj, Christian Dragatin, Christian Höfferer, Jürgen Feiel (HEALTH—Institute for Biomedicine and Health Sciences, Joanneum Research Forschungsgesellschaft m.b.H, Graz, Austria), Vladimir Bubalo (Medical University of Graz, Austria) for their help with the animal study and Selma Mautner (HEALTH – Institute for Biomedicine and Health Sciences, Joanneum Research GmbH, Graz, Austria) for critical review and editorial assistance with the manuscript. The authors further acknowledge the help of Daniel Paul, Maria Ratzer and Agnes Prasch with bioanalytics (HEALTH—Institute for Biomedicine and Health Sciences, Joanneum Research Forschungsgesellschaft m.b.H, Graz, Austria).
Additional information
Competing interests
None
Author contributions
JH, TB, FS initiated and designed the study. JH and SS conducted the animal study and JH, SS, RR contributed to data acquisition and collection. Data analysis was done by TA, JH, BB and GS. JH, BB, GS, TB, RR, FS and TA interpreted the data. JH, BB, GS and TB wrote the first draft of the manuscript and all authors critically revised the manuscript and gave their final approval for submission.
Funding
This project was funded by the Austrian Federal Ministry for Transport, Innovation and Technology.