RESEARCH PAPER
Determination of Depth-Dependent Intradermal
Immunogenicity of Adjuvanted Inactivated Polio Vaccine
Delivered by Microinjections via Hollow Microneedles
Pim Schipper
1
& Koen van der Maaden
1
& Stefan Romeijn
1
& Cees Oomens
2
& Gideon Kersten
1,3
&
Wim Jiskoot
1
& Joke Bouwstra
1
Received: 1 March 2016 /Accepted: 27 April 2016 /Published online: 17 June 2016
#
Springer Science+Business Media New York 2016
ABSTRACT
Purpose The aim of this study was to investigate the depth-
dependent intradermal immunogenicity of inactivated polio
vaccine (IPV) delivered by depth-controlled microinjections
via hollow microneedles (HMN) and to investigate antibody
response enhancing effects of IPV immunization adjuvanted
with CpG oligodeoxynucleotide 1826 (CpG) or cholera
toxin (CT).
Methods A novel applicator for HMN was designed to per-
mit depth- and volume-controlled microinjections. The appli-
cator was used to immunize rats intradermally with monova-
lent IPV serotype 1 (IPV1) at injection depths ranging from 50
to 550 μm, or at 400 μm for CpG and CT adjuvanted immu-
nization, which were compared to intramuscular
immunization.
Results The applicator allowed accurate microinjections into
rat skin at predetermined injection depths (50–90 0 μm),
-volumes (1–100 μL) and -rates (up to 60 μL/min) with min-
imal volume loss (±1–2%). HMN-mediated intradermal im-
munization resulted in similar IgG and virus-neutralizing an-
tibody titers as conventional intramuscular immunization. No
differences in IgG titers were observed as function of injection
depth, however IgG titers were significantly increased in the
CpG and CT adjuvanted groups (7-fold).
Conclusion Intradermal immunogenicity of IPV1 was not af-
fected by injection depth. CpG and CT were potent adjuvants
for both intradermal and intramuscular immunization,
allowing effective vaccination upon a minimally-invasive sin-
gle intradermal microinjection by HMN.
KEY WORDS dose sparing
.
hollow microneedles
.
intradermal immunization
.
inactivatedpolio vaccine
.
poliomyelitis
ABBREVIATIONS
CpG CpG oligodeoxynucleotide 1826
CT Cholera toxin
DC Dendritic cell
DDC Dermal dendritic cell
H&E Hematoxylin and eosin
HMN Hollow microneedle
IPV Inactivated polio vaccine
IPV1 Monovalent inactivated polio vaccine serotype 1
LC Langerhans cell
OPV Oral polio vaccine
PBS Phosphate buffered saline
TMB 3,3′5,5′-tetramethylbenzidine
VN Virus neutralizing
INTRODUCTION
Poliomyelitis can be prevented through vaccination by either
oral polio vaccine (OPV) or inactivated polio vaccine (IPV).
Although OPV is an inexpensive and easy to administer vac-
cine, it may cause outbreaks of vaccine-derived polioviruses
(1). Therefore, the World Health Organization aims to elim-
inate the use of OPV and substitute it by IPV in its goal
towards worldwide polio eradication (2). However, IPV vac-
cination is more costly because of higher production costs and
a higher dose requirement in comparison to OPV. Therefore,
* Joke Bouwstra
bouwstra@lacdr.leidenuniv.nl
1
Division of Drug Delivery Technology, Cluster BioTherapeutics, Leiden
Academic Centre for Drug Research, Leiden University, P.O. Box 9502,
2300 RA Leiden, The Netherlands
2
Soft Tissue Biomechanics and Engineering, Department of
Biomedical Engineering, Eindhoven University of
Technology, Eindhoven, The Netherlands
3
Intravacc (Institute for Translational
Vaccinology), Bilthoven, The Netherlands
Pharm Res (2016) 33:2269–2279
DOI 10.1007/s11095-016-1965-6
several strategies for dose sparing to reduce the cost of IPV
vaccination have been proposed, including intradermal im-
munization and IPV adjuvantation (3).
Although intradermal IPV immunization at reduced doses
in human resulted in seroconversion (4), intradermal injection
by the Mantoux technique is difficult to perform. Therefore,
there is an urgent need for novel intradermal injection
methods. Intradermal IPV immunization via jet injectors re-
sulted in seroconversion at reduced doses (5,6). Another novel
strategy for intradermal injection is microneedles, which are
micron-sized needles (7). Microneedles were investigated on
their ability to induce protective immune responses upon in-
tradermal IPV delivery on rats and rhesus macaques (8–12).
Hollow microneedle (HMN) mediated intradermal IPV im-
munization at reduced doses was also investigated in humans,
which resulted in similar seroconversion rates in comparison
to a full intramuscular dose (13).
Although intradermal immunization is a promising immu-
nization strategy, it is unknown whether the immunogenicity
of intradermally administered IPV is affected as a function of
injection depth. This is of interest, as several classes of dendrit-
ic cells (DCs) reside in either the epidermis or the dermis.
Langerhans cells (LCs) reside in the epidermis (topmost skin
layer) and dermal dendritic cells (DDCs) reside in the dermis
(lower layer) (14). Both LCs and DDCs have distinct immune
functions and may therefore have a different role in immuni-
zation (14). Therefore, targeting different skin depths may
affect the efficiency of intradermal IPV immunization.
Hence, the objective of the present study is to investigate in-
tradermal IPV immunization efficiency as function of skin
injection depth by using a HMN system. To this end, precise
and reproducible injection of IPV into the skin at a predefined
depth is a requirement. This requirement can be fulfilled by
using our previously developed in-house applicator (8 ).
However, this applicator only allowed a maximum flow rate
of 2 μL/min without leakage in the microinjection system.
Therefore, the system was thoroughly redesigned to allow
increased pressures, resulting in increased ranges of injection
rates, −volumes and -depths without leakages.
Besides establishing the potential effect of intradermal in-
jection depth on the immune response, the use of adjuvants
can improve IPV immunization efficiency and therefore may
result in IPV dose sparing and thus co st reduction (3,15).
Although colloidal aluminum hydroxide or aluminum phos-
phate salts have been used with IPV for intramuscular immu-
nization (16–20), they are not for intradermal use as they
cause severe side effects. Therefore, in this study we examined
the immune potentiating effects of CpG oligodeoxynucleotide
1826 (CpG) and cholera toxin (CT) as potentially suitable
adjuvants for intradermal IPV immunization.
In this study we demonstrate a HMN applicator that allows
for injection depth, -rate and -volume controllable minimally-
invasive intradermal microinjections. This applicator allowed
to investigate the dependence of intradermal immunogenicity
to the injection depth of IPV1. To our knowledge this is the
first systematic study where the influence of injection depth on
antigen-specific immune responses is investigated.
Additionally, the potential of adjuvants CpG and CT to en-
hance intradermal IPV immunization for dose sparing was
investigated.
MATERIALS AND METHODS
Materials
Polyimide coated fused silica capillary (375 μm outer diame-
ter, 100 and 20 μm inner diameter) was obtained from
Polymicro, Phoenix, USA. Silicone oil AK 350 was purchased
from Boom Chemicals, Meppel, the Netherlands. CapTite™
connections were obtained from Labsmith, USA. Parafilm
was purchased from Bemis, Monceau-sur-Sambre, Belgium.
Tissue-Tek O.C.T . compound was ordered at Sakura
Finetek, Alphen aan den Rijn, the Netherlands. Phosphate
buffered saline, pH 7.4 (PBS pH 7.4: 163.9 mM Na
+
,
140.3 mM Cl
−
, 8.7 mM HPO
4
2−
and 1.8 mM H
2
PO
4
−
,pH
7.4) was purchased from B. Braun Melsungen, Melsungen,
Germany. IsoFlo® (isoflurane 100% w/w) was obtained from
Abbott Laboratories, Maidenhead, UK. CpG was purchased
from Invivogen, Toulouse, France. Hydrofluoric acid (49%
w/w), concentrated sulfuric acid (95–98%), fluorescein,
trypan blue, CT (holotoxin) and 3,3′ 5,5′ -
tetramethylbenzidine (TMB) were obtained from Sigma-
Aldrich, Zwijndrecht, the Netherlands. Polystyrene 96 well
microtiter plates and 2.5 mL Vacuette® Z serum separator
clot activator premium tubes were ordered at Greiner Bio-
One, Alphen aan den Rijn, the Netherlands. Tween 80 and
30% w/w hydrogen peroxide were obtained from Merck,
Amsterdam, the Netherlands. PBS pH 7.2 (160.6 mM Na
+
,
155.2 mM C l
−
, 2.7 mM HPO
4
2−
, 1.5 mM H
2
PO
4
−
and
1.5 mM K
+
,pH7.2)waspurchasedfromGibco(Life
Technologies), Bleiswijk, the Netherlands. Protifar was ob-
tained from Nutricia, Zoetermeer, the Netherlands. Bovine
anti-poliovirus type 1 serum, monovalent IPV vaccine sero-
type 1 (IPV1), 1.1 M sodium acetate and 2 M sulfuric acid
were kindly provided by Intravacc, Bilthoven, the
Netherlands. Horseradish peroxidase-conjugated goat-anti-
rat IgG was obtained from Southern Biotech, Birmingham,
AL, USA. Female Wistar Han IGS rats (Crl:WI(Han), strain
code 273) of 175–225 g were ordered from Charles River
Laboratories, Saint-Germain-sur-l’Arbresle, France.
Fabrication of HMN
HMN were produced by an in-house process as described
previously (8). In short, the inner lumen of 20 μminner
2270 Schipper et al.
diameter polyimide coated fused silica capillaries were
filled overnight with silicone oil AK 3 50 by use of a
vacuum oven at 100°C. These silicone oil-filled capil-
laries were subsequently wet etched into HMN by im-
mersing the ends in a container with hydrofluoric acid
(49% w/w) for 4 h. To expose the microneedle tips, t he
polyimide coating at the etched ends of the cap illaries
was removed by immersing them in concentrated s ulfu-
ric acid (95–98%) a t 250°C for 5 min.
HMN Applicator
The HMN applicator that was previously developed in-house
(8 ) was thoroughly redesigned to enable better depth-
controlled intradermal microinjections. This HMN applicator
is depicted in Fig. 1a and b. Optimization was dedicated to
improve accuracy, precision and reproducibility of microin-
jections, to allow for increased injection rates to decrease the
injection time, to allow for higher injection volumes and to
achieve skin depth-controlled injections. To achieve these im-
provements, high-pressure resistance in the fluidics system was
required. Therefore, flexible materials were replaced by non-
flexible rigid materials and all previous connectors in the flu-
idics system were replaced by high-pressure resistant
CapTite™ connectors. For the fluidics system, an 100 μL
Hamilton gas-tight Luer-Lock syringe (model 1710 TLL,
Hamilton Robotics, Bonaduz, Switzerland) with a barrel in-
ner diameter of 1.46 mm was used in conjunction with a
syringe pump (NE-300, Prosense, Oosterhout, the
Netherlands) and 100 μm inner diameter polyimide coated
fused silica capillaries.
HMN Applicator Performance Validation with Ex-Vivo
Rat Skin
Microinjections into ex-vivo rat skin were performed to deter-
mine the range of injection rates, −volumes and –depths that
can be used for leakage-free microinjections. Ex-vivo shaved
rat skin was isolated from sacrificed female Wistar Han rats.
The skin was stretched on Styrofoam covered with parafilm.
Subsequently, microinjections of a solution of 10 μg/mL fluo-
rescein in PBS pH 7.4 were performed at various injection
rates, −volumes and -depths. To determine t he accuracy
and repeatability of the microinjections, volume loss that oc-
curred at i) connections of the fluidics system, ii) on the skin
surface at the injection site or iii) due to retained volume on
the microneedle after its withdrawal from the injection site
was measured by pipetting the lost volume with a 0.2–2 μL
pipette. A pipetted volume below 0.2 μL was measured as a
volume loss of 0.2 μL. The percentage volume loss was calcu-
lated as the percentage of volume loss (as measured by pipet-
ting) from the digitally-displayed dispensed volume (as indicat-
ed by the syringe pump).
In the first series of experiments, the injection rate perfor-
mance was investigated by varying injection rates from 1 to
60 μL/min, while both the injection volume and -depth were
kept constant at 10 μL and at 500 μm, respectively. The
maximum injection rate of the syringe pum p for the
1.46 mm inner diameter 100 μL Hamilton syringe was
62 μL/min. In the second series of experiments, the injection
volume was investigated by varying injection volumes from 1
to 100 μL, while both the injection rate and -depth were kept
constant at 20 μL/min and at 500 μ
m, respectively. Finally,
in
jection depths ranging from 50 to 900 μm were investigated
by performing 10 μL microinjections at an injection rate of
20 μL/min.
Visualization of Depth-Controlled Microinjections
in Ex-Vivo Rat Skin
In order to visualize microinjections at different preselected
depths in ex-vivo rat skin, trypan blue solution (0.4% w/v in
PBS pH 7.4) was used. Ex-vivo rat skin was isolated and pre-
pared as de scribed above. Next, trypan blue solution was
injected at 250, 400 and 550 μm depths with an injection
volume of 10 μL and at an injection rate of 20 μL/min. The
microinjections in ex-vivo rat skin were photographed on outer
and inner skin sides by utilizing a stereo microscope at 2.5×
magnification (Zeiss Stemi 2000-c, paired with a Zeiss
AxioCam ICc5 camera).
Furthermore, ex-vivo rat skin injected with 0.5 μLtrypan
blue solution at the same injection depths and at an injection
rate of 1 μL/min was embedded in Tissue-Tek O.C.T. com-
pound and subsequently snap-frozen in liquid nitrogen, from
which 10 μm thick cryosections were made on a Leica
CM3050s cryostat. Subsequently, unstained and hematoxylin
and eosin (H&E) stained images were made to visualize the
injection depth of the microinjections. Images were made with
a Zeiss 10× plan-Apochromat objective mounted onto a Zeiss
Axio Imager D2 microscope coupled with a MRc5 camera.
Investigation of Depth-Dependent Intradermal
Immunogenicity
Two immunization studies were performed under the guide-
lines and regulations enforced by the animal ethic committee
of the Netherlands, and were approved by the
BDierexperimentencommissie Universiteit Leiden (UDEC)^
under number 12084. Female Wistar Han rats with a weight
of 175–225 g on arrival were accommodated under standard-
ized conditions in the animal facilities of the Leiden Academic
Centre for Drug Research, Leiden University. The animals
were housed in groups of 5 and were assigned to different
immunization groups (10 rats per immunization group).
Prior to blood withdrawal or immunization, the rats were
anaesthetized with isoflurane. Anaesthetized animals that
Depth-Dependent and Adjuvanted Intradermal Immunogenicity of IPV 2271
were assigned to the intradermal injection groups were shaved
minimally (an area of 4 cm
2
on the left flank) prior to the
intradermal injection.
To investigate the depth-dependent intradermal immuno-
genicity of IPV1, the rats were immunized intradermally via
HMN mediated microinjections of 10 μL containing 5 DU of
IPV1 at an injection rate of 20 μL/min and at injection depths
of 250, 400, 550 μm for animal study 1, and 50, 150, 250 μm
for animal study 2. As a control in both animal studies, 5 DU
of IPV1 in 200 μL PBS pH 7.4 was administered intramus-
cularly divided over each hind leg (100 μL per hind leg).
Furthermore, both animal studies contained a mock treated
group via the intramuscular route (100 μL PBS pH 7.4 per
hind leg). In the first animal study the immunogenicity en-
hancing effects of adjuvants were investigated: two groups
were immunized intradermally via HMN mediated microin-
jections of 10 μL (injection rate 20 μL/min and injection
depth 400 μm) and two groups were immunized intramuscu-
larly (100 μL per hind leg) with 5 DU IPV1 adjuvanted with
either CpG or CT. These previously described immunization
procedures were performed at day 1 (prime immunization)
and were repeated at day 21 (booster immunization).
Collection of blood samples were performed at day 1, day
21 (prime) and day 42 (boost). Blood samples were collected
in 2.5 mL Vacuette® tubes and stored on ice before centrifu-
gation at 2000 g for 10 min to isolate serum.
Serum IgG Titers
To measure serum IgG titers, a capture ELISA was per-
formed. Bovine anti-poliovirus type 1 serum in PBS pH 7.2
was used to coat polystyrene 96 well microtiter plates over-
night at 4°C. Subsequently, washing was performed with
0.05% Tween 80 in tap water. Afterwards, assay buffer
consisting of PBS pH 7.2, 0.5% (w/v) Protifar and 0.05%
(v/v) Tween 80 was used to add 4.5 DU IPV1/well
(100 μL/well). Plates were incubated at 37°C for 2 h before
a wash step was performed. Afterwards, threefold serial dilu-
tions of serum samples in assay buffer were added at 100 μL/
well and subsequently incubated at 37°C for 2 h. Plates were
washed before horseradish peroxidase-conjugated goat-anti-
rat IgG was added to the wells (4000-fold dilution, 100 μL/
well) and subsequently incubated at 37°C for 1 h.
Subsequently, plates were thoroughly washed before adding
TMB substrate solution (100 μL/well) which consisted o f
1.1 M sodium acetate, 100 mg/mL TMB and 0.006% (v/v)
hydrogen peroxide. 2 M sulfuric acid was used after 10 min to
stop the reaction (100 μL/well). Finally, sample absorbance
was measured at 450 nm by a Biotek ELx808 plate reader
(Winooski, VT, USA).
The Biotek Gen5 2.0 data analysis software was used to
determine endpoint titers by 4-parameter analysis. The end-
point titer was defined as the reciprocal of the serum dilution
producing a 450 nm absorbance equal to that of the mean
450 nm absorbance with addition of three times the standard
deviation of eight samples of IPV1-specific-antibody negative
serum samples.
Virus-Neutralizing Antibodies
Determination of the antibody titers able to neutralize
wildtype poliovirus serotype 1 was outsourced to Bilthoven
Biologicals and performed as previously described (21,22). In
short, average virus-neutralizing (VN) antibody titers were
measured after pooling of the serum samples of all rats per
immunization group. Subsequently, two-fold serial dilutions
of the pooled sera were made (2
11
–2
22
) after inactivation of
sera at 56°C for 30 min prior to testing. 100TCID
50
of the
Mahoney wild-type strain (poliovirus type 1) was added to the
resulting serum dilutions and these virus/serum mixtures were
incubated for 3 h at 36°C and 5% CO
2
before incubation at
4°C overnight. Subsequently, to each sample 1 × 10
4
Vero
cells were added and the resulting mixtures were incubated
for 7 days at 36°C and 5% CO
2
. Finally, samples were fixed
with formalin, stained with crystal violet and were analyzed
macroscopically. VN titers were displayed as the last serum
dilution which did not exhibit cytopathogenic effects.
Statistical Analysis
Statistics were performed using G raphPad Prism (v.6.00,
GraphPad Software, LaJolla, CA, USA). Kruskall-Wallis tests
with Dunn’s post-hoc tests were performed as IgG titers were
non-normally distributed and considered significant at
p <0.05.
RESULTS
HMN Applicator Optimization and Performance
To allow for increased injection rates and thereby short injec-
tion times, connections in the fluidics system of the HMN
applicator were optimized by applying high-pressure resistant
CapTite™ connectors and high-pressure resistant polyimide
fused silica capillaries, as shown in Fig. 1a and b. Furthermore,
the dead volume in the fluidics system was kept to a minimum
to maximize injection volume accuracy and minimize the loss
of vaccine formulation. The hydrofluoric acid e tching
procedure of polyimide coated fused silica capillaries
resulted in sharp HMN (Fig. 1c).
The performance of the fluidics system was examined by
varying the injection rate during intradermal microinjections
into ex-vivo rat skin, mimicking back pressure during actual
microinjections on live animals. As shown in Fig. 2a,thevol-
ume loss was ±1–2% for all investigated injection rates at an
2272 Schipper et al.
injection volume and -depth of 10 μL and 500 μm, respec-
tively. Volume loss was only observed at the skin surface of the
microinjection sites or on the HMN aft er retracting the
microneedle from the skin. No additional volume loss was
observed. Owing to the achieved increase in injection rate, a
significant reduction in injection time was achieved.
As shown in Fig. 2b, volumes up to 100 μLcouldbe
injected into ex-vivo rat skin with volume loss that was ±1–
2%, independent of injection volume, at an injection depth
and -rate of 500 μmand20μL/min, respectively. This vol-
ume loss was observed on the skin surface at the microinjec-
tion sites.
Finally, variation in injection depth was studied by injecting
10 μLinex-vivo rat skin at different injection depths. As shown
in Fig. 2c, microinjections were performed between 50 and
900 μm injection depths. At all investigated injection depths,
all microinjections resulted in ±1–2% volume loss. Moreover,
there was no increase in volume loss at any particular
injection depth.
Visualization of Depth-Controlled Microinjections
in Ex-Vivo Rat Skin
To obtain evidence that skin layers at different preselected
depths were targeted, microinjections of a trypan blue solution
were performed on ex-vivo rat skin, which was then
photographed at both sides (Fig. 3a). Whereas the trypan blue
spot of the shallower microinjection at a skin injection depth of
250 μm was more clearly visible at the outer side of the skin,
the trypan blue spot of the deeper injection at 550 μmwas
hardly visible. Contrarily, the trypan blue spot of the deeper
microinjection at a skin injection depth of 550 μmwasmore
clearly visible at the inner side of the skin than the trypan blue
spot of the shallower injection (250 μm). Moreover, as shown
in Fig. 3b, visualization of the skin injection depth in cross-
sectioned skin indicated that the microinjections were indeed
performed at different pre-defined skin injection depths.
Injection Depth-Dependent Intradermal
Immunogenicity
To assess whether injection of IPV1 at different injection
depths in the skin affects IPV1-specific antibody responses,
two immunization studies in rats were conducted.
Intradermal immunization was performed using the HMN
applicator in a depth-controlled manner and this was com-
pared to intramuscular immunization with a conventional
26G hypodermic needle. IPV1-specific IgG titers obtained
after intradermal immunization with non-adjuvanted IPV1
at injection depths ranging between 50 and 550 μm are shown
in Fig. 4. Three weeks after prime immunization (Fig. 4a)and
boost immunization (Fig. 4b), no significant differences in
IPV1-specific IgG titers were observed at different injection
depths. Moreover, IPV1-specific IgG titers obtained after in-
tradermal immunization were similar to those obtained after
intramuscular IPV1 immunization. No IPV1-specific anti-
body responses were observed in the mock treated group.
Immunogenicity Enhancing Effects of Adjuvants CpG
and CT
To assess the potential increase in antibody responses and the
applicability for use as intradermal adjuvants, CpG and CT
were used as adjuvants in intradermal and intramuscular IPV
immunization (Fig. 4a and
b)
. Intradermal IPV1 immuniza-
tion adjuvanted with CpG led to statistically significant in-
creased IgG titers in comparison to non-adjuvanted 400 and
550 μm injection depths and intramuscular IPV1 group after
prime immunization. Contrarily, intramuscular IPV1 immu-
nization adjuvanted with CpG did not result in any significant
increase in IgG titer after prime immunization. After boost
immunization however, IPV1 immunization adjuvanted with
CpG via both the intradermal and intramuscular route result-
ed in significantly increased IgG titers in comparison to the
non-adjuvanted intramuscular IPV1 group.
a b c
1
2
5
4
7
9
8
6
3
10
Fig. 1 Images of the HMN applicator (a and b) and a scanning electron microscopy image of a HMN (c, bar represents 50 μm). The microneedle insertion
speed (1 to 3 m/s) is controlled by an electromagnet (1). Angled injections are possible via a guided rail (2). The injection depth of the microneedle into the skin is
accurately controlled by a micrometer actuator (10) and a guide plate (9). Fluid flow starts with a 100 μL Hamilton gas-tight Luer-Lock syringe (3) which is driven
by a controllable syringe pump (5) that is connected to a 100 μm inner diameter capillary (6) via a Luer-Lock-CapTite adapter (4). This capillary feeds the fluid flow
via a specially designed connection piece (7)intoaHMN(8) that enables the intradermal injections. Because the syringe pump is programmable, injection rates
and -volumes can be accurately controlled.
Depth-Dependent and Adjuvanted Intradermal Immunogenicity of IPV 2273