Growth-Factor-Releasing Polyelectrolyte Multilayer Films to Control

Subscriber access provided by UNIVERSITY OF MICHIGAN LIBRARY

Article

Growth factor-releasing polyelectrolyte multilayer films for controlling cell culture environment Ivan Ding, Dalia Shendi, Marsha W. Rolle, and Amy M Peterson Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b02846 • Publication Date (Web): 04 Oct 2017 Downloaded from http://pubs.acs.org on October 10, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Growth factor-releasing polyelectrolyte multilayer films for controlling cell culture environment Ivan Ding1, Dalia M. Shendi2, Marsha W. Rolle2, Amy M. Peterson1,* 1

Department of Chemical Engineering, Worcester Polytechnic Institute, 100 Institute Road,

Worcester, MA 01609 2

Department of Biomedical Engineering, Worcester Polytechnic Institute, 100 Institute Road,

Worcester, MA 01609 *[email protected]

KEYWORDS

Polyelectrolyte multilayer, cell culture substrate, surface topography, release kinetics

ABSTRACT

Polyelectrolyte multilayers (PEMs) are of great interest as cell culture surfaces due to their ability to modify topography and surface energy, and release biologically relevant molecules such as growth factors. In this work, fibroblast growth factor 2 (FGF2) was adsorbed directly onto polystyrene, plasma-treated polystyrene, and glass surfaces with a poly(methacrylic acid)

ACS Paragon Plus Environment

1

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 46

and poly-l-histidine PEM assembled above it. Up to 14 ng/cm2 of FGF2 could be released from plasma-treated polystyrene surfaces over the course of 7 days with an FGF2 solution concentration of 100 µg/mL applied during the adsorption process. This release rate could be modulated by adjusting adsorption concentration, decreasing to as low as 2 ng/cm2 total release over 7 days using a 12.5 µg/mL FGF2 solution. Surface energy and roughness could also be regulated using the adsorbed PEM. These properties were found to be substrate and first layerdependent, supporting current theories of PEM assembly. When released, FGF2 from the PEMs was found to significantly enhance fibroblast proliferation as compared to culture conditions without FGF2. The results showed that growth factor release profiles and surface properties are easily controllable through modification of the PEM assembly steps, and that these strategies can be effectively applied to common cell culture surfaces to control cell fate. INTRODUCTION Regenerative medicine is a rapidly growing field that has the potential to impact millions of patients. In 2012 alone, 160,000 patients in the United States were treated using cell-based therapies. Yet, in the same year, over 120,000 people remained on the organ transplant list.1,2 Factors such as limited space on 2D cell culture surfaces, inability to replicate conditions found in vivo, along with a combination of high costs and low shelf life have resulted in low throughput in cell expansion for regenerative medicine.1 Attempts to correct these issues include coating of 2D culture surfaces with materials such as silane3 or vitronectin4 along with the creation of 3D scaffolds.5 Despite improvements in their ability to more accurately replicate tissue microenvironments, these surfaces have their own limitations, including increases in batch inconsistencies or slowed cell migration and attachment.4,5 These limitations, along with higher


2

Page 3 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

cost and processing time, make simple synthetic surfaces such as tissue culture plastic (TCP) more popular despite their inability to provide ideal cell culture conditions.6 Growth factors are signaling molecules capable of modifying cell phenotype and function such as migration, proliferation, differentiation, and attachment.7,8 Fibroblast growth factor 2 (FGF2), in particular, has been implicated in a large number of functions such as suppressing stem cell differentiation while increasing proliferation rate,8–11 inducing angiogenesis,12–14 and repairing cartilage.15,16 While the use of growth factors provides a pathway towards controlling cell behavior, several limitations still exist. Growth factors have short half-lives, necessitating frequent media replacement. At physiological conditions, FGF2’s half-life ranges from 4.7 to 13.7 hours, depending on culture media.17,18 The depletion and spiking of FGF2 concentration over the course of the culture time may result in inconsistent and even slowed proliferation rates.18,19 Additionally, frequent manual additions of growth factors leads to potential sources of contamination and human error, especially in larger scale production.1,20 While cells react strongly to growth factors, emphasis must be placed on substrate properties such as surface energy,21–23 roughness9,24 and stiffness,3,25 which can also greatly impact cell function. For many cells, high surface energies may improve attachment but reduce proliferation.26 Likewise, cells have mechanosensing capabilities, probing as far as several microns into a surface, and preferentially attach to surfaces with stiffness and topography similar to their native physiological conditions.27,28. Modification of surface properties alone has even been shown to change the rate of cell differentiation.3,29 However, a single combination of surface properties fit for every cell type does not exist. For this reason, a fully tunable surface is necessary to meet the ideal culture conditions for specific cell types as they arise. One class of


3

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 46

materials that can both modify surface properties and control growth factor release to meet this need are polyelectrolyte multilayers (PEMs). PEMs are versatile materials formed through alternating layer-by-layer adsorption of polycations and polyanions. Many PEMs are known to be biocompatible and have been explored for use in biomedical surface modification, including the release of biologically relevant molecules such as growth factors.30,31 Growth factors can be incorporated within PEMs between layers of polyelectrolytes,14 through post-assembly loading,32 encapsulation,18,31 or as the first layer on a substrate.30,33 These methods can potentially be used in parallel for release of multiple growth factors simultaneously.14 Properties such as thickness, surface energy and roughness can be controlled by polyelectrolyte choice,34 polyelectrolyte and salt concentration,34 pH,24,33 substrate, first or last polyelectrolyte,35,36 and the number of polyelectrolyte layers.36 The use of copolymers allows even greater flexibility, permitting normally cytotoxic polyelectrolytes to be used on culture surfaces.37 Crosslinking or incorporation of nanomaterials can make further modifications to PEM mechanical properties and growth factor release rates.38,39 All of these options highlight PEMs as a versatile category of materials for cell culture surface modification. Peterson et al. described a PEM system consisting of poly(methacrylic acid) (PMAA) and poly-l-histidine (PLH) capable of releasing both bone morphogenetic protein 2 (BMP2) and FGF2.33,30 Growth factors were adsorbed directly on an anodized titanium substrate and were released from PEMs over several weeks while maintaining bioactivity. This was achieved with a PEM consisting of only 5 bilayers. Direct growth factor adsorption as the first layer may be advantageous due to ease of protein adsorption to surfaces and short processing time to form only 5 additional PEM bilayers. Furthermore, the protein loading concentration and release kinetics can be tuned without crosslinking or chemical modification of the polyelectrolytes.33,40


4

Page 5 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Previous research by Almodovar et al. has also shown that FGF2 can be adsorbed onto TCP coated with a polysaccharide-based PEM and fibronectin resulting in improved mesenchymal stem cell proliferation. However, it is unknown if direct surface adsorption is effective..41 The current work aims to use FGF2-(PMAA/PLH)5 PEMs to evaluate the effects of PEM modification on essential properties such as roughness, wettability and release rate. Also investigated are the influence of substrate, growth factor concentration, and long-term storage of PEMs on the performance of an FGF2-releasing surface. These considerations are critical to the deployment of PEM coatings for both in vivo and in vitro applications. While assembly of BMP2-(PMAA/PLH)5 PEMs on titanium is well understood, information on the impact of alternative substrates (e.g., TCP) on PEM assembly for PMAA/PLH PEMs is limited.33,30 Protein adsorption and desorption, and their potential roles in PEM assembly, structure, and release profile are also discussed.

EXPERIMENTAL SECTION Materials. Unmodified polystyrene Petri dishes and plasma-treated polystyrene Petri dishes were purchased from Corning. Soda lime glass microscope slides were obtained from GSC International. These will be described as polystyrene, tissue culture plastic (TCP) and glass, respectively, from this point forward. Poly(methacrylic acid) sodium salt (PMAA, 30 wt.% solution, MW ~9500), poly-L-histidine hydrochloride (PLH, MW ≥5000), phosphate buffered saline (PBS) tablets (pH=7.4), hydrochloric acid (12.1M), Tris base, acetone and reagent grade ethyl alcohol were obtained from Sigma-Aldrich. Sodium hydroxide (10N) was obtained from VWR. Recombinant Human FGF-basic (154 a.a.) was obtained from Peprotech. Unless otherwise noted, all materials were used as received.


5

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 46

CRL2352 fibroblasts were obtained from the American Type Culture Collection (ATCC). Gibco Iscove’s Modified Dulbecco’s Medium (IMDM) solution, Gibco fetal bovine serum (FBS), penicillin streptomycin, 0.25%/2mM trypsin-EDTA, Dulbecco’s Phosphate Buffered Saline (DPBS) and alamarBlue® Cell Viability Reagent were obtained from Thermo Fisher Scientific.

PEM Coating Preparation. Substrates were cut into 1 cm2 squares and cleaned with deionized water, ethanol, and deionized water again in the described order. The materials were otherwise used as obtained before PEM assembly.100 µg/mL FGF2 solution was applied to one side of the substrate and allowed to adsorb for 15 minutes. After adsorption, the substrates were washed in deionized water three times for 1 minute each. Using a dip coater (6 Position Compact SILAR Coating System, MTI), the substrates were subsequently immersed into 1 mg/mL PMAA at pH=4 for 15 minutes. Next, two 90 second wash steps in pH=4 water were performed, removing loosely attached polyelectrolyte from the surface. Another 15 minute adsorption step was performed for 1 mg/mL PLH at pH=4 followed by two more wash steps. This process was repeated until 5 polyelectrolyte bilayers were formed, resulting in a FGF2-(PMAA/PLH)5 PEM. After PEM assembly, the substrates were dried using nitrogen gas and stored at 4°C overnight. To investigate the stability of FGF2 within PEM-coated substrates, samples were prepared in the same manner as previously described, but stored for either 3 days or 5 days at 4°C or ambient temperature (22°C, four conditions total). Due to the large number of substrates, the samples for this study were prepared in two separate sets within the same day and randomized to minimize the effect of environmental variation during preparation and storage


6

Page 7 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

1M HCl and 1M NaOH were used to adjust the pH values for the dip coating solutions. Salvi et al. showed that the effect of adjusting pH with acids and bases on salt concentration in polyelectrolyte solutions was negligible.33 FGF2 was reconstituted in 5 mM Tris buffer as specified in manufacturer protocols. Further dilutions to 50 µg/mL, 25 µg/mL and 12.5 µg/mL were performed in the same buffer. PEM surfaces for cell culture in a 96-well plate were prepared in a sterile culture hood using the same conditions as described for dip coating. 30 µL of FGF2 was used for the protein adsorption step to completely cover the bottom while minimizing adsorption along the sides of the wells. 100 µL of polyelectrolyte and wash water solution was added to each well during PEM formation. After each time interval, the liquid was aspirated manually using a multi-pipette. PEM surfaces were prepared with FGF2 at the following concentrations: 100 µg/mL, 50 µg/mL, 25 µg/mL and 12.5 µg/mL. TCP and (PMAA/PLH)5 surfaces without FGF2 were prepared as negative controls. TCP and (PMAA/PLH)5 surfaces with addition of exogenous FGF2 were prepared as positive controls. FGF2 Release. FGF2 release rate was investigated by immersing the 1 cm2 substrates (n=5) in 1 mL of PBS each and incubating at 37 ˚C. At each time point, 1 mL aliquots were collected and immediately stored at -30 ˚C. Fresh PBS was used to replace the collected aliquot. After the final time point, the remaining FGF2 and PEM was removed through successive washes of 0.5 mL of 0.1M HCl and 0.5 mL of 0.1M NaOH.30,40 The HCl and NaOH aliquots were combined, pH adjusted to pH=7.4, and frozen at -30 ˚C. ABTS sandwich ELISA (Peprotech) was used to quantify FGF2 release and was performed in accordance with supplier instructions. Aliquots from the release studies were thawed and returned to room temperature immediately prior to their use. Color monitoring of the ELISA


7

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 46

plates was performed using a Perkin-Elmer Victor3 multilabel reader with a 405 nm filter and 650 nm correction filter. PEM Assembly. PEM assembly was studied using quartz crystal microbalance with dissipation monitoring (QCM-D) via a Q-Sense E4 (Biolin Scientific) using polystyrene-coated sensors at a flow rate of 50 µL/min. PLH served as a positively charged analogue for FGF2 in the first layer. The first layer was allowed to adsorb for 15 minutes, followed by a 10 minute wash step. This was repeated for PMAA, then PLH again until 5.5 bilayers were formed. Polyelectrolyte solutions were prepared as described in the PEM formation section. The system was maintained at 22 ˚C and both the frequency and dissipation were recorded. Frequency data was used to determine mass adsorption through the Sauerbrey equation:

=

(1)

where m is the calculated adsorbed mass, is the resonant frequency, n is the overtone number, and C is the sensitivity constant, which is 17.7 ng/(cm2 Hz) for a 5 MHz QCM-D sensor. The Sauerbrey equation is appropriate for thin, rigid surfaces. In addition, Vogt et al. established that the Sauerbrey equation is valid for PEM films below 40 nm in thickness.42 Surface Characterization. Uncoated surfaces, (PMAA/PLH)5, and FGF2-(PMAA/PLH)5 surfaces were prepared for each substrate for atomic force microscopy (AFM) and water contact angle analysis. Surfaces were characterized at three random locations for three different 1 cm2 specimens using a Nanosurf NaioAFM, resulting in 9 images for each condition. The root-mean squared (RMS) roughness values and peak-valley heights were obtained under contact mode. The collected images were analyzed using Gwyddion. Water contact angle analysis was performed with a Ramé-Hart contact angle goniometer. 5 µL droplets of deionized water were deposited by allowing direct contact of the droplet onto the


8

Page 9 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

surface before syringe removal. This process was completed for 3 replicates on 3 separate 1 cm2 specimens totaling in 9 measurements per condition. Fibroblast Proliferation. Fibroblasts were cultured and maintained in complete growth medium (IMDM containing 10% FBS, 1% L-glutamate, and 1% penicillin streptomycin). An alamarBlue viability assay was used to compare the relative cell number of CRL-2352 fibroblasts in response to culture on a range of surface conditions as described in PEM Assembly. Cells at passage 6 were seeded at a density of 2500 cells/cm (7800 cells/well) in a 96 well plate, and cultured in 200 µL of IMDM containing 1% FBS for 96 hours at 37ºC and 5% CO2. This low serum media was used in order to isolate the effects of FGF2 on cell proliferation. After 96 hours of culture, 10% v/v of alamarBlue reagent in low serum media was used to replace the media of each well, and cells were incubated for an additional 4 hours before measurement. A standard curve based on seeded cell count was used to determine the linear range of the alamarBlue fluorescent signal for fibroblasts.. Fluorescence was measured using a Perkin Elmer Victor3 multilabel reader at 544 nm excitation and 590 nm emission wavelength. The fluorescent intensity was recorded and converted to an estimated cell count based on the standard curve (See Supporting Information Figure S1). Statistical Analysis. Statistical analysis was performed on the surface characterization and alamarBlue results using a single-factor ANOVA. Post hoc comparison between different experimental conditions were performed using the Tukey Honest Significant Difference (HSD) method. Potential outliers in the datasets were investigated using both the Grubbs’ test and Tukey range test and removed if deemed as outliers in both tests.


9

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 46

RESULTS AND DISCUSSION FGF2 release – effect of substrate. Substrate surface properties such as stiffness, roughness, and wettability have a large effect on cell function. Understanding the effect of coating different substrates on release rate and surface properties would provide guidelines for which conditions can be optimized for cell expansion. This would create a general framework for substrate selection, and modification, such as through plasma treatment before growth factor and polyelectrolyte adsorption, to control the final properties of the PEM-coated surface. To account for these factors, polystyrene, TCP, and glass were chosen as model substrates. 7-day FGF2 release profiles on different substrates are shown in Figure 1. All conditions were prepared at a 100 µg/mL concentration during adsorption. For this study, the PMAA and PLH layers were assembled at pH=4, as this condition had previously shown the greatest BMP2 release from (PMAA/PLH)5-coated anodized titanium surfaces.33 PEM-coated polystyrene released 11.0 ± 3.2 ng/cm2 FGF2, while PEM-coated TCP, a plasma treated polystyrene, released 13.7 ± 4.0 ng/cm2 FGF2 over 7 days. The glass substrate released 8.0 ± 2.1 ng/cm2 of FGF2. For all substrates, the release rate was highest in the first hour after immersion in PBS, but greatly decreased over the course of 1 day. There was a significantly lower release of growth factor compared to the results of Salvi et al. At the same assembly pH and time scale, 1400 ng/cm2 of BMP2 was released from anodized titanium. These differences were most likely due to lower growth factor loading from differences in both growth factor and substrate charges, resulting in depletion of FGF2 over the course of 2-3 days. This explanation is plausible given the differences in isoelectric point (pI) between FGF2 (pI= 9.6) and BMP2 (pI=7.4).40,43 An attempt was made to quantify the amount of FGF2 lost during the adsorption step, however the ELISA method is unable to properly measure the value (See Supporting Information Table S2) likely


10

Page 11 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

due to interference polyelectrolytes in solution with adsorbed antibodies. The estimated lower bound of FGF2 lost during PEM assembly is 37.94 ng for five 1 cm2 substrates, assuming accurate final FGF2 concentrations in the PEM preparation solutions.

Figure 1. a. FGF2 release from PEM-coated polystyrene, TCP, and glass over 7 days. b. Comparison of released and remaining FGF2 recovered after release study and acid/base wash. The data is arranged to show the total FGF2 recovered from the system. In all cases, FGF2 was


11

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 46

adsorbed to the substrate from a 100 µg/mL solution, then the PEM was assembled on the FGF2coated surface. Error bars represent standard deviation (n=5).

Despite plasma treatment for TCP, the two polystyrene substrates did not show large differences in total released FGF2, suggesting that the difference in surface charge did not greatly impact the release of proteins. However, there was a substantial difference in total FGF2 adsorption. Untreated polystyrene had a 47.1 ± 9.8 ng/cm2 total loading on the surface compared to 14.3 ± 4.0 ng/cm2 and 8.6 ± 2.1 ng/cm2 for TCP and glass respectively. As TCP and glass are more hydrophilic than polystyrene, this suggests that FGF2 loading is dependent on hydrophobic adsorption. Alternatively, despite the net positive charge of the protein compared to the negative charge of the substrate, loading may be obstructed due to differences in configuration on the surface and increased charge-charge repulsion when FGF2 is below its pI.44 The combination of hydrophobic and hydrophilic forces result in a complex interaction between protein, surface and polyelectrolytes. Two important substrate properties that may be strongly related to the resulting differences in release profile and loading are surface roughness and charge. These two factors have been shown to influence both protein adsorption and polyelectrolyte formation.45,46

Surface Characterization. AFM was used to measure surface topography. Uncoated substrates, (PMAA/PLH)5-coated substrates and FGF2-(PMAA/PLH)5-coated substrates were characterized for each tissue culture surface. Representative AFM images for each surface type are shown in Supplemental Figure S2. Figure 2 shows the impact of substrate and PEM on the roughness of cell culture surfaces. The high standard deviation within each condition indicates large variations in the topography of the commercially obtained substrates.


12

Page 13 of 46

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 2. a. Average peak-valley height and b. RMS roughness of unmodified, (PMAA/PLH)5coated substrates, and FGF2-(PMAA/PLH)5-coated substrates. Error bars represent standard deviation (n=9). Statistical significance within the same substrate sets are labeled as * (p

Growth-Factor-Releasing Polyelectrolyte Multilayer Films to Control

Recommend Documents