A novel dynamic layer-by-layer assembled nano-scalebiointerface: functionality tests with platelet adhesionand aggregate morphology influenced by adenosine diphosphate

Melanie G. Watson • Juan M. Lopez •

Mihaela Paun • Steven A. Jones

Published online: 2 March 2013

� Springer Science+Business Media New York 2013

Abstract An improved biointerface was developed,

dynamic layer-by-layer self-assembly surface (d-LbL), and

utilized as a biologically-active substrate for platelet

adhesion and aggregation. Possible clinical applications for

this research include improved anti-coagulation surfaces.

This work demonstrated the functionality of d-LbL bioin-

terfaces in the presence of platelet-rich-plasma (PRP) with

the addition of 20 lM adenosine diphosphate (ADP), a

thrombus activator. The surface morphology of the

experimental control, plain PRP, was compared to PRP

containing additional ADP (PRP ? ADP) and resulted in

an expected increase of platelet adhesions along the

fibrinogen d-LbL substrate. The d-LbL process was used to

coat glass slides with fibrinogen, Poly (sodium 4-styrene-

sulfonate), and Poly (diallydimethlyammonium chloride).

Slides were exposed to PRP under flow and static condi-

tions with and without 20 lM of ADP. Fluorescence

microscopy (FM), phase contrast microscopy (PCM),

atomic force microscopy (AFM), and field emission-scan-

ning electron microscopy (FE-SEM) were used to evaluate

platelet adhesions under the influence of varied shear

conditions. PCM images illustrated differences between the

standard LbL and d-LbL substrates. FM images provided

percent surface coverage values. For high-shear conditions,

percent surface coverage values increased when using ADP

whereas plain PRP exposure displayed no significant

increase. AFM scans also displayed higher mean peak

height values and unique surface characteristics for

PRP ? ADP as opposed to plain PRP. FE-SEM images

revealed platelet adhesions along the biointerface and

unique characteristics of the d-LbL surface. In conclusion,

PRP ? ADP was more effective at increasing platelet

aggregation, especially under high shear conditions, pro-

viding further validation of the improved biointerface.

Keywords Fibrinogen � Layer-by-layer self-assembly �Nanotechnology � PlateletsThis work is dedicated to Claire J. Watson, a true fighter and

inspiration to all. May you continue to illuminate our path.

All the authors contributed equally to this work.

M. G. Watson

Department of Biomedical Engineering, LeTourneau University,

Longview, TX 75607, USA

M. G. Watson (&)

2100 Mobberly Ave., Longview, TX 75602, USA

e-mail: melaniewatson@letu.edu

J. M. Lopez

Department of Biomedical Engineering, Universidad Popular

Autonoma del Estado de Puebla, Puebla 72410, Mexico

J. M. Lopez

Department of Biotechnology, Universidad Popular Autonoma

del Estado de Puebla, Puebla 72410, Mexico

M. Paun

Department of Mathematics and Statistics, Louisiana Tech

University, Ruston, LA 71270, USA

M. Paun

Department of Bioinformatics, National Institute of Research

and Development of Biological Sciences, Splaiul Independentei

nr. 296, Sector 6, Bucharest 060031, Romania

S. A. Jones

Department of Biomedical Engineering, Louisiana Tech

University, Ruston, LA 71270, USA

123

J Thromb Thrombolysis (2013) 36:448–457

DOI 10.1007/s11239-013-0905-0

Abbreviations

PRP ? ADP 20 lM of ADP added to PRP

d-LbL Dynamic layer-by-layer

Introduction

Platelets are the essential components of hemostasis. Once

produced, platelets circulate within the blood system

awaiting activation [1, 2]. When platelets come into contact

with collagen and/or other factors, they activate other

platelets by discharging coagulation and platelet factors,

such as adenosine diphosphate (ADP), serotonin, thrombin,

von Willebrand factor, and fibrin-stabilizing factors [3].

These agents cause a cascade effect, producing a positive

feedback mechanism [4, 5]. During this process, platelets

bind to one another to form aggregates. Furthermore,

platelet activation increases with increased shear stresses

and exposure time [6]. These platelet aggregates, thrombi,

have the potential to damage biological systems, such as,

vasculature and organs by either obstructing blood flow at

the site of formation or by dislodging fragments in the form

of emboli [7, 8]. When atherosclerosis is present in an

arterial blood vessel, plaque ruptures can lead to thrombi,

which can lead to serious health conditions, including death

[9]. Thrombi can further promote stenosis leading to a

more rapid narrowing or occlusion of the vasculature [10].

Current studies of thrombosis use collagen and fibrinogen

biointerfaces to localize and control platelet depositions to

examine the effects of thrombi under shear rate, lumen

narrowing and occlusion conditions [9, 11, 12].

ADP is a known platelet activator that is released from

dense granules when they are secreted during platelet

activation [13]. Because ADP is a platelet activator, it can

be used as a positive control for platelet activation,

encouraging greater deposition along a surface [14, 15].

One step in characterizing the positive feedback aspects of

platelet adhesion is to quantify the relationship between

concentrations of activators, specifically ADP, and activa-

tion, adhesion and aggregation. Our laboratory examined

the adhesion of platelets to fibrinogen surfaces with the

addition of ADP concentration to further validate the

functionality as a biointerface of the newly developed

d-LbL surface [16, 17].

Biologically-active surfaces were generated with the

traditional static LbL (s-LbL) technique [18–20] and the

novel d-LbL technique [16]. ADP was added to citrated

platelet-rich plasma (PRP) under static and flow conditions.

Platelet adhesions were identified using an AFM, FE-SEM,

and FM. By demonstrating expected ADP responses, even

in a calcium-depleted environment [21], the experimental

potential of these improved d-LbL biointerfaces was fur-

ther examined beyond the initial studies. Other research has

thoroughly documented the role that ADP plays in platelet

activation and adhesion. It is not the intent of this paper

to demonstrate a new mechanism for ADP activation

in platelets or to present the ADP behavior as novel.

Rather, because the role ADP plays as an agonist is so

well-documented, even at physiological shear stress of

approximately 15 dynes-cm2, it serves as a useful reference

to show no undue interactions are occurring between the

platelets and the novel surface preparation technique,

showing the expected ADP behavior under shear [22, 23].

Earlier work from our laboratory has demonstrated the

expected behavior with a platelet activation antagonist

under static and low-shear conditions [4–6]. The experi-

mental shear stresses used for this work ranged from 0.002

to 0.02 dynes-cm2 and served as a test bed for an expected

behavior. For future work, shear stresses can be scaled up

to physiological shear stresses. This study builds upon the

prior antagonist demonstration and adds the agonist ADP

functionality test, as well as advanced imaging to validate

the presence of platelets [21, 23]. Moreover, s-LbL sub-

strates consistently displayed bulk acridine orange fluo-

rescence dye across the surface with and without the

presence of platelet adhesions even after a rising procedure

due to the proposed rougher, porous s-LbL surface char-

acteristics trapping the dye within the LbL itself [16].

Therefore, our laboratory designed the d-LbL technique to

minimize the rough, porous effects allowing the acridine

orange dye, not directly attached to platelet adhesions, to

easily evacuate the d-LbL surface. Our laboratory dem-

onstrated the possibility of using d-LbL for surface gen-

eration, instead of statically layered biointerfaces, to

produce homogenous substrates that are thrombogenic,

generating no observable unwanted agonist or antagonist

reactions. By using the proposed d-LbL method, our lab-

oratory and others might gain better insight into the pre-

vention of coagulation for the aforementioned diseases.

One potential clinical application of the d-LbL technique is

the modification of cardiovascular stents to control

thrombus formation after angioplasty.

Methods

Chemicals

Poly (diallydimethlyammonium chloride) (PDDA), Poly

(sodium 4-styrene-sulfonate) (PSS), sodium phosphate

monobasic (PBS), AO hemi (zinc chloride) salt, tris buffer

saline tablet, fibrinogen fraction 1 type 1-S, hydrochloric

Acid (HCl), ADP and sodium citrate were purchased for

Sigma Aldrich (St. Louis, MO). Sodium hydroxide (NaOH)

pellets were purchased from Fluka, a subsidiary of Sigma

Aldrich (St. Louis, MO).

Platelet adhesion on d-LbL biointerfaces 449

123

Preparation of glass slides

Premiere� 75 9 25 mm plain glass microscope slides were

cleaned with 91 % isopropyl alcohol and de-ionized (DI)

water. Each slide was first inserted into a 50 mL polypro-

pylene conical tube filled with isopropyl alcohol and

oscillated on a VWR advanced digital shaker table at 120

rotations per minute for 7 min. The slides were then

inserted into 50 mL tubes containing DI water and oscillated

on the shaker table for 7 min, at a rate of 120 rotations per

minute. Immediately after this procedure, the slides were

dried and stored until needed for a PRP experiment.

Static layer-by-layer self assembly (s-LbL)

Fifteen clean glass slides were immersed into 750 mL

rectangular dishes, each containing of cationic polyelec-

trolyte PDDA. The slides were removed after 10 min, and

immersed into another set of rectangular dishes containing

DI water. After this 1-min rinse, the slides were placed in

temperature- and humidity-controlled chamber during the

drying process. Following drying, the slides were

immersed into dishes containing PSS. The slides were

removed from the PSS solution after 10 min. Finally, the

slides were rinsed and dried following the same procedure

as previously described. This process was repeated four

more times for a total of five bilayers to form the initial

substrate.

To add the biologically active interface, a layering pro-

cedure similar to the method previously described was fol-

lowed. Each glass slide was exposed to PDDA for 10 min,

rinsed with DI water for 4 min, exposed to fibrinogen for

20 min, and rinsed again with DI water for 4 min. The pro-

cess was repeated to generate a total of 3 bilayers. The

completed surface was thus consisted of 8 bilayers.

Dynamic layer-by-layer self assembly (d-LbL)

Fifteen clean glass slides were immersed into fifteen 50 mL

tubes, each containing 35 mL of cationic polyelectrolyte

PDDA, and oscillated in a horizontal position. The slides

were removed after 10 min, and inserted into another set of

50 mL tubes containing 45 mL of DI water, which was used

as a rinse. After a 1-min rinse, the slides were placed in

temperature- and humidity-controlled chamber during the

drying process. Following drying, the slides were inserted

into tubes containing 35 mL of PSS. The slides were

removed from the PSS solution after 10 min. Finally, the

slides were rinsed and dried following the same procedure as

before. This process was repeated four more times for a total

of 5 bilayers to form the initial substrate.

To add the biologically active interface, a layering

procedure similar to the method previously described was

followed. Each glass slide was exposed to PDDA for

10 min, rinsed with DI water for 4 min, exposed to

fibrinogen for 20 min, and rinsed again with DI water for

4 min. The process was repeated to generate a total of 3

bilayers. The completed surface was thus consisted of 8

bilayers.

Platelet preparation

Approximately 45 mL of bovine whole blood samples

were collected into 50 mL conical tubes at the university

dairy from the same bovine source for each experiment.

Each tube contained 5 mL sodium citrate prior to veni-

puncture allowing a blood-citrate ratio of 9:1. Our blood

collection procedures were approved by Louisiana Tech

University’s Institutional Review Board for ethical han-

dling of animal subjects. Blood samples were obtained

from the subcutaneous abdominal (milk) vein located on

the lower abdomen in front of the udder sack. After col-

lection, the tubes were sealed and gently rocked back and

forth to ensure homogeneous sodium citrate/whole blood

mixing. Sodium citrate is known to form a complex with

calcium [24], and the further dilution with calcium-free

PBS provided a potential for calcium-poor platelets after

incubation for 60–90 min between collection and the

experiments. Thus, all blood samples were taken to the

laboratory directly after collection and centrifuged within

30 min.

The blood from each 50 mL tube was divided into

smaller 15 mL centrifuge tubes and centrifuged at 1,000 rcf

for 25 min, at a temperature of 30 �C using the Eppendorf

Centrifuge 5804R model. A brake setting of 6 up and down

was applied. After spinning, PRP was extracted from each

tube. PBS solution was then added to the PRP bringing the

total mixture to back to the original 50 mL volume for each

tube. Validation of PRP required that platelet counts be

analyzed within the normal bovine platelet count range of

2.0–5.0 9 105 platelet/lL [25–27] and were determined to

consistently fall within range. The platelet counts were

examined using an Agilent 2100 Bioanalyzer�. Platelet

samples were loaded into wells on the lap-on-a-chip assay

tray along with Agilent Cell Assay kit solutions and analyzed.

Exposure of slides to PRP

The petri dish/slide setups were placed directly in the tem-

perature- and humidity-controlled chamber (35 ± 3 �C,

20 ± 3 % relative humidity). Each protein-coated glass

slide was placed in a Petri dish and covered with either plain

PRP or PRP ? ADP for 30 min. To simulate flow condi-

tions, the petri dishes were placed on a VWR Advanced

Digital shaker table set at 60 rotations per minute. After the

exposure time, slides were gently rinsed, dried, stained with

450 M. G. Watson et al.

123

AO, rinsed and dried again. No further platelet fixation,

counterstaining, mounting or preparation methods were

necessary due to a potential for de-staining and/or masking

the fluorescence. Furthermore, the AO staining method

described above does not lead to cellular autolysis for

platelets and allows for preservation of cellular detail under

ultraviolet light [28].

Microscopy and imaging systems

FM

All slides were imaged using the Olympus IX51 fluores-

cence microscope with an Olympus DP71 camera. The

settings for the microscope are as follows: an exposure

time of 667 ms with a 109 magnification. The AO staining

was imaged with a TRITC filter. Slides that were subjected

to flow conditions were partitioned into three regions,

corresponding to low, mid and high shear. For each shear

region, 40 non-overlapping, *2 mm by 1.5 mm images

were obtained.

IrfanView was used to convert all image files from tiff

format to jpeg. An in-house MATLAB m-file resized each

image to 640 by 480 pixels, performed an image subtrac-

tion, converted the image to black and white, and counted

the percent of pixels above a threshold level that indicated

AO staining. However for the purposes of this research, the

original FM tiff files are presented as images in this article.

An additional in-house MATLAB program was used to

determine the sizes of the platelet aggregations in the

images.

PCM

Phase contrast images were collected by a Leica Micro-

systems LTD phase contrast microscope with a DFC290

camera. The settings for the microscope are as follows:

brightfield with a phase contrast ring using with a 109

magnification.

AFM

Dry nitrogen gas was washed over the sample slides to

remove any particulate matter before AFM imaging. The

slides were then placed on the ministage platform of the

Quesant Instrument AFM for scanning. Several

40 9 40 lm sections were taken of each slide.

FE-SEM

A Hitachi S-4800 FE-SEM was used to provide qualitative

characteristic results of platelets adhesions and the bio-

materials. Dry nitrogen gas was washed over the sample

slides to remove any particulate matter before FE-SEM

imaging. Each glass slide was cut into 1-inch squares and

placed on the scanning disk. Samples were loaded into the

microscope and imaged with a voltage setting of approxi-

mately 1–2 kV.

Statistical analysis

All PRP data obtained from the MATLAB program as a result

of the FM images were expressed as percent platelet sur-

face coverage ± SD. Additional statistical analysis tools

were used to calculate means, SDs, and non-parametric

Kruskal–Wallis tests. For the Kruskal–Wallis tests, p-val-

ues (p \ 0.05) were used to compare statistical significance

between groups and sets as a one-way analysis of variance

used to determine if three or more samples originate from

the same distribution. The test makes no assumptions about

the distribution of the data unlike the parametric indepen-

dent group ANOVA. The Kruskal–Wallis test was utilized

for this study due to the non-normally distributed biologi-

cal results obtained. A p-value of less than 0.05 was con-

sidered statistical significant.

Results

Surface morphology

Figure 1a, b illustrate PCM images characterizing the

surface roughness differences when using s-LbL (Fig. 1a)

and d-LbL (Fig. 1b) substrates. The image of Fig. 1a

suggests a substrate rougher in nature than the d-LbL

substrate of Fig. 1b. Figure 1c, d display FM images of the

s-LbL (Fig. 1c) and d-LbL (Fig. 1d) samples stained with

AO and exposed to plain PRP. Figure 1c displays the bulk

staining effect possibility due to the rougher, porous s-LbL

surface as compared to the d-LbL surface (Fig. 1d).

Statistical comparisons for s-LbL versus d-LbL surfaces

suggest significant differences in mean percent surface

coverage of platelet adhesions along the biointerfaces as

shown in Table 1. Under static conditions, s-LbL presented

a mean surface coverage of 5.515 % as compared to d-LbL

results of 0.472 % coverage. At higher shears, s-LbL mean

surface coverage results were 3.392 % versus d-LbL cov-

erage at 0.509 %.

Figure 2a, b show FM images of d-LbL samples stained

with AO and exposed to plain PRP (Fig. 2a) and

PRP ? ADP (Fig. 2b). The images were taken in the low

shear area of the slide. The number of platelet adhesions

appear similar between both images. However, the platelet

adhesions sizes of the PRP ? ADP samples, Fig. 2b, are

larger as compared to the plain PRP platelet adhesion sizes.

Additional, many of the PRP ? ADP platelet adhesions are

Platelet adhesion on d-LbL biointerfaces 451

123

elongated along the surface. The platelet adhesion elon-

gations, as evidenced in Fig. 2b, appeared in numerous

PRP ? ADP images of multiple slides. These elongations

were not found on any of the plain PRP images. When first

discovering these elongations, our laboratory was con-

cerned that they were simply scratches along the surfaces.

However, after further testing with the use of an AFM we

were able to eliminate concerns for surface scratches (see

Fig. 3b).

Figure 3a, b display AFM scans of d-LbL-fibrinogen

surfaces exposed to plain PRP (Fig. 3a) and PRP ? ADP

(Fig. 3b). Figure 3a exhibits isolated peaks, the largest of

which is 660 nm. Figure 3b displays elongated features

along the surface that are as much as 2,100 nm in height.

The PRP ? ADP mean peak heights consistently displayed

greater values (1521 ± 657 nm) than plain PRP (893 ±

464 nm). A Kruskal–Wallis test was performed to evaluate

the raw peak height data of plain PRP as compared to

PRP ? ADP. The p-value for this comparison was 0.003.

This result suggests a statistical significance between peak

height values of plain PRP to PRP ? ADP.

Figure 3c, d display FE-SEM images of the surface

morphology and platelet adhesion potential of the d-LbL

interface. The slides used for FE-SEM imaging were lay-

ered with fibrinogen, oscillated with PRP, and coated with

AO. Figure 3c indicates surface roughness and irregulari-

ties of the d-LbL biointerface. Figure 3d displays two

activated platelet adhesions fixed to the d-LbL surface

oscillated with plain PRP.

Additionally, AFM height histograms estimated the

average height and mean deviation of the d-LbL substrates

after immersion and drying of d-LbL bilayers 1 and 4. The

height histogram for bilayer 1 indicated an average height of

29.467 nm with a mean deviation of 14.22 nm along the

d-LbL substrate. The height histogram for bilayer 4 esti-

mated an average height of 48.878 nm with a mean deviation

of 19.54 nm along the d-LbL substrate. The histogram

Fig. 1 PCM images of LbL surfaces statically layered (a) and dynamically layered (b). FM images of LbL surfaces statically layered (c) and

dynamically layered (d)

Table 1 Mean percent surface coverage values and SD of d-LbL

versus s-LbL surfaces

Surface type Shear condition Mean % coverage SD

s-LbL Static 5.515 ±4.843

High 3.392

d-LbL Static 0.472 ±0.046

High 0.509

452 M. G. Watson et al.

123

results suggest an increase in average height and slight

increase in the mean deviation when dynamically adding

layers to the substrate.

Platelet adhesion sizes

Figure 4a, b illustrate the size distribution of platelet

adhesions along PRP ? ADP (Fig. 4a) and plain PRP

(Fig. 4b) images. Each curve in Fig. 4a, b represent a

different glass slide. The results were obtained by using a

particle-size identification program developed by Lopez

[16], setting different size thresholds for particles and

counting the number of particles identified at that threshold

level. The results for individual trials are shown as indi-

vidual lines, though for the purposes of this illustration it

was simply important to show the spread of the

Fig. 2 FM images of a d-LbL fibrinogen surface exposed to PRP under flow conditions without (a) and with added ADP (b) and stained with AO

Fig. 3 AFM and FE-SEM images of d-LbL fibrinogen surfaces

exposed to PRP under flow conditions. a Displays an AFM image of

plain PRP oscillated over the surface at low shear conditions.

b Displays an AFM image of PRP ? ADP oscillated over the surface

at low shear conditions. c Displays an FE-SEM image of the d-LbL

biointerface. d Displays an FE-SEM image of two activated platelets

adhered to the d-LbL biointerfaces

Platelet adhesion on d-LbL biointerfaces 453

123

size groupings rather than any individual trial, for

which grayscale is functional. In Fig. 4a, PRP ? ADP

displayed a greater number of aggregates over a

broader size range when compared to Fig. 4b’s plain

PRP results.

Mean percent surface coverage

Figure 5 summarizes the mean percent surface coverage

values, as calculated from FM images of d-LbL-fibrinogen

surfaces with and without ADP. For static and low to mid

shear dynamic conditions, results suggest PRP ? ADP did

not significantly increase nor decrease platelet adhesion

mean percent surface coverage values. For higher shears,

the PRP ? ADP mean percent surface coverage values

increased as compared to plain PRP. This result suggests

increasing shear rates while adding ADP to PRP does

increase the surface coverage of platelet adhesions along

the d-LbL biointerface, when compared to plain PRP. This

is the expected behavior from ADP addition, and serves as

a further validation of the surface functionality.

Fig. 4 Platelet aggregate size

distributions for d-LbL

fibrinogen surfaces exposed to

PRP ? ADP (a) and plain PRP

(b) and stained with AO

Fig. 5 Percent surface coverage of platelets on d-LbL-fibrinogen

surfaces after exposure to PRP, with and without added ADP, under

dynamic flow and static conditions. Error bars indicate one SD of

percent surface coverage

454 M. G. Watson et al.

123

For all imaging results, the automated standard image

processing removed individual images from the data set for

a slide if the percent surface area was 3 or more SDs

removed from the mean for the slide and thus considered

an outlier. This only dealt with the outliers in individual

slide sets, but did not consider the slide sets as a whole.

When the data were analyzed for entire slides, two of the

slides were excluded as being 3 or more SDs away from the

means for the rest of the slides in a test group.

Statistical analysis

Table 2 shows mean percent surface coverage values with

SD calculated from the raw percent surface coverage data

of plain PRP and PRP ? ADP experiments. Each mean

percent surface coverage value was determined by ana-

lyzing the raw data for each experiment over multiple days.

All plain PRP experiments performed under static condi-

tions taken over multiple days resulted in data that were

combined into a single file and processed using Excel. This

procedure was repeated for the other coverage values. For

plain PRP, the mean percent surface coverage values

suggested little difference between each shear condition as

evidenced from the SD values. The PRP ? ADP mean

percent surface coverage values with SD indicate an

increase in platelet adhesions along the surface from the

static to high shear conditions when compared to plain PRP

results.

Table 3 shows Kruskal–Wallis test p-values and mean

percent surface coverage values dependent upon shear

conditions for plain PRP and PRP ? ADP samples using

the d-LbL fibrinogen surface. Under static and low to mid

shear dynamic conditions, the Kruskal–Wallis p-values

suggest no significant effect between plain PRP and

PRP ? ADP. However, results for high shear conditions

indicate an increase in the surface coverage of platelet

adhesions. Furthermore, the high shear condition Kruskal–

Wallis p-value of 0.0002 indicates statistical significance

between plain PRP and PRP ? ADP results.

Kruskal–Wallis p-values were used to confirm consis-

tence between each experiment and the individual methods

performed. When observing the plain PRP sessions from a

single day, p-values indicated that the low shear condition

p = 0.331 and the mid shear condition p = 0.384 sug-

gesting consistence between percent surface coverage

values from individual images. For high shear conditions

when observing a single day’s experimental trials, these

results indicated statistical significance. These results

imply that platelet adhesion percent surface coverage val-

ues maintain continuity during experimental trials while

performed under static and low to mid shear conditions, but

varied greatly at higher shears. PRP ? ADP displayed

similar p-values dependent upon the shear condition.

Discussion

The purpose of this study was to examine, in combination,

the effects of adding ADP to PRP and increasing shear

stress on platelet surface morphology while using an

improved nano-scale biointerface made with the novel

d-LbL process. PCM and FM images (Fig. 1) of the s-LbL

and d-LbL substrates illustrated key differences between

the LbL techniques. Imaging results of s-LbL and d-LbL

suggest that s-LbL consistently maintained rougher sur-

faces possibility facilitating a bulk staining effect regard-

less of platelet adhesion attachment thus leading to false

positives in FM when using s-LbL substrates. Statistical

comparisons of the mean percent surface coverage values

between s-LbL and d-LbL (Table 1) indicate quantitative

differences between the two methods. The increased s-LbL

surface coverage results are possibly due to the bulk

staining effect and further validate a need for the improved,

novel technique.

The FM images (Fig. 2) of the d-LbL surfaces oscillated

with plain PRP or PRP ? ADP displayed platelet adhe-

sions along the biointerfaces and are confirmed at platelets

in FE-SEM images (Fig. 3). In the case of PRP ? ADP,

platelet adhesion sizes were larger and elongated. This

result was further validated by statistical AFM mean peak

height comparisons and Fig. 4’s platelet adhesion size

Table 2 FM used to obtain mean percent surface coverage values

with SD of PRP experiments

PRP experiment Shear condition Mean % coverage SD

Plain PRP Static 0.472 ±0.046

Low 0.417

Medium 0.517

High 0.509

PRP ? ADP Static 0.434 ±0.105

Low 0.368

Medium 0.450

High 0.614

Table 3 Statistical analysis information of FM images when com-

paring plain PRP results versus PRP ? ADP

PRP experiment

comparison

Shear

condition

Mean %

coverage

K–W test:

p-value

Plain PRP vs. PRP ? ADP Static 0.461 0.1477

Low 0.407 0.7507

Mid 0.451 0.4292

High 0.571 0.0002

Platelet adhesion on d-LbL biointerfaces 455

123

graphs. In these three cases, the information suggests that

platelet adhesion sizes increase respectively when adding

ADP to PRP and when increasing shear rates in comparison

to the plain PRP control variable. The surface morphology

of the d-LbL biointerface illustrated in FM, AFM, and

FE-SEM images indicate a substrate capable of activating

and adhering platelets with size and shape differences

between plain PRP and PRP ? ADP samples.

The graphs shown in Fig. 4 compare the platelet adhe-

sion sizes of plain PRP samples to PRP ? ADP samples.

These graphs demonstrate the larger platelet adhesion sizes

of PRP ? ADP versus plain PRP. As previously discussed,

ADP is known to activate platelets and encourage surface

deposition. When using the d-LbL method for biointerface

generation, the addition of ADP to PRP performs as

expected along the surface as compared to similar depo-

sition methods. The platelet adhesion elongation effect

observed in FM and AFM images from PRP ? ADP

experiments may contribute to the increased platelet

adhesion sizes. AFM scans for PRP ? ADP also demon-

strated significantly higher peak heights as compared to the

other plain PRP AFM images. The larger platelet adhesion

sizes and platelet adhesion elongation effect imply that the

addition of ADP to PRP modifies the general surface

morphology of adhered and activated platelets.

Mean percent surface coverage values for higher shear

conditions indicated a statistical difference when compar-

ing PRP ? ADP to plain PRP. This, however, is not the

case for all other shear conditions (Fig. 5; Table 3). This

information suggests that under high shear conditions,

PRP ? ADP experiments performed differently from that

of plain PRP. Because the shear rates were on the order of

10-3 when compared to physiological shear rates, more work

needs to be done to clarify the shear effects suggested here.

Table 2 results suggested that plain PRP produced

decreased mean percent surface coverage values when

compared to PRP ? ADP values. These results indicate

that increased shear stress and the addition of ADP to PRP

increase platelet adhesion percent surface coverage values.

Table 3’s mean percent surface coverage and Kruskal–

Wallis p-values implied that when comparing plain PRP to

PRP ? ADP the shear rates play an important factor in

prompting an ADP effect. Thus, results suggest that ADP

only affects platelet adhesions when oscillated along a

biointerface at higher shear rates.

Conclusion

This study was designed to determine whether the effect of

ADP on platelet adhesion at higher shear conditions would

demonstrate the expected increase in adhesion while per-

formed on the d-LbL surfaces, providing additional

validation to d-LbL process as a viable technology. Clini-

cal and experimental applications of this novel d-LbL

method include, but are not limited to, new types of bio-

materials producing increased thrombogenic surfaces or

anti-coagulation biointerfaces. Under static conditions, the

effect of ADP was not measurable, whereas added ADP

increased percent surface coverage at higher shear rates.

The addition of ADP tended to increase the platelet adhe-

sion size in the AO-stained images and led to a platelet

adhesion elongation effect that was seen in both FM and

AFM images. Although the different shear stresses appear

to modulate the effect of ADP, all shear stresses examined

in this study were much lower than typical physiological

values. The novel dynamic layering method, introduced in

this paper, for biointerface generation produced homoge-

neous surfaces capable of platelet adhesion as displayed in

FM, PCM, AFM and FE-SEM images. The d-LbL bioin-

terfaces performed well and eliminated a bulk staining

effect, providing a thrombogenic surface, platelet aggre-

gations arrested on the surface, and interacting as expected

with the addition of ADP under shear.

Acknowledgments We would like to acknowledge Dr. Mark A.

DeCoster for his assistance in using FM, Jim Spaulding for interesting

scientific discussions as well as the workforce at the Louisiana Tech

dairy for assisting in bovine blood acquisition. This work was sup-

ported in part from the REU Grant award #1062937 by providing

additional research assistance and Dr. Mihaela Paun’s Grant award #

CNCSIS TE-92/2010.

Conflict of interest No benefits in any form have been or will be

received from a commercial party related directly or indirectly to the

subject of this manuscript.

References

1. Platt W (1979) Color atlas and textbook of hematology, 2nd edn.

J. B. Lippincott Company, Philadelphia

2. Alberts B et al (2008) Molecular biology of the cell, 5th edn.

Garland Science, Taylor & Francis Group, LLC, New York

3. Guyton A, Hall J (2000) Text book of medical physiology, 11th

edn. W.B. Saunders Company, Philadelphia

4. Vyavahare K (2006) Computational modeling of nitric oxide

transport in a channel. Thesis, Louisiana Tech University

5. Frilot C (2001) Role of nitric oxide as a modulator of platelet

dense granule release. Thesis, Louisiana Tech Univerisity

6. Eshaq R (2006) The effect of the local concentrations of nitric

oxide and ADP on platelet adhesion and thrombus formation.

Thesis, Louisiana Tech University

7. Mannucci PM, Gorini S (1972) (eds) Advances in experimental

medicine and biology: platelet function and thrombosis: a review

of methods, vol 34. Plenum Press, New York

8. Sherry S, et al. (1969) (eds) Thrombosis. National Academy of

Science, Washington

9. Hinds M, Park Y, Giddens D, Aleviadou B (2001) Local hemo-

dynamics affect monocytic cell adhesion to a three-dimensional

flow model coated with E-selectin. J Biomech 34:95–103

456 M. G. Watson et al.

123

10. Jones SA, Fronek A (1986) Analysis of break frequencies

downstream of a constriction in a cylindrical tube. J Biomech

20(3):319–327

11. Matsusaki M, Ajiro H, Kida T, Serizawa T, Akashi M (2012)

Layer-by-layer assembly through weak interactions and their

biomedical applications. Adv Mater 24(4):454–474

12. Berny M, Patel I, White-Adams T, Simonson P, Gruber A,

Rugonyi S (2010) Rational design of an ex vivo model of

thrombosis. Cell Mol Bioeng 3(2):187–189

13. Spaet T, Zucker M (1964) Mechanism of platelet plug formation

and role of adenosine diphosphate. Am J Phys 206(6):1267–1274

14. Turner N, Moake J, McIntire L (2001) Blockade of adenosine

diphosphate receptors P2Y12 and P2Y1 is reguired to inhibit

platelet aggregation in whole blood under flow. Blood 98:

3340–3345

15. Remijn J, Wu Y, Jeninga E, IJssedldijk M, Willigen G, Groot P,

Sixma J, Nurden A, Nurden P (2002) Role of ADP receptor

P2Y12 in platelet adhesion and thrombus formation in flowing

blood. Arterioscler Thromb Vasc Biol 22:686–691

16. Lopez J (2010) An improved layer-by-layer technique to generate

biointerfaces for platelet adhesion studies: dynamic LbL. Dis-

sertation, Louisiana Tech University

17. Watson M (2010) Surface morphology of platelet adhesion

influenced by activators, Inhibitors and Shear Stress. Dissertation,

Louisiana Tech University

18. Zhao Q, Li H, Li B (2011) Nanoencapsulating living biological

cells using electrostatic layer-by-layer self-assembly: platelets as

a model. J Mater Res 26(2):347–351

19. Yu DG, Jou CH, Lin WC, Yang MC (2007) Surface modification

of poly (tetramethylene adiopate-co-terephthalate) membrane via

layer-by-layer assembly of chitosan and dextran sulfate poly-

electrolyte multiplayer. Colloids Surf B 54:222–229

20. Zhao Z, Qi Y, Wei M, Zhang F, Xu S (2012) Layer-by-layer

assembly and morphological characterizations of DNA/layered

double hydroxide thin films. Mater Lett 78:62–65

21. Rinder C, Student L, Bonan J, Rinder H, Smith B (1993) Aspirin

does not inhibit adenosine diphosphate-induced platelet a-granule

release. Blood 82(2):505–512

22. Weiss H, Turitto V, Baumgartner H (1986) Platelet adhesion and

thrombus formation on subendothelium in platelets deficient in

glycoporoteins IIb-IIa, Ib, and storage granules. Blood 67(2):

322–330

23. Anfossi G et al (2002) Adenosine increases human platelet levels

of cGMP through nitric oxide possible role in its antiaggregating

effect. Thromb Res 105:71–78

24. Kingston J, Bayly W, Sellon D, Meyers K, Waldrop K (2001)

Effects of sodium citrate, low molecular weight heparin, and

prostaglandin E1 on aggregation, fibrinogen binding, and enu-

meration of equine platelets. Am J Vet Res 62(4):547–554

25. Woodell-May J, Ridderman D, Swift M, Higgins J (2005) Pro-

ducing accurate platelet counts for platelet rich plasma: validation

of a hematology analyzer and preparation techniques for count-

ing. J Craniofacial Surg 16(5):749–756

26. Schnabel LV, Mohammed HO, Miller BJ, McDermott WG,

Jacobson MS, Santangelo KS, Fortier LA (2006) Platelet rich

plasma (PRP) enhances anabolic gene expression patterns in flexor

digitorum superficialis tendons. J Orthop Res 25(2):230–240

27. Ehrenfest DM, Rasmusson L, Albrektsson T (2008) Classification

of platelet concentrates: from pure platelet-rich plasma (P-PRP)

to leucocyte- and platelet-rich fibrin (L-PRF). Trends Biotechnol

27(3):158–167

28. Jahanmehr SAH, Hyde K, Geary CG, Cinkotai KI, Maciver JE

(1987) Simple technique for fluorescence staining of blood cells

with acridine orange. Tech Methods 40(8):926–929

Platelet adhesion on d-LbL biointerfaces 457

123