Levofloxacin Loaded Microneedles Produced Using 3D Printed Molds For Noso Infection

RESEARCH ARTICLE
www.advtherap.com
Levofloxacin-Loaded Microneedles Produced Using

3D-Printed Molds for Klebsiella Pneumoniae Biofilm Control
K. B. Vinayakumar,* Maria Daniela Silva, Artur Martins, Stephen Mundy,
Pedro González-Losada, and Sanna Sillankorva*
various methods are being adapted to

Additive manufacturing advancements contribute considerably to several deliver drugs across the skin layers. A
fields and its use in the medical field is gaining attention due to its easily well-established method of drug delivery
customizable option (patient-specific), low cost, and fast turnout time in through the skin involves hypodermic nee-
developing drug delivery and diagnostic tools. Herein, the fabrication of a dles. However, this conventional drug ad-
ministration method does not deliver the
microneedle (MN) platform is reported using a stereolithography 3D printer,
drug effectively and accurately within the
varying the 3D printing angle and aspect ratio (2:1, 3:1, and 4:1). The optimal therapeutic range or window.[1] Hence,
printing angle is 30°, resulting in needle tip and base diameters of ≈50 and in the 1970s, a transdermal route was
≈330 μm and heights of ≈550/850/1180 μm. Polyvinyl alcohol (PVA) MNs proposed to deliver the drug painlessly,
produced with varying levofloxacin concentrations show variability of ≈4% in more effectively, and accurately. Transder-
mal drug delivery technology has many
tip and 3% base diameters and 15% in height compared to the 3D-printed
advantages, such as targeting a specific
MNs. Geometry B is used to produce levofloxacin-loaded PVA MNs and skin area, avoiding the stomach environ-
tested against Klebsiella pneumoniae colony biofilms. Levofloxacin is released ment, reducing medical costs and med-
gradually, as assessed by spectrofluorimetry. The minimum inhibitory ical wastes, allowing a dose reduction,
concentration of levofloxacin against the K. pneumoniae clinical isolate is and more precise control over drug vol-
4 μg mL−1 but this concentration is insufficient to cause any effect on K. ume that can be achieved.[2] However, the
main drawback of using a traditional trans-
pneumoniae biofilms. Only concentrations ≥32 μg mL−1 are statistically
dermal patch is that the skin barrier is
different compared to the unloaded MNs. 3D printing is an attractive solution mostly impermeable, limiting the transport
to produce molds for fabricating biopolymeric MNs for topical drug delivery. of large/hydrophilic molecules. Therefore,
many alternative techniques have been pro-
posed to improve transdermal drug deliv-
ery, such as jet injectors, iontophoresis,
1. Introduction sonophoresis, chemical penetration enhancer, skin ablation, and
The skin has been a lucrative target for drug delivery and diagno- microneedle (MN).[3,4] Among these approaches, the micronee-
sis in medicine. Since the skin is the largest organ in the body, dle approach is considered one of the best methods to deliver the
drug with fewer side effects.[3–6]
Since MNs are short, they do not reach the nerve-rich region
K. B. Vinayakumar, A. Martins, S. Mundy, P. González-Losada, resulting in less or no pain during insertion.[4,7] Also, MNs will
S. Sillankorva
INL—International Iberian Nanotechnology Laboratory
help deliver macromolecules (e.g., insulin and vaccines), offer-
Av. Mestre José Veiga, Braga 4715-330, Portugal ing a patient-friendly drug administration. The primary use of
E-mail: [email protected]; [email protected] microneedling is to create a pathway to an object by physically
M. D. Silva scratching or rupturing the top layer. In most applications, this
Centre of Biological Engineering barrier is the skin. The first concept of the microneedle array for
University of Minho the drug delivery application was reported by Gerstel and Place in
Campus de Gualtar, Braga 4710-057, Portugal
1971.[8] This patent presents a concept of improved transdermal
M. D. Silva
LABBELS—Associate Laboratory drug delivery using solid and hollow microneedle arrays.[8] Al-
Braga/Guimarães Portugal though the concept of microneedle was presented earlier, it was
not fabricated until the 1990s, when the microfabrication revolu-
The ORCID identification number(s) for the author(s) of this article tion took place. Henry et al. reported the first 150 μm long mi-
can be found under https://doi.org/10.1002/adtp.202200320
croneedle array for the drug delivery application, showing a four
© 2023 The Authors. Advanced Therapeutics published by Wiley-VCH
orders magnitude increase in drug permeability through the skin
GmbH. This is an open access article under the terms of the Creative
Commons Attribution-NonCommercial License, which permits use, compared to the traditional transdermal patches.[9] The strength
distribution and reproduction in any medium, provided the original work of MNs as technological platforms has been demonstrated for dif-
is properly cited and is not used for commercial purposes. ferent applications, such as electrical signal recording from the
DOI: 10.1002/adtp.202200320 visual cortex of the brain,[10] skin impedance measurement for
Adv. Therap. 2023, 6, 2200320 2200320 (1 of 10) © 2023 The Authors. Advanced Therapeutics published by Wiley-VCH GmbH
23663987, 2023, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adtp.202200320, Wiley Online Library on [21/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advtherap.com
cancer,[11] insulin delivery,[12] blood collection,[13] among others the most common pathogens in nosocomial infections. In recent
(see ref. [14] for a more detailed summary of the varied uses). years, this bacteria has been associated with chronic ulcers and
Different technologies such as deep reactive-ion etching burn wound infections, besides severe pulmonary, and urinary
(DRIE) silicon, laser machining metal, imprinting polymer, tract infections.[34,35] The presence of this pathogen in wounds is
3D printing polymer molds, and extruding polymer mold- problematic since it is often associated with biofilms that are in-
ing have been previously reported to fabricate microneedle creasingly resistant to antibiotic treatment.[36] Nonetheless, the
arrays.[4,13,15–19] Recent advancement in 3D-printing technology delivery using MNs can prompt a drug distribution to layers,
is increasingly accepted for developing MNs, microsensors, and often not accessible due to the biofilm matrix surrounding the
microactuators.[20–23] Due to the skin thickness varying with eth- bacterial cells, and the mechanical effect can already cause some
nicity, gender, and age of people, the customizable MNs so- damage in the 3D biofilm structures. Therefore, MNs with dif-
lution is crucial for applications related to drug delivery, anti- ferent levofloxacin concentrations were assessed for their effec-
body delivery, and interstitial fluid sampling, among others. The tiveness in reducing the number of viable K. pneumoniae cells in
3D printed molding approach will provide a customizable solu- biofilms.
tion to produce a microneedle of interest in less time without
the need for expensive equipment and trained microfabrication 2. Results and Discussion
professionals.[24,25] The small cost, customizable approach, and
easy-use nature of 3D printing were recently adopted to develop 2.1. Microneedle Design and 3D Printing
a microneedle array for drug delivery applications.[26,27] Krieger
et al. reported using a stereolithography (SLA) 3D printer to fab- The MNs designed in the paper have a tip diameter of 50 μm, base
ricate high aspect ratio microneedle masters and reproducing diameter of 300 μm, and height of 600/900/1200 μm. A study by
them in female silicone molds to fabricate dissolvable polymer Gill et al. demonstrated that a microneedle with a tip radius of
MNs.[27] The developed approach showed a promising route to 50 μm could enable easy and pain-free insertion into the skin,[37]
develop an MN array (MNA) with a tip of ≈20–30 μm. However, and the microneedle height will depend on the region at which
detailed studies on printing angle, MNs with different aspect ra- the drug molecule that needs to be delivered under the skin. The
tios, and geometry variations concerning 3D-printed MNs and typical needle height required to rupture the top layer of the skin
biopolymer needles have not been studied in detail. Further, the is ≈150 μm.[9] However, different aspect ratios (2:1, 3:1, and 4:1)
flexibility of the 3D printer for the specific application will allow of MNs were demonstrated to address the 3D-printed micronee-
the increase of the number of MNs per chip and modify the size, dle scope in delivering the drug in different skin layers for poten-
shape, and geometry of each microneedle just by modifying the tial application in drawing interstitial fluids.[24–27]
design in the software. The Z resolution influence on stepping, i.e., the smoothness
MNs can be made of solid (e.g., metal, silicon, glass, and ce- of the microneedle, was studied by Johnson et al.[26] A higher
ramics) or polymeric materials. The type of polymer used for printing resolution takes a long time to print but should create
the fabrication of MNs will have an impact on the character- a smoother surface on features. The smoothness of the needle
istics of the MNs produced. For instance, the polymers cho- is crucial to reduce the insertion force and ensure the easy ex-
sen can create dissolving or hydrogel-forming MNs. Dissolv- traction of the cured part from the master mold. Another param-
ing MNs are suitable for the delivery of low molecular weight eter optimized is the distance between adjacent needles. When
drugs, enzymes, vaccines, peptides (reviewed in ref. [28]), and the distance between the needles is too small, the resin between
even bacteriophages.[29] These are entirely or almost wholly dis- them can occasionally and randomly be fused. This consequently
solved once inserted into the skin. The materials used for this affects the microneedle outer wall angle out in a concave shape.
are water-soluble ones, including, for instance, maltose, dextran, The orientation of the printing angle showed a considerable in-
and hyaluronic acid.[30] Hydrogel-forming MNs are swellable fluence on the microneedle geometry.
and are suitable, for instance, for the extraction of intersti- Along with the variation in needle geometry, needles were
tial fluid and can be fabricated with diverse polymers, such tilted or twisted with printing angle orientation. Figure 1 shows
as poly(methyl vinyl ether-alt-maleic anhydride) (PMVE/MA), optical and scanning electron microscope (SEM) images of the
MeHA, silk, among others (reviewed in ref. [31]). In recent years, MNs printed considering the parameters detailed in the Experi-
polymeric MNs have received more attention in the biomedical mental Section.
area than inorganic and metal MNs, due to cost, biodegradability, The variation in tip, base, and height of printed MNs with dif-
biocompatibility, low toxicity, and strength.[32,33] ferent printing angle orientations and the designed computer-
This paper demonstrates the stereolithography-based 3D- aided-design (CAD) model geometries for each parameter were
printed approach to producing a microneedle array for drug (lev- analyzed (Figure 2).
ofloxacin, LVX) delivery for medical applications. The design and MNs printed with 0° showed a more repeatable result concern-
fabrication of the different printing angles have been evaluated to ing the needle tip, base, and height with the designed CAD model
find the optimum printing conditions. An optimized printing an- geometries. However, the drawback observed in the 0° printing
gle was chosen and used to produce arrays of the 3D-printed MNs is tilting the MNs toward one direction. The tilting of a tip can
with different aspect ratios. Printed MNs were subsequently used be due to the pulling force generated during the resin pulling all
to produce polydimethylsiloxane (PDMS) molds and biopolymer along the printing process. Printing with 30° orientation showed
microneedle arrays loaded with levofloxacin. In addition, MNs a promising result with the reduction in tilting of the needles.
with different base geometries were produced to find the robust- However, the obtained needle geometry is not closely related
ness of the 3D printing system. Klebsiella pneumoniae is one of to the 0° printed needles. The reduction in the tip diameter is
Figure 1. Optical and SEM microscopy of different MNs. 1) Optical microscopy of the different MNs. The upper figures are enlarged MNs from the
strip with three MNs printed on the strip edge taken using a 5× objective. 2) SEM microscopy of arrays with a height of 600 μm and printed at different
angles. 3) MNs with heights of 900 and 1200 μm printed at 0° and 30° angles.
MNs arrays. The 3D-printed microneedle arrays fabricated were

used to produce PDMS molds, which were used to produce
biopolymer MNs.
The 3D printed microneedle arrays using the 30° orientation
were observed by SEM (Figure 3(1–3)).
The SEM images show that >95% of the 3D printed MNs
were intact (few broken needle tips were observed in the high
aspect ratio MNs). We believe the broken needles were primarily
related to the handling of the sample. It has been validated that
an optimized SLA recipe can generate a robust microneedle
array with a large variety of geometries.
The traditional approach to producing biopolymer-based MNs
still depends on cleanroom-processed silicon molds. One of the
main drawbacks of silicon-processed mold is that it is difficult
to achieve 3D features on the molds. However, the capability
of achieving complex features/geometries, fast turnaround
Figure 2. Microneedle angle orientations used during printing and the dif-
time, and low-cost production have helped the adoption of 3D
ferent geometries designed using CAD (tip—horizontal black, base—red,
and height—blue lines) and the 3D printed MNs (columns using the same printing in different application sectors, such as surgical tools,
colors). medical applications, organ-on-chip, tissue engineering, and
patient-specific implants.[38] On the scalable fabrication side,
SLA 3D printers can help to produce robust and custom molds
interesting for the microneedle application. The reduction in using the wide variety of material libraries available from the
microneedle tip diameter can be due to the printing angle, which supplier (https://formlabs.com/materials/engineering/). These
needs to be studied in detail. However, the observed reduction in materials withstand high temperatures and pressure making
tip diameter is relatively consistent. Other printed orientations them suitable for injection molding of the polymer microneedle
(45°, 60°, and 90°) showed promising repeatable results, but the array.[39,40] Table 1 summarizes the literature on 3D-printed
needle tip diameter is greater than the 30° orientation printed microneedle molds used for drug delivery applications.
needles. Along with obtaining similar results concerning needle ge-
After understating the variability of each parameter during ometry, the effect of printing angle on the needle tip tilting is
different printing orientations, 30° was used to print 11 × 11 studied in the paper. Furthermore, MNs with different base
Figure 3. Microneedle arrays. 1) 3D printed microneedle mold with 2:1 aspect ratio, 2) 3D printed microneedle mold with 3:1 aspect ratio, 3) 3D printed
microneedle mold with 4:1 aspect ratio, 4) PVA microneedle using mold-1, 5) PVA microneedle using mold-2, and 6) PVA microneedle using mold-3.
Table 1. 3D printed microneedle molds produced using different methods and their geometries.
Ref. Method Height [μm] Base [μm] Tip [μm]

[41] Continued liquid interphase production (CLIP) 1000 333
[42] Direct laser printing (DLP) 700 250
[27] SLA 3000/1000/600 40–80
[26] SLA (with 25 μm layer thickness) 707 437 40
This work SLA 1180/850/550 330 40–50
geometry (square, circle, triangle, pentagon, and hexagon) (cell viability >90%), being minimally toxic to these cells.[43] In
have also been successfully printed with good repeatability another study, fibroblasts and keratinocytes exposed for 24 h
(Figure 4). to highly porous membranes of N, O-carboxymethyl chitosan
Achieving silicon micromachining with such diverse geome- (CMCh) and PVA resulted in a 80–100% cell viability for both
tries is always challenging and complicated. Hence, this result cell lines, showing no significant cytotoxic effect. In addition,
validates the robustness and ease of use nature of the 3D printer PVA has low protein adsorption and high water solubility and
in micromolding applications compared to traditional silicon chemical resistance characteristics. Moreover, PVA is an excip-
processing. ient available in pharmaceutical-quality raw material in quanti-
ties required for large-scale commercial use and is widely used
in wound dressings due to its moisturizing effect and ability to
2.2. Biopolymer Microneedle Fabrication absorb wound exudate.[45,46]
Figure 5 summarizes the geometrical variation between 3D-
Three different aspect ratio MNs were 3D-printed and used to printed MNs with biopolymer MNs, using the MNs with a height
produce PDMS molds. The produced PDMS molds were used of 900 μm.
to fabricate biopolymer MNs. Polyvinyl alcohol (PVA) MNs, fab- The tip and base diameter of the biopolymer MNs are consis-
ricated using different aspect ratios, were observed by SEM (Fig- tent with the 3D-printed microneedle, with a variability of less
ure 3(4–6)). PVA is a synthetic polymer approved by the Food and than ≈2%. However, the height of the microneedle showed a
Drug Administration (FDA) and commonly used in medical de- considerable variation between 3D-printed and biopolymer MNs.
vices because of its biocompatibility, nontoxicity, and noncarcino- The significant variation is seen in the microneedle with a 2:1 as-
genicity due to its inertness and stability. For instance, Bushan pect ratio, where the reduction in biopolymer microneedle height
et al. tested the cytotoxicity of different wearable wound sensor is ≈32% compared to the 3D printed microneedle. In the 3:1
constituents and found that their polyvinyl alcohol-based poly- and 4:1 MN aspect ratios, a reduction of ≈11% in the biopoly-
mer (PVA-SbQ) had excellent biocompatible with keratinocytes mer needle height was seen compared to the 3D-printed MNs.
Figure 4. 3D-printed microneedle molds with different base geometries. Top) Optical microscope images (5× objective). Bottom) SEM images of the
array and an enlarged microneedle.
as this gram-negative species is frequently isolated from infected

wounds, including chronic wounds. In addition, the antibiotic
LVX was chosen since it can be used for treatment and can be
detected by fluorescence spectroscopy, which is advantageous for
its rapid quantification.
2.3.1. Minimum Inhibitory Concentration (MIC)
The MIC of LVX against K. pneumoniae PC28 planktonic cells

was 4 μg mL−1 (Figure 6). According to the European Committee
on Antimicrobial Susceptibility Testing (EUCAST) clinical break-
points, K. pneumoniae is resistant to LVX at dosings higher than
1 μg mL−1 .
Figure 5. Printed and biopolymer microneedle geometries for different as-

pect ratio MNs. 2.3.2. Activity of Levofloxacin-Loaded MNs against K. Pneumoniae
Biofilms
This difference in height phenomena of the polymer MNs versus
the master mold was observed already with other polymers (algi- K. pneumoniae biofilms were formed for 24 h, treated with LVX-
nate, carboxymethyl cellulose, and hyaluronic acid), which varies loaded MNs for an additional 24 h, and quantified (Figure 7a).
with the concentration of polymer used.[47] For instance, the mas- Although different geometries were printed in this work, for the
ter mold was 1200 μm in height, and the MNs produced using release and antimicrobial efficacy studies, only one of these ge-
the PDMS replica varied in height from 971 μm [carboxymethyl ometries was tested (B geometry, see Table 2). Furthermore, al-
cellulose 5% (w/w)] up to 1077 μm [hyaluronic acid 1% (w/w)], though in planktonic cells, the MIC concentration obtained was
corresponding to reductions ≈19% to ≈10%. Furthermore, in- 4 μg mL−1 , it is well known that antibiotic concentrations needed
creasing the polymer concentration also caused an increase in to cause a reduction of biofilm cells are generally higher. Within
the height of hyaluronic acid MNs [1012 μm with 1% (w/w) to biofilms, their self-produced extracellular polymeric matrix pro-
1077 μm with 5%] but decreased the height of carboxymethyl cel- tects cells from antibiotic action. Moreover, the lower cell growth
lulose MNs [1086 μm with 1% (w/w) to 971 μm with 5%]. The rate and the presence of metabolically inactive cells also con-
difference in microneedle height can be controlled by optimizing tribute to the increased antibiotic tolerance of biofilms compared
the PDMS hardness, PDMS curing temperature, controlling the to planktonic cells.[48] Therefore, in this study, a minimum con-
vacuum processing during PVA needle development, and using centration equal to the MIC (4 μg mL−1 ) and a maximum concen-
a more suitable polymer selection and its concentration. tration of 128 μg mL−1 (32× MIC) was loaded onto the MNs.
Levofloxacin is only available in oral and intravenous formu-
lations, and its administration involves daily dosages of 250–
2.3. Production and Efficacy of the LVX Loaded MNs 750 mg.[49,50] The disadvantages of systemic administration of
antibiotics are well-known, including several side-effects (par-
Envisioning using these MNs in biofilm-related infections, K. ticularly in the gastrointestinal tract), impact on commensal
pneumoniae was selected as the bacterial pathogen for the studies flora, low bioavailability at the infected region, and increase of
Figure 6. Determination of the minimum inhibitory concentration of levofloxacin on K. pneumoniae PC28. Image acquired using Chemi XT4.
Figure 7. Levofloxacin-loaded MNs. a) antibiofilm effect measured as the logarithmic number of K. pneumoniae PC28 viable cells from 24 h biofilms
after 24 h of treatment with levofloxacin-loaded MNs. Data are shown as mean ± SD for n = 3. Differences were considered statistically significant if
P ≤ 0.05 (*). b) Efficacy of the levofloxacin-loaded MNs on colony biofilms.
Table 2. Geometries of the 3D-printed MNs. The viable cells reached a concentration of 9.76 log10
CFU mL−1 (Figure 7a). As expected, LVX-loaded MNs signifi-
Label Diameter Height Ratio Diameter Pitcha) Printing orientation cantly reduced K. pneumoniae biofilm cells only at LVX concen-
[μm] [μm] [°] trations ≥32 μg mL−1 (P ≤ 0.05) compared to the unloaded MNs.
A 300 600 2:1 50 90 0 Therefore, a concentration eight times higher than the MIC was
B 30 necessary to produce a noticeable effect on biofilms. From this
C 45
concentration forward, all viable cell counts were statistically dif-
ferent (P ≤ 0.05) not only from the control (unloaded MNs) but
D 60
also between each concentration. Nevertheless, biofilm eradica-
E 90
tion was not achieved.
F 900 3:1 0
The delivery of LVX using MNs was previously studied to con-
K 30 trol biofilms formed by other bacterial species. Polydopamine
G 1200 4:1 0 nanoparticles containing levofloxacin and 𝛼-amylase were incor-
L 30 porated into MNs and used to treat Staphylococcus aureus and
a)
Pitch—The distance between the needles.
Pseudomonas aeruginosa biofilms. A residual biofilm biomass of
only 12.63% ± 1.86% was obtained after treatment (vs 100% of
the untreated control). The antibiofilm effect was also evaluated
antimicrobial resistance.[51] The topical application of antibiotics, in vivo in a rat model of infected full-thickness skin defect. The
for instance, using MNs, minimizes these effects. Moreover, MNs incorporating polydopamine nanoparticles loaded with lev-
MNs are an attractive choice for biofilm infections since they can ofloxacin with and without 𝛼-amylase reduced the bacterial levels.
penetrate the extracellular polymeric matrix barrier and deliver Photothermal therapy further reduced the number of bacteria,
the loaded antimicrobial into the biofilm interior.[29,52,53] with almost eradication ability.[53]
Figure 8. A–F) MN force-displacement curves and G,H) images of MNs subjected to the test. A) Arrays B, K, and L; B) triangular geometry; C) square
geometry; D) pentagon geometry; E) hexagon geometry, and F) circle geometry.
There are also major visual differences between the unloaded 2.3.3. Release of Levofloxacin from the PVA MNs
and LVX-loaded MNs on colony biofilms formed on the Petri
dishes (Figure 7b). As visible, the biofilms where unloaded MNs The release of LVX from the microneedle arrays at different con-
were used (control) spread throughout the agar layers, occupy- centrations was assessed for 60 min (Figure 9).
ing a large area, while in the LVX-treated colony biofilms, the The release started immediately, and after ≈10 min, the curves
biomass is restricted to the area where the polycarbonate mem- reached a plateau, with approximately the total concentration
brane was placed. Only the viable cells in the polycarbonate mem- (100%) loaded being released. This is because the polyvinyl al-
brane were counted for the viable cell counts since the area cov- cohol used is known to dissolve quickly (7–11 min).[29]
ered by the biomass on the Petri dishes using different antibiotic The MNs can, therefore, deliver a high amount of antibiotic
concentrations was highly dependent on the LVX concentration quickly, resulting in higher bioavailability and faster onset of ac-
used. tion than, for instance, topical creams and transdermal patches.
The mechanical behavior of the PVA MN arrays varying in It is reported that only 10–20% of drugs applied using topical
length and geometry was investigated (Figure 8). creams can permeate across the skin. In the case of transdermal
There were no significant differences between the PVA MN patches, the drug also still has to permeate the stratum corneum.
arrays varying in height and geometry. Furthermore, the curves Using hypodermic needles, the drug is delivered directly into the
did not show a transition point which could indicate buck- dermis, providing high bioavailability and fast action like MNs,
ling failure. This is in agreement with other studies reported but it is very painful, and self-administration is not possible.[57]
in the literature.[54–56] The only exception is the geometry de- Also, as stated before, MNs can penetrate the extracellular poly-
picted in Figure 8F, which showed a curve with a changing meric matrix of biofilms, which does not happen with topical
slope before fracture force was reached, most likely due to bend- creams and transdermal patches.
ing. There were, however, slight differences in the mechani-
cal behavior over the displacement ranges of PVA arrays vary- 3. Conclusion
ing the geometry (Figure 8B–F). Since all were fabricated with
PVA, the change is attributed merely to the geometry of the The paper reports desktop SLA printer optimization to print high
MNs. aspect ratio microneedle arrays for biomedical applications. In
ter printing, the MNA was washed in clean isopropyl alcohol (IPA) bath for
10 min. The bath used a magnetically coupled impeller to agitate the IPA
to remove the nonsolidified resin effectively. Later, the MNA was dried at
room temperature (RT) for 30 min and cured by UV (405 nm) for 60 min at
60 °C. Printing angles of 0°, 30°, 45°, 60°, and 90° (Figure 10) were tested
to study the microneedle variations in height, length, and base diameter
compared to the designed CAD model.
The CAD model of the microneedle in the study was designed with a
tip diameter of 50 μm, base diameter of 300 μm, and height of 600, 900,
or 1200 μm to obtain an aspect ratio of 2:1, 3:1, and 4:1 (Table 2).
After optimizing the printing angle, using just three printed MNs
printed at the edge, a master MNA was created, with which a female PDMS
mold (see Section 2.2) and the biopolymeric MNA (see Section 2.3) were
developed (Figure 11).
PDMS Mold Fabrication: The 3D-printed MNA master mold was used
to produce a PDMS replica. First, the mold was placed into a desicca-
tor with a drop of trichloro (1,1,2,2-perfluorooctyl) silane (from Sigma-
Aldrich) and placed under vacuum conditions, causing the silane to evap-
orate and form a layer on the surface of the mold to prevent any PDMS
adhesion. The PDMS (Sylgard 184 from Dow Corning) was prepared by
mixing elastomer and curing agent (10:1 ratio), followed by 3 min of son-
ication to mix both components homogeneously. Then, the PDMS was
Figure 9. Percentage of levofloxacin released from the MNs loaded with
placed for 15 min in the desiccator under vacuum conditions to remove all
different levofloxacin concentrations (4–128 μg mL−1 ). Data are shown as
air bubbles generated during the sonication. Next, the PDMS was poured
mean ± SD for n = 3. Differences were considered statistically significant
onto the MNA master, and a vacuum was applied for 30 min to remove
if P ≤ 0.05 (*).
any remaining bubbles. Finally, the PDMS mold was cured overnight in an
oven at 70 °C. The resulting PDMS replica was easily removed from the
master mold without causing damage to the 3D-printed gray resin MNA.
addition, the effect of microneedle geometry due to printing an- Biopolymer Microneedle Fabrication Process: PVA (Mowiol 4:88, Sigma-
gle orientation is reported. Fabricated different aspect ratios (2:1, Aldrich, MO, USA) was used to fabricate MNs with and without LVX using
3:1, 4:1) MNs were used to produce PDMS molds and subsequent a molding technique. PVA [10% w/v] was hydrated in sterile water for 4 h
biopolymer (PVA) microneedle arrays using the PDMS molds. at RT under agitation (350 rpm). After, the temperature was increased to
The fabricated biopolymer MNs with different aspect ratios were 85 °C and, under agitation (350 rpm), PVA was allowed to dissolve com-
compared with the 3D-printed MNs, and although they were pletely for ≈3 h. The solution was allowed to cool, and after, LVX in water
was mixed with the solution to have final concentrations ranging from 4 to
closely related, the variation in microneedle height needs to be 128 μg mL−1 . A volume of 150 μL of PVA was placed in the PDMS molds,
studied by optimizing the PDMS molding process and biopoly- and a vacuum was applied (Agilent Technologies IDP3, Massachusetts,
mer microneedle fabrication process. The dissolving polymeric USA). Molds were centrifuged at 1000 rpm (RT, 10 min, Universal 320,
MNs had an effect in locally reducing the concentration of viable Hettich GmbH & Co. KG., Tuttlingen, Germany) to fill the microneedle
cells of K. pneumoniae from biofilms. Antimicrobial-loaded MNs cavities and remove air bubbles completely. The polyvinyl MNs were dried
can have benefits when dealing with hard-to-remove biofilms for three h at 40 °C in a ventilated oven (Termaks B8054, Bergen, Nor-
way). MNs in their respective PDMS molds were stored in vacuum-sealed
from, for instance, chronic and infected wounds and should be
bags at 4 ˚C until their manual removal from the PDMS molds for further
explored more as possible alternatives to systemic drug delivery. use. To perform antibiofilm experiments, MNA were fabricated using filter
sterilized PVA solution containing different LVX concentrations that were
poured to UV sterilized (15 min UV) PDMS molds. Vacuum filling and
drying were performed in closed Petri dishes.
4. Experimental Section Compression Experiments: Compression experiments were performed
The 3D Microneedle Printing Process: Microneedle master molds, vary- using a Shimadzu AGX-10 kN Texture Analyzer (Shimadzu, Japan)
ing in geometry, were fabricated using a Formlabs Form2 desktop 3D SLA equipped with a 500 N load cell and a 10 mm (diameter) cylindrical probe.
printer. A standard gray resin (RS-F2-GPGR-04, Formlabs) was used, and At least three replicates of each sample were tested. The tested conditions
during the SLA 3D printing, the liquid resin was cured by a highly accu- involved a compression speed of 0.01 mm s−1 (equivalent probe pulling
rate laser to form each layer achieving much finer details and repeatedly speed) with a complete travel of up to 1.2 mm maximum.
achieving high-quality results. This is particularly important for printing Bacteria and Growth Conditions: The K. pneumoniae PC28 (clinical iso-
fine intricate geometry with smooth surfaces such as MNs. Although SLA late, Hospital de Braga) used in this study was previously described to be a
printers can usually achieve significantly smaller layer thickness (Z-axis), K. pneumoniae carbapenemase.[58] The strain was grown at 37 ˚C in Tryptic
the reason for improved print quality compared to fused deposition model Soy Broth or Tryptic Soy Agar (TSA).
(FDM) printers lies in their much higher XY resolution. To ensure optimal MIC: The MIC of LVX (Alfa Aesar, MA, USA) for K. pneumoniae PC28
printing conditions, the Form2 SLA printer was inspected and checked for was determined using the broth microdilution method according to the
potential issues that might affect the print quality before printing. The op- guidelines of the Clinical and Laboratory Standards Institute. The 96-well
tical glass window was cleaned using a dry lint-free wipe (Photographic plate was imaged using a high-performance imaging apparatus (Chemi
Emulsion Cleaner (PEC) Pad), and the resin was filtered using a 190 μm XT4, GBOX-CHEMI-XT4-E, AlphaMetrix Biotech, Rödermark, Germany),
mesh oil. coupled with a 4.2 megapixel (MP) imaging 16-bit charge-coupled device
The gray resin MN fabrication workflow consisted of three steps: de- (CCD) camera.
sign, printing, and postprocessing. First, a 3D CAD model of the MNA was Antibiofilm Activity of LVX-Loaded PVA MNs: The activity of LVX-loaded
created using Autodesk Inventor 2014 CAD software. Next, a .stl file was MNs was evaluated against 24 h biofilms of K. pneumoniae PC28. Biofilms
exported from Inventor and imported into the proprietary SLA software, were produced according to the colony biofilm procedure,[29] with some
Preform. Preform enabled the user to define the printing parameters. Af- modifications. Briefly, overnight cultures of K. pneumoniae were diluted to
Figure 10. A) Printing angles used in microneedle fabrication steps and B) a schematic representation of where the height, tip, and base dimensions
were determined.
Figure 11. Biopolymer microneedle array fabrication steps.
an approximate concentration of 1 × 108 CFU mL−1 . Sterile polycarbonate Tukey’s multiple comparisons test through GraphPad Prism 7 software.
membranes (Whatman Nuclepore Track-etched, Maidstone, UK, 25 mm Differences were considered statistically different if P ≤ 0.05 (95% confi-
diameter, 0.1 μm pore size) were cut to the size of the MNA and placed dence interval).
on TSA plates with the shiny side facing up and inoculated with 50 μL of
this culture. The polycarbonate membranes were incubated at 37 ˚C un-
der static conditions for 24 h and transferred to fresh TSA plates. Next, Acknowledgements
unloaded and LVX-loaded MNs were placed on top of the 24 h-biofilms
formed on the polycarbonate membranes. After, the plates were incubated S.S. acknowledges funding by FCT through the individual scientific em-
at 37 ˚C. After 24 h, the polycarbonate membranes were removed and ployment program contract (2020.03171.CEECIND). The radius was cor-
placed on 1 mL of saline (0.9% w/v NaCl). The samples were vortexed rected to diameter on June 12, 2023.
thoroughly, serially diluted in saline, and plated on TSA to quantify the
number of CFUs.
Levofloxacin Release from the MNs: The release of LVX from the MNs Conflict of Interest
at 37 ˚C was evaluated. Briefly, MNs loaded with different concentrations
of LVX (4–128 μg mL−1 ) were removed from the PDMS molds, placed in The authors declare no conflict of interest.
5 mL of water, and incubated at 37 ˚C under agitation (120 rpm) (Advanced
3500 Orbital Shaker, VWR). Samples (200 μL) were taken every 2 min until
10 min, every 5 min until 30 min, and every 10 min until 60 min. The LVX Data Availability Statement
concentration was determined by spectrofluorimetry (excitation at 295 nm
and emission at 440 nm) using a microplate reader (SYNERGY H1 Biotek). The data that support the findings of this study are available from the cor-
The data were presented as the cumulative percentage of LVX released. responding author upon reasonable request.
Optical and Scanning Electron Microscopy Visualization: Lateral views
of the 3D-printed MNA master molds and PVA MNs were acquired with
a wide-field upright optical microscope (Nikon—Eclipse Ni-E, Japan) cou- Keywords
pled with a one-color camera (DS-Fi2, Nikon, Japan). The microneedle ar-
biofilms, 3D printing, Klebsiella pneumoniae, levofloxacin, microneedles,
rays (3D printed and PVA) were sputtered with gold (Leica Microsystems,
polyvinyl alcohol
EM ACE200, Wetzlar, Germany) and visualized by scanning electron mi-
croscopy (Quanta FEG 650, FEI Europe B.V., Eindhoven, Netherlands).
Statistical Analysis: Mean and standard deviations (SD) were deter- Received: November 23, 2022
mined for at least three independent experiments (n). Statistical compar- Revised: February 24, 2023
ison was performed using One-Way Analysis of Variance (ANOVA) and Published online: April 19, 2023
[30] Y. Ye, J. Yu, D. Wen, A. R. Kahkoska, Z. Gu, Adv. Drug Delivery Rev.
2018, 127, 106.
[1] S. Adepu, S. Ramakrishna, Molecules 2021, 26, 5905.
[31] J. G. Turner, L. R. White, P. Estrela, H. S. Leese, Macromol. Biosci.
[2] N. Dragicevic, H. I. Maibach, Nanocarriers 2016. https://doi.org/10.
2021, 21, 2000307.
1007/978-3-662-47862-2
[32] R. F. Donnelly, M. T. C. McCrudden, A. Z. Alkilani, E. Larrañeta, E.
[3] D. Ramadon, M. T. C. McCrudden, A. J. Courtenay, R. F. Donnelly,
McAlister, A. J. Courtenay, M. C. Kearney, T. R. Raj Singh, H. O. Mc-
Transl. Res. 2022, 12, 758.
Carthy, V. L. Kett, E. Caffarel-Salvador, S. Al-Zahrani, A. D. Woolfson,
[4] W. Y. Jeong, M. Kwon, H. E. Choi, K. S. Kim, Biomater. Res. 2021, 25,
PLoS One 2014, 9, e111547.
24.
[33] K. Ahmed Saeed AL-Japairai, S. Mahmood, S. Hamed Almurisi, J.
[5] J. Hadgraft, M. E. Lane, Phys. Chem. Chem. Phys. 2011, 13, 5215.
Reddy Venugopal, A. Rebhi Hilles, M. Azmana, S. Raman, Int. J.
[6] Y. C. Kim, J. H. Park, M. R. Prausnitz, Adv. Drug Delivery Rev. 2012, 64,
Pharm. 2020, 587, 119673.
1547.
[34] L. K. Branski, A. Al-Mousawi, H. Rivero, M. G. Jeschke, A. P. Sanford,
[7] I. H. Bajwa, N. Kaur, J. M. Dsouza, J. L. Mathew, Syst. Rev. 2022, 11,
D. N. Herndon, Surg. Infect. 2009, 10, 389.
170.
[35] D. Gu, N. Dong, Z. Zheng, D. Lin, M. Huang, L. Wang, E. W. C. Chan,
[8] M. S. Gerstel, V. A. Place, US-3964482-A, 1976.
L. Shu, J. Yu, R. Zhang, S. Chen, Lancet Infect. Dis. 2018, 18, 37.
[9] S. Henry, D. V. McAllister, M. G. Allen, M. R. Prausnitz, J. Pharm. Sci.
[36] D. M. P. De Oliveira, B. M. Forde, T. J. Kidd, P. N. A. Harris, M. A.
1998, 87, 922.
[10] P. K. Campbell, K. E. Jones, R. J. Huber, K. W. Horch, R. A. Normann, Schembri, S. A. Beatson, D. L. Paterson, M. J. Walker, Clin. Microbiol.
IEEE Trans. Biomed. Eng. 1991, 38, 758. Rev. 2020, 33, e00181.
[11] P. Åberg, P. Geladi, I. Nicander, J. Hansson, U. Holmgren, S. Ollmar, [37] H. S. Gill, D. D. Denson, B. A. Burris, M. R. Prausnitz, Clin. J. Pain
Skin Res. Technol. 2005, 11, 281. 2008, 24, 585.
[12] K. B. Vinayakumar, P. G. Kulkarni, M. M. Nayak, N. S. Dinesh, G. [38] Y. Bozkurt, E. Karayel, J. Mater. Res. Technol. 2021, 14, 1430.
M. Hegde, S. G. Ramachandra, K. Rajanna, J. Micromech. Microeng. [39] A. Sachan, R. J. Sachan, J. Lu, H. Sun, Y. J. Jin, D. Erdmann, J. Y. Zhang,
2016, 26, 065013. R. J. Narayan, MRS Adv. 2021, 6, 61.
[13] K. Oka, S. Aoyagi, Y. Arai, Y. Isono, G. Hashiguchi, H. Fujita, Sens. [40] M. Gülçür, J. M. Romano, P. Penchev, T. Gough, E. Brown, S. Dimov,
Actuators, A 2002, 97–98, 478. B. Whiteside, CIRP J. Manuf. Sci. Technol. 2021, 32, 311.
[14] P. Dardano, M. Battisti, I. Rea, L. Serpico, M. Terracciano, A. Cam- [41] C. L. Caudill, J. L. Perry, S. Tian, J. C. Luft, J. M. DeSimone, J. Controlled
marano, L. Nicolais, L. De Stefano, Adv. Ther. 2019, 2, 1900036. Release 2018, 284, 122.
[15] L. R. Pires, K. B. Vinayakumar, M. Turos, V. Miguel, J. Gaspar, J. Pers. [42] S. C. Balmert, C. D. Carey, G. D. Falo, S. K. Sethi, G. Erdos, E. Kork-
Med. 2019, 9, 49. maz, L. D. Falo, J. Controlled Release 2020, 317, 336.
[16] M. Azmana, S. Mahmood, A. R. Hilles, U. K. Mandal, K. A. Saeed [43] P. Bhushan, Y. Umasankar, J. D. Hutcheson, S. Bhansali, Toxicol. Vitr.
Al-Japairai, S. Raman, J. Drug Delivery Sci. Technol. 2020, 60, 101877. 2019, 58, 170.
[44] A. de Lacerda Bukzem, D. M. dos Santos, I. S. Leite, N. M. Inada, S.
[17] N. Uchida, M. Yanagi, H. Hamada, Nanomaterials 2021, 11, 335.
P. Campana-Filho, Int. J. Biol. Macromol. 2021, 156, 459.
[18] K. B. Vinayakumar, G. M. Hegde, S. G. Ramachandra, M. M. Nayak,
[45] E. R. Kenawy, E. A. Kamoun, M. S. Mohy Eldin, M. A. El-Meligy, Arab.
N. S. Dinesh, K. Rajanna, Biointerphases 2015, 10, 021008.
J. Chem. 2014, 7, 372.
[19] K. B. Vinayakumar, G. M. Hegde, M. M. Nayak, N. S. Dinesh, K. Ra-
[46] Y. Liu, S. Zhou, Y. Gao, Y. Zhai, Asian J. Pharm. Sci. 2019, 14, 130.
janna, Microelectron. Eng. 2014, 128, 12.
[47] G. Bonfante, H. Lee, L. Bao, J. Park, N. Takama, B. Kim, Micro Nano
[20] M. Dehghan, M. Tahmasebipour, S. Ebrahimi, Microelectron. Eng.
2022, 254, 111695. Syst. Lett. 2020, 8, 13.
[21] R. D. Sochol, E. Sweet, C. C. Glick, S. Y. Wu, C. Yang, M. Restaino, L. [48] M. D. Silva, S. Sillankorva, Crit. Rev. Microbiol. 2019, 45, 595.
Lin, Microelectron. Eng. 2018, 189, 52. [49] G. J. Noel, Clin. Med. Ther. 2009, 1, CMT.S28.
[22] A. I. Bunea, M. H. Jakobsen, E. Engay, A. R. Bañas, J. Glückstad, Micro [50] S. J. Martin, R. Jung, C. G. Garvin, Drug Saf. 2001, 24, 199.
Nano Eng. 2019, 2, 41. [51] A. M. Pinto, M. D. Silva, L. M. Pastrana, M. Bañobre-López, S. Sil-
[23] N. Convery, N. Gadegaard, Micro Nano Eng. 2019, 2, 76. lankorva, FEMS Microbiol. Rev. 2021, 45, fuab019.
[24] S. R. Dabbagh, M. R. Sarabi, R. Rahbarghazi, E. Sokullu, A. K. Yetisen, [52] M. S. Arshad, A. T. Zahra, S. Zafar, H. Zaman, A. Akhtar, M. M. Ayaz,
S. Tasoglu, iScience 2021, 24, 102012. I. Kucuk, M. Maniruzzaman, M. W. Chang, Z. Ahmad, Pharm. Res.
[25] C. Yeung, S. Chen, B. King, H. Lin, K. King, F. Akhtar, G. Diaz, B. Wang, 2021, 38, 165.
J. Zhu, W. Sun, A. Khademhosseini, S. Emaminejad, Biomicrofluidics [53] X. Yu, J. Zhao, D. Fan, Chem. Eng. J. 2022, 437, 135475.
2019, 13, 064125. [54] J. W. Lee, J. H. Park, M. R. Prausnitz, Biomaterials 2008, 29, 2113.
[26] A. R. Johnson, A. T. Procopio, 3D Print. Med. 2019, 5, 2. [55] B. Xu, G. Jiang, W. Yu, D. Liu, Y. Zhang, J. Zhou, S. Sun, Y. Liu, J. Mater.
[27] K. J. Krieger, N. Bertollo, M. Dangol, J. T. Sheridan, M. M. Lowery, E. Chem. B 2017, 5, 8200.
D. O’Cearbhaill, Microsyst. Nanoeng. 2019, 5, 42. [56] M. J. Kim, S. C. Park, B. Rizal, G. Guanes, S. K. Baek, J. H. Park, A. R.
Betz, S. O. Choi, Front. Bioeng. Biotechnol. 2018, 6, 54.
[28] M. Dalvi, P. Kharat, P. Thakor, V. Bhavana, S. B. Singh, N. K. Mehra,
[57] T. Waghule, G. Singhvi, S. K. Dubey, M. M. Pandey, G. Gupta, M.
Life Sci. 2021, 284, 119877.
Singh, K. Dua, Biomed. Pharmacother. 2019, 109, 1249.
[29] S. Sillankorva, L. Pires, L. M. Pastrana, M. Bañobre-López, Viruses
[58] J. Sousa, S. Sillankorva, A. Faustino, C. Carvalho, Curr. Genet. 2022,
2022, 14, 964.
68, 393.

Levofloxacin Loaded Microneedles Produced Using 3D Printed Molds For Noso Infection

Uploaded by

Copyright:

Available Formats

Levofloxacin Loaded Microneedles Produced Using 3D Printed Molds For Noso Infection

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Levofloxacin Loaded Microneedles Produced Using 3D Printed Molds For Noso Infection

Uploaded by

Copyright:

Available Formats

RESEARCH ARTICLE

Levoﬂoxacin-Loaded Microneedles Produced Using

various methods are being adapted to

MNs arrays. The 3D-printed microneedle arrays fabricated were

Ref. Method Height [μm] Base [μm] Tip [μm]

as this gram-negative species is frequently isolated from infected

2.3.1. Minimum Inhibitory Concentration (MIC)

The MIC of LVX against K. pneumoniae PC28 planktonic cells

Figure 5. Printed and biopolymer microneedle geometries for diﬀerent as-

Figure 11. Biopolymer microneedle array fabrication steps.

You might also like