nature biomedical engineering
https://doi.org/10.1038/s41551-023-01059-5
Artic�e
A wireless patch for the monitoring of
C-reactive protein in sweat
In the format provided by the
authors and unedited
Content
s
Supplementary Note 1 |
Simulated CRP
-
dAb levels on the working electrode over time.
Supplementary Note 2 |
Real
-
time CRP sensor calibration during on
-
body studies
.
Supplementary Fig
.
1 |
Fabrication process of the microfluidic multimodal sensor patch.
Supplementary Fig
.
2 |
SEM images of the LEG electrodes
.
Supplementary Fig
.
3 |
The integrated wireless wearable electronics for CRP sensing.
Supplementary Fig
.
4 |
Surface functionaliza
tion process of the LEG
-
AuNPs working electrode of the CRP
sensor.
Supplementary Fig. 5 |
Influence of the self
-
assembled monolayer (SAM) modification on electrode
performance.
Supplementary Fig
.
6
|
Electrochemical performance of the LEG CRP sensors prepa
red by different
functionalization methods.
Supplementary Fig
.
7
|
Characterization of the CRP sensor functionalization using XPS.
Supplementary Fig
.
8
|
Electrochemical impedance spectroscopy (EIS) for the LEG
-
AuNPs CRP sensor
after each surface
modification step.
Supplementary Fig
.
9
|
Batch to batch variations in electrochemical performance of the LEG electrodes and
LEG
-
AuNPs electrodes.
Supplementary Fig
.
10
|
Conjugation process of the dAb
-
conjugated AuNPs complex for signal
amplification.
Sup
plementary Fig
.
11
|
Comparison of the electrochemical performances of redox probe conjugated dAb
and dAb
-
conjugated AuNPs.
Supplementary Fig
.
1
2
|
UV
-
Vis absorbance of the dAb
-
AuNP conjugate after each modification step.
Supplementary Fig. 13 |
The
reproducibility of the CRP sensor.
Supplementary Fig
.
1
4
|
Selectivity of the CRP sensor to potential interferences in sweat.
Supplementary Fig. 1
5
|
The influence of incubation pH, time, temperature,
sample volume,
and detection
ionic strength on the CRP
sensor responses.
Supplementary Fig. 16 |
SWV voltammograms of the CRP sensors in CRP solutions with varying Cl
-
concentrations.
Supplementary Fig
.
1
7
|
The stability of the CRP sensor.
Supplementary Fig
.
1
8
|
Representative mass spectra of CRP derived pe
ptides and chromatograms.
Supplementary Fig
.
1
9
|
CRP levels in serum, sweat, saliva, and urine.
Supplementary Fig
.
20
|
Box
-
and
-
whisker plot of sweat electrolyte and pH levels between iontophoresis
and exercise sweat.
Supplementary Fig
.
21
|
The
influence of electrolyte concentration and pH on the antigen capturing
measured with ELISA.
Supplementary Fig
.
22
|
Sweat rate of current and former smokers with and without COPD after 5 minutes
of iontophoretic sweat induction.
Supplementary Fig. 2
3
|
Inf
luence of the high flow rates on microfluidic automatic CRP sensing.
Supplementary Fig. 2
4
|
Influence of the interferent molecules on microfluidic automatic CRP sensing.
Supplementary Fig
.
2
5
|
Flow test characterization of the multimodal sensor patch in
response to different
initial pHs.
Supplementary Fig
.
2
6
|
Detailed circuit schematic of the InflaStat.
Supplementary Fig
.
2
7
|
LEG
-
based sensor calibration plots obtained with the InflaStat.
Supplementary Fig. 2
8
|
Performance of the CRP, pH, and temperat
ure sensors under mechanical
deformation
.
Supplementary Fig
.
29
|
On
-
body evaluation of the wearable sensor on healthy
never smokers
.
Supplementary Fig
.
3
0
|
On
-
body evaluation of the wearable sensor on healthy smokers.
Supplementary Fig
.
3
1
|
On
-
body evaluation of the wearable sensor on post
-
covid subjects.
Supplementary Fig. 3
2
|
Time
-
lapse images showing the automatic microfluidic reagent routing and
washing.
Supplementary Fig
.
3
3
|
Influence of solution pH in peak potential and current of
the redox molecule
thionine.
Supplementary Fig
.
3
4
|
Influence of the pH, ionic strength, and temperature on the CRP sensor reading.
Supplementary Fig
.
3
5
|
Box
-
and
-
whisker plot of CRP levels in sweat and serum samples from post
-
COVID
subjects with mild an
d moderate symptoms.
Supplementary Fig
.
3
6
|
Evaluation of system reproducibility via dynamic monitoring of sweat and blood
CRP.
Supplementary Fig
.
3
7
|
Evaluation of sweat CRP levels across anatomic locations.
Supplementary Table 1 |
Commercial CRP
kits and laboratory tests that can perform ng
-
level CRP
detection.
Supplementary Table 2 |
Commercial point
-
of
-
care CRP monitors.
Supplementary Table 3 |
Inclusion list of precursor ions used for targeted CRP protein.
Supplementary Table 4 |
Patient inform
ation for the COPD study.
Supplementary Table 5 |
Patient information for the heart failure study.
Supplementary Table 6 |
Patient information for the active infection study.
Supplementary Table 7 |
The currently reported microfluidic wearable sweat
sensors
.
Supplementary Table 8 |
Patient information for the post
-
COVID
-
19 infection study
.
Supplementary Note 1 | Simulated CRP
-
dAb levels on the working electrode over time.
As sweat samples containing CRP molecules enter the microfluidic patch, the detector antibodies deposited
in solid state are expe
cted to dissolve and diffuse within the detection chamber along the concentration
gradient. The collision between CRP molecules with antibodies will lead to the antigen
-
antibody binding
events along the microfluidic channels before they eventually reach th
e detection chamber. The introduction
of a serpentine microfluidic channel is also expected to facilitate the mixing and binding of the antigen
-
antibody complex.
Therefore, to visualize and estimate the time scale of the binding events at various location
s of the
microfluidic module, simulation of the CRP
-
antibody reversible binding reaction and the mass transport
process of reactant and product were conducted through finite element analysis (FEA) using the commercial
software COMSOL Multiphysics (see
Meth
ods
and
Fig. 4d,e
).
Based on the results illustrated in
Fig. 4d
, the binding and transport of CRP with detection antibodies can be
categorized into four stages. The heat maps represent the concentration of CRP
-
detection antibody complex
formed. In the reconstitution stage (I), detection antibodies diffuse along the co
ncentration gradient. Binding
of CRP starts to occur within the center of the reagent reservoir as indicated by the red color. As more sweat
containing CRP molecules enter the reagent reservoir, more antigen
-
antibody complexes are formed as
indicated by th
e larger area of red
-
color species. The antigen
-
antibody complex travels along the flow
direction to enter the detection chamber (circular chamber). After the serpentine mixing channels, antigen
-
antibody complex slowly distributes evenly across the detecti
on chamber, allowing binding with capture
antibodies immobilized at the bottom of the detection chamber to occur (II. Incubation).
After all the pre
-
deposited detection antibodies in the reagent reservoir are reconstituted, formed antigen
-
antibody complex
with sweat CRP or flushed into the detection reservoir, the concentration of detection
antibodies in the reagent reservoir is gradually depleted. The continuous flow of sweat into the microfluidic
module will no longer lead to the formation of more antibo
dy
-
antigen complexes as indicated by the blue
color in the reagent reservoir in step III (Refreshment). Hence, fresh sweat stream deplete of antigen
-
antibody complexes continues to enter the detection chamber and flush the unbound antibody
-
antigen
complexe
s in the chamber towards the outlet. Eventually, all unbound antibody
-
antigen complexes and
detection antibodies (which are labeled with electroactive molecules) will be refreshed out of the detection
chamber as shown in step IV. At this stage, detection i
s performed, and the electrochemical signal obtained
is specific and correlated to the antigen
-
antibody complexes bound on the working electrode surface since
the concentration of the complex in the detection chamber converges to zero (indicated by the blu
e color).
Supplementary Note 2
|
Real
-
time CRP sensor calibration during on
-
body studies
We investigated the influence of pH, electrolyte and temperature and found that all factors influence the
sensor readout of CRP based on
Supplementary Fig.
34
.
To account for the influences from binding environments, a multivariate model consisting of fou
r
independent variables: temperature, pH, electrolyte, CRP concentration ([CRP]) and a dependent variable:
peak current expressed in potential (mV) was constructed based on the following equation:
푝푒푎푘
푐푢푟푟푒푛푡
=
퐴
×
[
퐶푅푃
]
×
푝
퐻
!
×
[
푒푙푒푐푡푟표푙
푦푡푒
]
"
×
푡푒푚푝푒푟푎푡푢푟
푒
#
The coefficients were solved using non
-
linear least square fitting in Matlab and found to be:
A
=
-
0.5117;
m
= 0.6862;
n
= 0.1068;
j
=
-
0.6135
The model demonstrate
s
good accuracy in predicting signals measured by the sensors (
r
2
= 0.94). During
on
-
body operation, readings from the pH, temperature, electrolyte, and CRP sensors can thus be used to
real
-
time back
-
calculate the actual concentration of CRP based on the fi
tted model.