micromachines

Article
Aspartate Aminotransferase and Alanine
Aminotransferase Detection on Paper-Based
Analytical Devices with Inkjet
Printer-Sprayed Reagents
Hsiang-Li Wang, Chien-Hung Chu, Sing-Jyun Tsai and Ruey-Jen Yang *
Received: 30 October 2015; Accepted: 11 January 2016; Published: 15 January 2016
Academic Editor: Sergey S. Shevkoplyas

Department of Engineering Science, National Cheng Kung University, Tainan 70101, Taiwan;
sunshinesun5815@yahoo.com.tw (H.-L.W.); as195735@yahoo.com.tw (C.-H.C.); andrey.tsai@gmail.com (S.-J.T.)
* Correspondence: rjyang@mail.ncku.edu.tw; Tel.: +886-6-275-7575 (ext 63343); Fax: +886-6-276-6549

Abstract: General biochemistry detection on paper-based microanalytical devices (PADs) uses pipette
titration. However, such an approach is extremely time-consuming for large-scale detection processes.
Furthermore, while automated methods are available for increasing the efficiency of large-scale
PAD production, the related equipment is very expensive. Accordingly, this study proposes a
low-cost method for PAD manufacture, in which the reagent is applied using a modified inkjet
printer. The optimal reaction times for the detection of aspartate aminotransferase (AST) and alanine
aminotransferase (ALT) are shown to be 6 and 7 min, respectively, given AST and ALT concentrations
in the range of 5.4 to 91.2 U/L (R2 = 0.9932) and 5.38 to 86.1 U/L (R2 = 0.9944). The experimental results
obtained using the proposed PADs for the concentration detection of AST and ALT in real human
blood serum samples are found to be in good agreement with those obtained using a traditional
spectrophotometric detection method by National Cheng Kung University hospital.

Keywords: paper-based analytical devices; biochemistry detection; low-cost; inkjet printer

1. Introduction
Liver disorders such as hepatitis and hepatocirrhosis are among the most common health problems
globally, and are caused by many different factors, including excessive alcohol consumption, irregular
sleep patterns, lack of exercise, irregular meal times, and an over consumption of high calorie foods [1,2].
In clinical testing, the levels of aspartate aminotransferase (AST) and alanine aminotransferase (ALT)
provide important indicators of potential liver disease. In the blood serum of healthy adults, AST has
a concentration of around 5 to 40 U/L, while ALT has a concentration ranging from 5 to 35 U/L [3].
However, when the liver, heart, or muscle is diseased or damaged, the levels of AST or ALT rise.
Consequently, in detecting liver disease, effective methods for measuring the AST and ALT levels in
human blood are urgently required.
The concentrations of AST and ALT are generally evaluated using spectrophotometric [4,5],
chemiluminescence [6], colorimetric [7,8], or chromatography [9] methods. However, such methods
require the use of expensive bulky equipment and the intervention of skilled clinical staff. Accordingly,
the problem of developing low-cost and simple paper-based analytical devices (PADs) for disease
diagnosis purposes has attracted significant interest in recent years [10–12]. The first propose patterned
paper as diagnostic platform by the Whitesides group. PADs comprise cellulose fiber as the main
component, and hence sample transport is achieved by means of capillary forces alone without the
need for external driving mechanisms such as pumps. PADs have many advantages compared to

Micromachines 2016, 7, 9; doi:10.3390/mi7010009 www.mdpi.com/journal/micromachines

Micromachines 2016, 7, 9 2 of 10

traditional bench top systems, including lower reagent costs, reduced sample volumes, faster detection
time, lower cost, and greater portability. As a result, they have found increasing use in analytical and
clinical chemistry, food safety testing, environmental monitoring, and others [13,14].
PADs can be produced through photolithography, handcrafting, cutting, or printing methods.
Photolithography methods [15] require a UV lamp, hotplate, and a mask, and embed the photoresist
into the paper by means of UV exposure through the mask. In handcrafted methods, the devices are
fabricated via a manual wax drawing or stamping process [16]. Such methods are simple and require
no expensive equipment. However, they have a poor resolution and low throughput. Cutting methods
utilize computer-controlled knife-shaping [17] or laser cutting [18] techniques, and are thus ideally
suited to mass production. However, the equipment cost is typically rather high. The other low-cost
and repeatable cutting method to make PADs is craft punch patterning [19,20]. In wax-printing
methods [21–23], wax is printed on the surface of the paper in the desired configuration and a heating
process is then performed such that the wax melts and penetrates into the paper to form hydrophobic
channels. Through a careful control of the wax quantity, the heating time, and the heating temperature,
PADs with either hemi-channel or fully-enclosed channels can be easily formed [24]. Compared
to photolithography, handcrafting, and cutting methods, wax printing is inexpensive, rapid, and
extremely versatile. As a result, it is one of the most commonly used methods for producing PADs for
both prototype and mass production purposes.
In performing a biochemistry detection using PADs, the reagents and samples are generally
applied using pipette titration. However, such an approach is extremely time-consuming for large-scale
detection processes. To address this problem, several recent studies have demonstrated the use of
inkjet printing technology in depositing biomaterials or functional materials in carefully defined
regions of the paper surface [25–27]. Inkjet printing has many advantages for PAD manufacturing,
including a rapid throughput, low cost, a straightforward process, good versatility, and the potential
for mass production. Moreover, the cartridges of the printer can be filled with different reagents
in order to realize multi-assay devices in a single printing run [28,29]. The present study uses a
modified inkjet printer to spray biochemical reagents onto PADs for the detection of AST and ALT.
The feasibility of the proposed approach is demonstrated by comparing the experimental results for
the AST and ALT concentrations of human blood samples with the results obtained using a traditional
spectrophotometric approach by National Cheng Kung University hospital.

2. Materials and Methods

2.1. Modified Inkjet Printer

The biochemical reagents were sprayed onto the PADs using an Epson T50 inkjet printer
(Seiko Epson Corporation, Nagano-Ken, Japan). The printer was specifically chosen since it is a
piezoelectric printer and thus, compared to thermal inkjet printers, avoids the risk of heat-induced
damage to the reagent solution. In addition, the Epson T50 printer has a compact disc (CD) printing
mode, and thus allows the printing region to be defined with a high degree of precision.
Circuit damage and nozzle chamber clogging may render the inkjet printer head inoperable. Thus,
the exposed circuits in the printer head were sealed using Scotch tape (3M, St. Paul, MN, USA) to
protect against accidental reagent leakage. In addition, clogging was resolved by injecting ethanol into
the jet head until gas bubbles were no longer observed.

2.2. Fabrication of Paper-Based Analytical Devices

The PADs had dimensions of 20 mm ˆ 20 mm and contained four circular regions; each with a
diameter of 5 mm (Figure 1). (Note that the PADs were patterned with four detection regions in order
to facilitate their mass production. However, biological detection was performed using PADs with one
detection region only.) The PADs were designed using AutoCAD 2012 software (Autodesk, Marin, CA,
USA) and printed on ADVANTEC chromatography filter paper (No. 51B, Advantec MFS, Inc., Tokyo,
Micromachines 2016, 7, page–page

Micromachines 2016, 7, 9 3 of 10
with one detection region only.) The PADs were designed using AutoCAD 2012 software
(Autodesk, Marin, CA, USA) and printed on ADVANTEC chromatography filter paper (No.51B,
Advantec
Japan) usingMFS, Inc., Tokyo,
a commercial waxJapan)
printerusing
(Fuji aXerox
commercial
ColorQube wax8570
printer
solid (Fuji Xerox Fuji
ink printer, ColorQube 8570
Xerox, Tokyo,
solid inkFollowing
Japan). printer, Fuji Xerox, Tokyo,
the printing process,Japan). Following
the PADs the printing
were placed process,
in a furnace the PADs
(Vulcan™ were
A-550, placed
Dentsply,
in a furnace
ON, Canada) (Vulcan™
and bakedA-550, Dentsply,
at 155 ˝ C for 90 Ontario, Canada)
s to facilitate theand baked at of
penetration 155
wax°C into
for 90 s to
the facilitate
filter paper.
the penetration
Finally, of wax
the reagent into theonto
was sprayed filterthe
paper. Finally,
detection theofreagent
region the PADs was sprayed
using onto theEPSON
the modified detection
T50
region
printerof the PADsabove.
described using the modified EPSON T50 printer described above.

Figure
Figure 1.
1. (a)
(a) Schematic
Schematic illustration
illustration of
of paper-based
paper-based analytical
analytical device;
device; (b)
(b) Photograph
Photograph of
of paper-based
paper-based
analytical device.
analytical device.

2.3.
2.3. Preparation
Preparation of
of Reagent
Reagent Solutions
Solutions
The
The AST
ASTreagent
reagent(Roche,
(Roche,Mannheim,
Mannheim, Germany)
Germany) comprised
comprisedreagent 1 (R1)
reagent and and
1 (R1) reagent 2 (R2)2
reagent
solutions
(R2) solutionsmixedmixedin ainratio of of
a ratio 5:1.5:1.TheThe R1 R1solution
solutionconsisted
consistedofof100
100 mmol/L,
mmol/L, pH pH 7.8
7.8
tris(hydroxymethyl)aminomethane
tris(hydroxymethyl)aminomethane (TRIS) (TRIS) buffer,
buffer, 300 mmol/L
300 mmol/L L-aspartate,
L-aspartate, 0.23 mmol/L 0.23 mmol/L
nicotinamide
nicotinamide adenine dinucleotide
adenine dinucleotide (NADH), and (NADH),
0.53 U/mL and 0.53 U/mL
malate malate dehydrogenase
dehydrogenase (MDH). The(MDH). The R2
R2 solution
solution
consistedconsisted of 75 mmol/L
of 75 mmol/L α-ketoglutarate.
α-ketoglutarate. The equilibrium
The equilibrium reactionreaction
betweenbetween the and
the reagent reagent and
the AST
the ASThas
sample sample has the
the form formbelow.
shown shownThe below. The in
increase increase in oxaloacetate
oxaloacetate following following theprocess
the reaction reactionis
process
determinedis determined via reaction
via an indicator an indicator
catalyzedreaction catalyzed
by malate by malateMore
dehydrogenase. dehydrogenase. More
specifically, NADH
specifically,
is oxidized NADH is oxidized
to nicotinamide to nicotinamide
adenine dinucleotide adenine
(NAD+). dinucleotide (NAD+).
The reduction The reduction
in NADH in
is directly
NADH is directly
proportional to theproportional to theof
rate of formation rate of formation
oxaloacetate andofthus
oxaloacetate
of the ASTand thus of the AST activity.
activity.
AST
α-ketoglutarate + L-aspartate (1)
Øglutamate
�� L- + oxaloacetate
AST
α-ketoglutarate ` L-aspartate L -glutamate ` oxaloacetate (1)
MDH
oxaloacetate + NADH
oxaloacetate + H`+ H
` NADH Ømalate
�⎯`�MDH
L- + NAD
L -malate
+
` NAD` (2)(2)

The
The ALT
ALT reagent
reagent (Roche)
(Roche) also
also comprised
comprised R1R1 and
and R2
R2 solutions
solutions mixed
mixed inin aa ratio
ratio of
of 5:1.
5:1. The
The R1
R1
solution consisted of 125 mmol/L, pH 7.3 TRIS buffer, 625 mmol/L L -alanine, 0.23
solution consisted of 125 mmol/L, pH 7.3 TRIS buffer, 625 mmol/L L-alanine, 0.23 mmol/L NADH, mmol/L NADH,
and 1.5U/mL
and 1.5 U/mL lactate
lactate dehydrogenase
dehydrogenase (LDH).(LDH). The R2consisted
The R2 solution solutionof consisted
94 mmol/Lofα-ketoglutarate.
94 mmol/L
α-ketoglutarate. The enzyme
The addition of ALT addition prompts
of ALT enzyme promptsreaction
the equilibrium the equilibrium reaction
shown below, shown
in which thebelow, in
increase
which the increase in pyruvate prompted by the ALT activity is determined in an
in pyruvate prompted by the ALT activity is determined in an indicator reaction catalyzed byindicator reaction
catalyzed by lactate dehydrogenase.
lactate dehydrogenase.
AST
α-ketoglutarate + L-alanine AST
�� L-glutamate + pyruvate (3)
α-ketoglutarate ` L-alanine Ø L-glutamate ` pyruvate (3)
LDH
�� L-lactate + NAD+
pyruvate + NADH + H+ LDH (4)
pyruvate ` NADH ` H` Ø L-lactate ` NAD` (4)

2.4.
2.4. Reagent
ReagentSpraying
Spraying Process
Process
As shown ininFigure
As shown Figure2, 2,
thethe Epson
Epson T50 T50
printerprinter has
has six inksix ink cartridges
cartridges (i.e., six
(i.e., six single single
color color
channels),
channels), namely
namely yellow, yellow,
black, lightblack,
blue,light
lightblue, lightand
red, red, red,blue.
red, However,
and blue. However, the RGB
the RGB values values
of the lightofblue
the
light bluered
and light andcolor
light red color
channels are channels
not easilyare not easily
defined. Hence,defined. Hence, were
these channels thesenotchannels wereinnot
considered the
considered in theprocess.
reagent spraying reagent The
spraying process.
print head The print
contained head contained
540 nozzles (each with540 annozzles (each diameter
approximate with an

3
Micromachines 2016,
Micromachines 2016, 7,
7, 9page–page 4 of 10
Micromachines 2016, 7, page–page

approximate diameter of 18.09 μm) arranged in the form of a 90 × 6 array. The printing region in the
approximate diameter of 18.09 μm) arranged in the form of a 90 × 6 array. The printing region in the
CD18.09
of tray µm)
was arranged
defined using
in theEpson
form CD
of aPrinting Software
90 ˆ 6 array. The (Figure
printing3a,b). The
region inPADs
the CDwere
traythen
wasattached
defined
CD tray was defined using Epson CD Printing Software (Figure 3a,b). The PADs were then attached
to the surface of a CD in the corresponding region and inserted into the printing tray (Figure
using Epson CD Printing Software (Figure 3a,b). The PADs were then attached to the surface 3c).
of a CD
to the surface of a CD in the corresponding region and inserted into the printing tray (Figure 3c).
Finally,
in the reagent was
the corresponding injected
region into the into
and inserted yellow
the ink cartridge
printing in the printer
tray (Figure head the
3c). Finally, andreagent
sprayedwason
Finally, the reagent was injected into the yellow ink cartridge in the printer head and sprayed on
the papers.
injected into the yellow ink cartridge in the printer head and sprayed on the papers.
the papers.

Figure 2.
2. Epson
Epson T50
T50 printer
printer with
with six
six color
color channels.
channels.
Figure
Figure 2. Epson T50 printer with six color channels.

Figure 3. (a) Design of printing area using bespoke CD software; (b) Designated printing area on CD
Figure
Figure 3. (a)
(a)Design
Design of
ofprinting
printing area using
usingbespoke
areaattached
bespoke CD
CDsoftware;
software; (b) Designated
Designated printing
(b)area. printing area
area on
on CD
CD
printing3.tray; (c) Analytical device to designated printing
printing tray; (c) Analytical device attached to designated printing area.
printing tray; (c) Analytical device attached to designated printing area.
2.5. Color Changes with Reaction Time
2.5.
2.5. Color
ColorChanges
Changeswith withReaction
Reaction TimeTime
The color changes to be brighter with reaction time, a typical example is shown in Figure 4. In
The
The color changes to bebebrighter with reaction time, a typical example is shown in Figure 4. In4.
this case, we spray
color changesASTto reagent ontowith
brighter the PADs
reaction with
time, an ainkjet
typical printer
example andisdripshown clinical human
in Figure
this case, we spray AST reagent onto the PADs with an inkjet printer and drip clinical human
serum
In this onto
case, thewe reaction
spray AST areareagent
of the onto
PADsthe with PADsa pipette.
with an The images
inkjet show
printer andthedripcolor changes
clinical with
human
serum onto the reaction area of the PADs with a pipette. The images show the color changes with
reaction
serum ontotime thefor the reagent
reaction area ofand humanwith
the PADs serum from 6The
a pipette. to images
11 min.show The the
captured images were
color changes with
reaction time for for thereagent
reagent and human serum
fromfrom 6 min.
to 11The min. The captured were images were
analyzedtime
reaction by the the commercialand software
human Adobe
serum Photoshop
6 to 11 7.0 (Adobe Systems,images
captured San Jose, CA, USA).
analyzed
analyzed by the commercial software Adobe Photoshop 7.0 (Adobe Systems, San Jose, CA, USA).
Thethe
by Adobe Photoshop
commercial 7.0 software
software offers a built-in
Adobe Photoshop lasso function
7.0 (Adobe Systems,to Sanhighlight
Jose, CA, theUSA).
reaction
The zone
Adobe of
The Adobe Photoshop 7.0 software offers a built-in lasso function to highlight the reaction zone of
the devices, for which the RGB color gradation was selected for analysis. In this case, we selected
Photoshop 7.0 software offers a built-in lasso function to highlight the reaction zone of the devices, for
the devices, for color
which the RGBwas color gradation was selected thisfor analysis. In this thecase, we+ selected
the R +the
which G +RGBB values as the representative
gradation selected color intensity,Inbecause
for analysis. case,thewe(Red + Green
selected +RBlue)
+G RB+values
G+B
the R + G + B values as the representative color intensity, because the (Red + Green + Blue) R +G +B
values
as showed bettercolor
the representative linear distribution
intensity, becausewith theconcentration.
(Red + GreenIn the Adobe
+ Blue) R + GPhotoshop
+ B values 7.0 software,
showed better a
values showed better linear distributionInwith concentration. In the Adobe Photoshop 7.0 software, a
darker color will yield a smaller intensity value. However, in order to show darker colors clearly,
linear distribution with concentration. the Adobe Photoshop 7.0 software, a darker color will yield
darker
awe color will yield a smaller intensity value. However, in orderclearly,
to show darker colors clearly,
used the
smaller value 200
intensity value.to subtract
However, theindetection
order valuesdarker
to show to reverse
colors the results. weThe usedformula
the is: reverse
value 200 to
we used the value 200 to subtract the detection values to reverse the results. The formula is: reverse
color intensity
subtract = 200 −values
the detection color intensity.
to reverseTherefore,
the results.the Thecolor intensity
formula increased
is: reverse colorwith concentration
intensity = 200 ´ colorbut
color intensity = 200 −thecolor intensity. Therefore, the color intensity increased with with
concentration but
decreased with reaction time.
intensity. Therefore, color intensity increased with concentration but decreased reaction time.
decreased with reaction time.

4
4
Micromachines 2016, 7, 9 5 of 10
Micromachines 2016, 7, page–page
Micromachines 2016, 7, page–page

Figure 4. The color changes with reaction time for the AST reagent and human serum.
Figure
Figure 4.
4. The
The color
color changes
changes with
with reaction
reaction time
time for
for the
the AST
AST reagent
reagent and
and human
human serum.
serum.
3. Results
3.
3. Results
Results
3.1. Reagent Spraying Time
3.1.
3.1. Reagent
Reagent Spraying
Spraying Time
Time
Table 1 shows the time required to spray 4, 16, and 86 devices, respectively. As shown, the time
Table 1 shows
Table 1 shows the
the time
time required
required to
to spray
spray 4,4, 16,
16, and
and 86
86 devices,
devices, respectively.
respectively. As
As shown,
shown, the
the time
time
required to complete a 15-run spray process is 300, 600, and 1650 s, respectively. In other words, the
required
required to
to complete
complete aa 15-run
15-run spray
spray process
process is
is 300,
300, 600,
600, and
and 1650
1650 s,
s, respectively.
respectively. In
In other
other words,
words, the
the
average printing time per device (15-off) is equal to 75, 37.5, and 19.18 s, respectively. In other
average printing time per device (15-off) is equal to 75, 37.5, and 19.18 s, respectively. In
average printing time per device (15-off) is equal to 75, 37.5, and 19.18 s, respectively. In other words,other
words, the feasibility of the inkjet reagent spraying method for mass production purposes
words, the feasibility
the feasibility of reagent
of the inkjet the inkjet reagent
spraying spraying
method for massmethod for mass
production production
purposes purposes
is confirmed.
is confirmed.
is confirmed.
Table 1. Inkjet
Table 1. Inkjet printing
printing time
time for
for 4,
4, 16,
16, and
and 86
86 devices.
devices.
Table 1. Inkjet printing time for 4, 16, and 86 devices.

Inkjet
Inkjet Printing
Printing TimeTime 4 Devices
4 Devices 16Devices
16 Devices 86 Devices
86 Devices
Inkjet Printing Time 4 Devices 16 Devices 86 Devices
TimeTime of 15
of print print
runs15
(s)runs (s) 300 300 600
600 1650
1650
Timeofof print 15 runs (s) 300 600 1650
Average
Average time time ofaprint
print device a (s)
device (s) 75 75 37.5
37.5 19.18
19.18
Average time of print a device (s) 75 37.5 19.18

3.2.
3.2. Detection
Detection Results
Results for
for PADs Prepared via
PADs Prepared via Pipette
Pipette Titration
Titrationand
andInkjet
InkjetSpraying
Spraying
3.2. Detection Results for PADs Prepared via Pipette Titration and Inkjet Spraying
The feasibilityofofthethe
The feasibility inkjet
inkjet reagent
reagent spraying
spraying process
process was investigated
was investigated by comparing
by comparing the
the detection
The feasibility of the inkjet reagent spraying process was investigated by comparing the
detection resultsusing
results obtained obtained using the
the sprayed sprayed
PADs PADs
for AST andfor
ALT AST and ALT
samples withsamples with concentrations
concentrations ranging from
detection results obtained using the sprayed PADs for AST and ALT samples with concentrations
ranging from
5.4 to 91.2 U/L5.4
andto 5.38
91.2toU/L
86.1and
U/L,5.38 to 86.1 U/L,
respectively, respectively,
with with using
those obtained those PADs
obtained using with
prepared PADs a
ranging from 5.4 to 91.2 U/L and 5.38 to 86.1 U/L, respectively, with those obtained using PADs
prepared with a conventional pipette titration method. The corresponding results are shown in
conventional pipette titration method. The corresponding results are shown in Figure 5a,b, respectively.
prepared with a conventional pipette titration method. The corresponding results are shown in
Figure 5a,b,
Note that for respectively.
each sample, Note that for eachis sample,
the concentration evaluated theinconcentration is evaluated
terms of the mean R + G + Bincolor
terms of the
intensity
Figure 5a,b, respectively. Note that for each sample, the concentration is evaluated in terms of the
mean
in the R + G + B color
detection regionintensity in the Itdetection
of the PADs. region
is seen that of the
a good PADs. It is
qualitative seen that exists
agreement a goodbetween
qualitative
the
mean R + G + B color intensity in the detection region of the PADs. It is seen that a good qualitative
agreement
two sets of exists
resultsbetween
for eachthe two sets
sample. of results for
Consequently, each
the basicsample. Consequently,
feasibility the basic
of the proposed feasibility
inkjet reagent
agreement exists between the two sets of results for each sample. Consequently, the basic feasibility
of the proposed
spraying processinkjet reagent spraying process is confirmed.
is confirmed.
of the proposed inkjet reagent spraying process is confirmed.

Figure 5. (a) Comparison of AST detection results obtained using printer and pipette reagent
reagent
Figure 5. (a)
Figure 5. (a) Comparison
Comparison of
of AST
AST detection
detection results
results obtained
obtained using
using printer
printer and
and pipette
pipette reagent
application methods; (b) Comparison of ALT detection results obtained using printer and pipette
application
application methods;
methods; (b)
(b) Comparison
Comparison of
of ALT
ALT detection
detection results
results obtained
obtained using
using printer
printer and pipette
and pipette
reagent application methods.
reagent application
application methods.
reagent methods.

5
5
Micromachines 2016, 7, page–page
Micromachines 2016, 7, 9 6 of 10

3.3. Optimal Reaction Time

3.3. Optimal Reaction Time
When the sample (AST or ALT) is placed on the detection region of the PADs, the ensuing
reaction
When prompts a change
the sample (AST in or the
ALT)color intensity
is placed on theover time. For
detection practical
region of the detection
PADs, theapplications,
ensuing reaction the
optimal
promptsreaction
a changetime,in thei.e.,
colortheintensity
time at which the For
over time. color intensity
practical varies linearly
detection with the
applications, the optimal
sample
concentration
reaction time, must first
i.e., the timebeatdetermined. The variation
which the color intensity in the color
varies intensity
linearly with the was thus observed
sample over
concentration
time
mustfor
firstAST and ALT samples
be determined. with three
The variation different
in the concentrations
color intensity was thususing a digital
observed overcamera (Canon
time for AST
650D,
and ALTCanon, Tokyo,
samples withJapan). The captured
three different images were
concentrations usingprocessed using Adobe
a digital camera (CanonPhotoshop
650D, Canon, 7.0
software to determine
Tokyo, Japan). the corresponding
The captured change inusing
images were processed the mean
Adobe R +Photoshop
G + B intensity value in
7.0 software each case.
to determine
Figure 6a,b shows the
the corresponding changecorresponding
in the meanresults
R + G for
+ B AST solutions
intensity valuewith concentrations
in each case. Figureof 5.7,shows
6a,b 45.6 andthe
91.2 U/L, respectively,
corresponding results for AST andsolutions
ALT with solutions with concentrations
concentrations of 5.7, 45.6 andof 91.25.38, 43.05, and
U/L, respectively,
86.1 U/L, solutions
and ALT respectively.
with As shown, a linear
concentrations relation
of 5.38, between
43.05, and the color
86.1 U/L, intensity
respectively. Asand the asample
shown, linear
concentration
relation between is obtained
the colorafter approximately
intensity 6 minconcentration
and the sample for the AST samples
is obtainedandafter
7 min for the ALT
approximately
samples.
6 min for Intheother
AST words,
samplesthe andoptimal reaction
7 min for the ALT times for AST
samples. and ALT
In other detection
words, are equal
the optimal to 6times
reaction and
7for
min,
ASTrespectively.
and ALT detection are equal to 6 and 7 min, respectively.

Figure
Figure 6.
6. (a)
(a) Change
Change inin mean
mean R+G+B
R+G+B intensity
intensity value
value over
over time
time for
for three
three AST concentrations (5.7
AST concentrations (5.7 U/L,
U/L,
45.6 U/L and 91.2 U/L); (b) Change in mean R + G + B intensity value over time
45.6 U/L and 91.2 U/L); (b) Change in mean R + G + B intensity value over time for three ALT for three ALT
concentrations (5.38U/L,
concentrations (5.38 U/L,43.05
43.05U/L,
U/L,andand
86.186.1 U/L).
U/L). NoteNote that dotted
that dotted lines indicate
lines indicate the optimal
the optimal reaction
reaction timecase.
time in each in each case.

3.4.
3.4. Calibration
Calibration Curves
Curves
Figure
Figure 7a7a shows
showsthe
themeasured
measuredvaluesvaluesofof thethe mean
mean R +RG+ +GB+color
B color intensity
intensity afterafter
6 min6 min
givengiven
AST
AST concentrations of 5.7, 22.8, 45.6, 68.4, and 91.2 U/L, respectively. Similarly, Figure
concentrations of 5.7, 22.8, 45.6, 68.4, and 91.2 U/L, respectively. Similarly, Figure 7b shows the mean 7b shows the
mean intensity
intensity valuesvalues
obtainedobtained afterfor
after 7 min 7 min
ALTfor ALT concentrations
concentrations of 5.38,
of 5.38, 21.53, 21.53,
43.05, 43.05,
64.58, and64.58, and
86.1 U/L,
86.1 U/L, respectively. (Note that prior to the detection experiments, the reagents
respectively. (Note that prior to the detection experiments, the reagents and samples were placed in an and samples were
placed
ice bucketin toanprevent
ice bucket to prevent temperature-induced
temperature-induced deterioration. For each deterioration. For each detection
sample, concentration sample,
concentration
was then performeddetection was
under then performed
ambient temperature under ambient
conditions (37 temperature conditions
˝ C)). The correlation (37 °C)).
coefficient The
values
correlation coefficient values of the AST and ALT detection
2 results were 2 found
of the AST and ALT detection results were found to be R = 0.9932 and R = 0.9944, respectively. Thus, to be R2 = 0.9932 and

R
the=optimality
2 0.9944, respectively.
of the ASTThus,
and ALT the detection
optimalitytimesof the (6 AST
and 7andmin,ALT detection times
respectively) (6 and 7Inmin,
is confirmed. the
respectively) is confirmed. In the blood serum of normal healthy adults,
blood serum of normal healthy adults, the AST concentration is around 5~40 U/L while the ALT the AST concentration is
around 5~40 U/L while the ALT concentration is around 5~35 U/L. Consequently,
concentration is around 5~35 U/L. Consequently, for the PADs developed in the present study, mean for the PADs
developed in theintensities
R + G + B color present study,
greater mean
thanR166 + G(6+min)
B color
andintensities
160 (7 min)greater thanpossible
represent 166 (6 min) and 160
indicators of
(7 min) represent
liver disorder. possible indicators of liver disorder.

6
Micromachines 2016, 7, 9 7 of 10
Micromachines
Micromachines 2016,
2016, 7,
7, page–page
page–page

Figure
Figure 7.
Figure 7. (a)
7. (a) AST
(a) AST calibration
AST calibration curve;
calibration curve; (b)
curve; (b) ALT
(b) calibration
ALT calibration
ALT curve.
calibration curve.
curve.

3.5.
3.5. Detection
3.5. Detection Results
Detection Results Obtained
ResultsObtained using
Obtainedusing Paper-Based
usingPaper-Based Analytical
Paper-BasedAnalytical Devices
AnalyticalDevices
Devices and
and
and Traditional
Traditional Spectrophotometric
Traditional Spectrophotometric
Method
Method
Spectrophotometric Method
The feasibility
feasibility of
The feasibility
The the
of the
of reagent-sprayed
reagent-sprayed PADs
the reagent-sprayed PADs was
PADs was further
was further evaluated
further evaluated
evaluated by by comparing
by comparing
comparing the the AST
the AST
AST and and
and
ALT detection
ALT detection
ALT results
detection results obtained
results obtained
obtained for for real
for real human
real human
human bloodblood serum
blood serum samples
serum samples
samples with with those
with those obtained
those obtained using
using aaa
obtained using
traditional
traditional spectrophotometric
traditional spectrophotometric method.
spectrophotometric method. (Note
method. (Note that
(Note that the
that spectrophotometric
spectrophotometric assays
the spectrophotometric
the assays were
assays were conducted
were conducted
conducted
by a certified
by aa certified
by certified testtest laboratory
test laboratory
laboratory in in National
in National
National ChengCheng
Cheng KungKung University
Kung University Hospital,
University Hospital,
Hospital, Tainan,Tainan, Taiwan.)
Taiwan.) Table
Tainan, Taiwan.) Table 222
Table
presents
presents aaa qualitative
presents qualitative comparison
qualitative comparison
comparison of of the
of the two
the two methods
two methods
methods in in terms
interms
termsof of the
ofthe reagent/sample
thereagent/sample
reagent/sample volumes volumes
volumes and and
and
detection
detection time.
detection time.
time. As As shown,
As shown,
shown, the the spectrophotometric
thespectrophotometric
spectrophotometricassay assay requires
assayrequires
requires180 180
180µL μL R1
μLR1R1 andand
and 36
3636µLμL μL R2
R2R2 reagents
reagents
reagents to
to
to detect
detect eacheach
detect each
sample.sample.
sample. By
By contrast,
By contrast, contrast, the
the PAD-based
the PAD-based PAD-based colorimetric
colorimetric
colorimetric assay requiresassay
assay arequires
requires aa total
total reagent total reagent
reagent
volume of
volume
volume
just 2 µL.ofofInjust
just 22 μL.
μL. In
addition, In
the addition,
addition, the
the colorimetric
colorimetric colorimetric
assay requires assay
assay requires
requires
a sample aa sample
volume sample of 2 volume
volume of
of 22same
µL (i.e., the μL
μL (i.e.,
(i.e., the
the
as that
same
same
of as
the as that of the
that of whereas
reagent), reagent),
the reagent), whereas the spectrophotometric
whereas the spectrophotometric
the spectrophotometric method requires method
method
a sample requires
requires
volume a sample
a sample
of 7 µL.volume
volume
In other of
of
77 μL.
μL. In
words, other
Inthe
other words,
words, the
PAD-based the PAD-based
PAD-based
colorimetric colorimetric
colorimetric
method reduces method
method reduces
the reagentreduces andthe
the reagent
reagent
sample and
and sample
volumes sample volumes
volumes
by 99.07% and
by
by 99.07%
71.4%,99.07% and
and 71.4%,
71.4%, respectively.
respectively. respectively.
Figures 8
Figures 88 and
Figures and
and 999 compare
compare
compare the the detection
the detection results
detection results obtained
results obtained
obtained using using
using the the
the twotwo methods
two methods
methods for for the
for the AST
the AST
AST and and
and
ALT
ALT concentrations
concentrationsofof
ALT concentrations 20 20
of 20 human
human
human blood blood
blood samples,
samples,
samples, respectively.
respectively.
respectively. It is seen It is
is seen
It that seen that
that the
the correlation the correlation
correlation
coefficient
coefficient
coefficient
between thebetween
between the
the results
results obtained results
usingobtained
the twousing
obtained using
methods the two
twoamethods
the has value of Rhas
methods 2 =a
has value
a0.9319
valuefor of
oftheR 2 = 0.9319 for the
R2 AST
= 0.9319
sample forandthe
AST
RAST
2 sample
sample
= 0.9588 forand
and RR2 == sample.
the ALT 2 0.9588
0.9588 for
for the
the ALT
In other ALT
words,sample.
sample. In
In other
the practical other words,
words,ofthe
feasibility thethepractical
proposedfeasibility
practical feasibility
sprayed-reagentof
of the
the
proposed
proposed
PADs sprayed-reagent
sprayed-reagent
is confirmed. However, PADs
PADs is confirmed.
is confirmed.
it is observed However,
However,
that the detection it is
it is observed
observed
results obtained that the
that using detection
the detection results
results
the PAD-based
obtained
obtained using
colorimetric using the
the PAD-based
method PAD-based
are generally colorimetric
colorimetric
higher than method
method are
are generally
those obtained usinghigher
generally higher than
than those
those obtained
the spectrophotometer. obtainedTheusing
using
light
the
the spectrophotometer.
spectrophotometer.
intensity The
The light
of the shooting environment light intensity
intensity
was found of
of the
the
to beshooting
shooting environment
environment
a very influential factor,wasandfound
was found
affectedto be
to the aa very
be picture
very
influential
influential
colors. Hence,factor,
factor, and
and light
a fixed affected
affected
sourcethe picture
the environment colors.
picture colors. wasHence,
Hence,
needed a
a tofixed
fixed
avoid light
light source
anysource environment
environment
distortion in the image was
was
needed
needed to
analysis to avoid any distortion in the image
avoid any distortion in the image analysis results.
results. analysis results.

Figure
Figure 8.
8. (a) AST concentration detection results obtained using paper-based analytical device and
Figure 8. (a)
(a)AST
ASTconcentration
concentrationdetection results
detection obtained
results using
obtained paper-based
using analytical
paper-based device
analytical and
device
traditional
traditional spectrophotometric
spectrophotometric method
method from
from NCKU
NCKU hospital;
hospital; (b)
(b) Correlation
Correlation between
between
and traditional spectrophotometric method from NCKU hospital; (b) Correlation between
spectrophotometric results
spectrophotometric results for
results for AST
for AST concentration
AST concentration and
concentration and paper-based
and paper-based analytical
paper-basedanalytical device
analyticaldevice results.
deviceresults.
results.
spectrophotometric

77
Micromachines 2016, 7, 9 8 of 10
Micromachines 2016, 7, page–page

Figure 9.
Figure 9. (a)
(a)ALT
ALTconcentration
concentrationdetection results
detection obtained
results using
obtained paper-based
using analytical
paper-based device
analytical and
device
traditional spectrophotometric method from NCKU hospital; (b) Correlation
and traditional spectrophotometric method from NCKU hospital; (b) Correlation between between
spectrophotometric results
spectrophotometric results for
for ALT
ALT concentration
concentration and
andpaper-based
paper-basedanalytical
analyticaldevice
deviceresults.
results.

Table 2.
Table 2. Comparison
Comparison between
between colorimetric
colorimetric method
method and
and spectrophotometric
spectrophotometric method
method for
for detection
detection of
of
AST and
AST and ALT.
ALT.

Methods
Methods Roche
Roche Computing and and
Computing Microfluidics Lab
Microfluidics Lab
Detection method
Detection method Spectrophotometer
Spectrophotometer Paper-based devices
Paper-based devices
R1: 180 μL µL
R1: 180 R1 +R1R2:+ 2R2:
μL2 µL
Volume ofreagent
Volume of reagent
R2: 36
R2:μL
36 µL (R1:R2 = 5:1)
(R1:R2 = 5:1)
Volume ofsample
Volume of sample 7 μL7 µL 2 μL 2 µL
AST reaction time 10 min 6 min
AST reaction time 10 min 6 min
ALT reaction time 10 min 7 min
ALT reaction time 10 min 7 min

4. Conclusions
4. Conclusions
This study has presented the use of a modified inkjet printer to spray biochemical reagent solutions
This study has presented the use of a modified inkjet printer to spray biochemical reagent
onto paper-based analytical devices (PADs) fabricated using a wax-printing method. The feasibility of
solutions onto paper-based analytical devices (PADs) fabricated using a wax-printing method. The
the proposed approach has been demonstrated by performing the colorimetric concentration detection
feasibility of the proposed approach has been demonstrated by performing the colorimetric
of AST and ALT. The experimental results have shown that the optimal reaction time for AST detection
concentration detection of AST and ALT. The experimental results have shown that the optimal
is 6 min for AST concentrations in the range of 5.4 to 91.2 U/L (R2 = 0.9932). Similarly, the optimal
reaction time for AST detection is 6 min for AST concentrations in the range of 5.4 to 91.2 U/L
reaction time for ALT detection is 7 min for ALT concentrations in the range of 5.38 to 86.1 U/L
(R22 = 0.9932). Similarly, the optimal reaction time for ALT detection is 7 min for ALT concentrations
(R = 0.9944). In addition, it has been shown that a good agreement exists between the concentration
in the range of 5.38 to 86.1 U/L (R2 = 0.9944). In addition, it has been shown that a good agreement
detection results obtained using the sprayed-reagent PADs for real human blood samples and those
exists between the concentration detection results obtained using the sprayed-reagent PADs for real
obtained using a traditional spectrophotometer (i.e., correlation coefficients of R2 = 0.9319 for AST and
human blood samples and those obtained using a traditional spectrophotometer (i.e., correlation
R2 = 0.9588 for ALT). However, the concentration results obtained using the colorimetric approach are
coefficients of R2 = 0.9319 for AST and R2 = 0.9588 for ALT). However, the concentration results
slightly higher than those obtained from the spectrophotometric assay due to variations in the lighting
obtained using the colorimetric approach are slightly higher than those obtained from the
conditions during the detection process. Consequently, it is recommended that the PAD calibration
spectrophotometric assay due to variations in the lighting conditions during the detection process.
and assay processes should be performed using a fixed light source within a closed detection box.
Consequently, it is recommended that the PAD calibration and assay processes should be
In addition, it is noted that the reagents and samples should be stored in a chilled environment prior
performed using a fixed light source within a closed detection box. In addition, it is noted that the
to testing in order to prevent thermal deterioration under long-term room temperature conditions.
reagents and samples should be stored in a chilled environment prior to testing in order to prevent
In general, the results presented in this study confirm that the proposed inkjet reagent spraying
thermal deterioration under long-term room temperature conditions.
method provides a low-cost, versatile, and effective method for the mass-production of paper-based
In general, the results presented in this study confirm that the proposed inkjet reagent
analytical devices, and is thus ideally suited to the realization of point-of-care (POC) applications.
spraying method provides a low-cost, versatile, and effective method for the mass-production of
paper-based analytical
Supplementary devices,
Materials: The andare
following isavailable
thus ideally suited
online at to the realization of point-of-care
http://www.mdpi.com/2072-666X/7/1/9/S1.
(POC)S1:
Figure applications.
Adobe Photoshop 7.0 software analyzed RGB values as color intensity. Figure S2: (a) Change in mean
R + G + B values color intensity over time for AST; (b) Use the value 200 to subtract the detection values from
(a) to reverse the results. Figure S3: (a) The result of AST detection obtained by printer method; (b) The result of
Supplementary Materials: The following are available online at http://www.mdpi.com/2072-666X/7/1/9/S1.
AST detection obtained by pipette method. Figure S4: (a) The result of ALT detection obtained by printer method;
Figure
(b) The S1: Adobe
result Photoshop
of ALT detection 7.0 software
obtained analyzed
by pipette RGB values
method. as AST
Table S1: colordata.
intensity. Figure
Table S2: ALTS2: (a) Change in
data.
mean R + G + B values color intensity over time for AST; (b) Use the value 200 to subtract the detection values
from (a) to reverse the results. Figure S3: (a) The result of AST detection obtained by printer method; (b) The
result of AST detection obtained by pipette method. Figure S4: (a) The result of ALT detection obtained by

8
Micromachines 2016, 7, 9 9 of 10

Acknowledgments: The authors gratefully acknowledge the financial support provided to this study by the
Ministry of Science and Technology of Taiwan under grant number MOST-102-2221-E-006-104-MY3. The authors
are also indebted to Chiao-Hsiung Chuang and Chin-Chung Tseng of NCKU hospital for their assistance
throughout the course of this study.
Author Contributions: Hsiang-Li Wang and Chien-Hung Chu designed the experiment, prepared required
equipment and medicines, collected clinical serum of patients at NCKU hospital, and also completed all the goals
of experiment. Sing-Jyun Tsai and Ruey-Jen Yang supervised the work, analyzed data, and edited the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Mukai, M.; Ozasa, K.; Hayashi, K.; Kawai, K. Various S-GOT/S-GPT ratios in nonviral liver disorders and
related physical conditions and life-style. Dig. Dis. Sci. 2002, 47, 549–555. [CrossRef] [PubMed]
2. Mussoa, G.; Gambinob, R.; Cassaderb, M. Recent insights into hepatic lipid metabolism in non-alcoholic
fatty liver disease (NAFLD). Prog. Lipid Res. 2009, 48, 1–26. [CrossRef] [PubMed]
3. Huang, X.J.; Choi, Y.K.; Im, H.S.; Yarimaga, O.; Yoon, E.; Kim, H.S. Aspartate aminotransferase (AST/GOT)
and Alanine aminotransferase (ALT/GPT) detection techniques. Sensors 2006, 6, 756–782. [CrossRef]
4. Karmen, A. A note on the spectrometric assay of glutamic-oxalacetic transaminase in human blood serum.
J. Clin. Investig. 1955, 34, 131–133. [PubMed]
5. Itoh, H.; Srere, P.A. A new assay for glutamate-oxaloacetate transaminase. Anal. Biochem. 1970, 35, 405–410.
[CrossRef]
6. Janasek, D.; Spohn, U. Chemiluminometric flow injection analysis procedures for the enzymatic
determination of L-alanine, α-ketoglutarate and L-glutamate. Biosens. Bioelectron. 1999, 14, 123–129.
[CrossRef]
7. Babson, A.L.; Shapiro, P.O.; Williams, P.A.R.; Phillips, G.E. The use of a diazonium salt for the determination
of glutamic-oxalacetic transaminase in serum. Clin. Chim. Acta 1962, 7, 199–205. [CrossRef]
8. Lippi, U.; Guidi, G. A new colorimetric ultramicromethod for serum glutamicoxalacetic and glutamic-pyruvic
transaminase determination. Clin. Chim. Acta 1970, 28, 431–437. [CrossRef]
9. Canepari, S.; Carunchio, V.; Girelli, A.M.; Messina, A. Determination of aspartate aminotransferase activity
by high-performance liquid chromatography. J. Chromatogr. B 1994, 656, 191–195. [CrossRef]
10. Martinez, A.W.; Phillips, S.T.; Butte, M.J.; Whitesides, G.M. Patterned paper as a platform for inexpensive,
low-volume, portable bioassays. Angew. Chem. Int. Ed. 2007, 46, 1318–1320. [CrossRef] [PubMed]
11. Bruzewicz, D.A.; Reches, M.; Whitesides, G.M. Low-cost printing of poly(dimethylsiloxane) barriers to
define microchannels in paper. Anal. Chem. 2008, 80, 3387–3392. [CrossRef] [PubMed]
12. Pollock, N.R.; Rolland, J.P.; Whitesides, G.M. A paper-based multiplexed transaminase test for low-cost,
point-of-care liver function testing. Sci. Transl. Med. 2012, 4, 152ra129. [CrossRef] [PubMed]
13. Lee, W.G.; Kim, Y.G.; Chung, B.G.; Demirci, U.; Khademhosseini, A. Nano/microfluidics for diagnosis of
infectious diseases in developing countries. Adv. Drug Deliv. Rev. 2010, 62, 449–457. [CrossRef] [PubMed]
14. Martinez, A.W.; Phillips, S.T.; Whitesides, G.M. Diagnostics for the developing world: Microfluidic
paper-based analytical devices. Anal. Chem. 2010, 82, 3–10. [CrossRef] [PubMed]
15. Martinez, A.W.; Phillips, S.T.; Wiley, B.J.; Gupta, M.; Whitesides, G.M. FLASH: A rapid method for
prototyping paper-based microfluidic devices. Lab Chip 2008, 8, 2146–2150. [CrossRef] [PubMed]
16. Lu, Y.; Shi, W.; Jiang, L.; Qin, J.; Lin, B. Rapid prototyping of paper-based microfluidics with wax for low-cost,
portable bioassay. Electrophoresis 2009, 30, 1497–1500. [CrossRef] [PubMed]
17. Fenton, E.M.; Mascarenas, M.R.; López, G.P.; Sibbett, S.S. Multiplex lateral-flow test strips fabricated by
two-dimensional shaping. ACS Appl. Mater. Interfaces 2009, 1, 124–129. [CrossRef] [PubMed]
18. Nie, J.; Liang, Y.; Zhang, Y.; Le, S.; Lia, D.; Zhang, S. One-step patterning of hollow microstructures in paper
by laser cutting to create microfluidic analytical devices. Analyst 2013, 138, 671–676. [CrossRef] [PubMed]
19. Mu, X.; Zhang, L.; Zheng, Z. Multiplex microfluidic paper-based immunoassay for the diagnosis of hepatitis
C virus infection. Anal. Chem. 2014, 86, 5338–5344. [CrossRef] [PubMed]
20. Sun, M.; Johnson, A. Measurement of total antioxidant capacity in sub-µL blood samples using craft
paper-based analytical devices. RSC Adv. 2015, 5, 55633–55639. [CrossRef]
21. Carrilho, E.; Martinez, A.W.; Whitesides, G.M. Understanding wax printing: A simple micropatterning
process for paper-based microfluidics. Anal. Chem. 2009, 81, 7091–7095. [CrossRef] [PubMed]
Micromachines 2016, 7, 9 10 of 10

22. Lu, Y.; Shi, W.; Qin, J.; Lin, B. Fabrication and characterization of paper-based microfluidics prepared in
nitrocellulose membrane by wax printing. Anal. Chem. 2010, 82, 329–335. [CrossRef] [PubMed]
23. Zhong, Z.W.; Wang, Z.P.; Huang, G.X.D. Investigation of wax and paper materials for the fabrication of
paper-based microfluidic devices. Microsyst. Technol. 2012, 18, 649–659. [CrossRef]
24. Renault, C.; Koehne, J.; Ricco, A.J.; Crooks, R.M. Three-dimensional wax patterning of paper fluidic devices.
Langmuir 2014, 30, 7030–7036. [CrossRef] [PubMed]
25. Calvert, P. Inkjet printing for materials and devices. Chem. Mater. 2001, 13, 3299–3305. [CrossRef]
26. Abe, K.; Suzuki, K.; Citterio, D. Inkjet-printed microfluidic multianalyte chemical sensing paper. Anal. Chem.
2008, 80, 6928–6934. [CrossRef] [PubMed]
27. Su, S.; Ali, M.M.; Filipe, C.D.M.; Li, Y.; Pelton, R. Microgel-based inks for paper-suppor biosensing
applications. Biomacromolecules 2008, 9, 935–941. [CrossRef] [PubMed]
28. Yamada, K.; Henares, T.G.; Suzuki, K.; Citterio, D. Paper-based inkjet-printed microfluidic analytical devices.
Angew. Chem. Int. Ed. 2015, 54, 5294–5310. [CrossRef] [PubMed]
29. Liang, W.H.; Chu, C.H.; Yang, R.J. Bio-sample detection on paper-based devices with inkjet printer-sprayed
reagents. Talanta 2015, 145, 6–11. [CrossRef] [PubMed]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons by Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).