Basic Research—Biology
Hypoxia Modulates the Differentiation Potential
of Stem Cells of the Apical Papilla
Julie Vanacker, PhD,* Aiswarya Viswanath,* Pauline De Berdt,* Amandine Everard, PhD,†
Patrice D. Cani, PhD,† Caroline Bouzin, PhD,‡ Olivier Feron, PhD,‡ Anibal Diogenes, DDS, PhD,§
Julian G. Leprince, DDS, PhD,* and Anne des Rieux, PhD*
Abstract
Introduction: Stem cells from the apical papilla (SCAP)
are a population of mesenchymal stem cells likely
involved in regenerative endodontic procedures and
have potential use as therapeutic agents in other tissues. In these situations, SCAP are exposed to hypoxic
conditions either within a root canal devoid of an
adequate blood supply or in a scaffold material immediately after implantation. However, the effect of hypoxia
on SCAP proliferation and differentiation is largely unknown. Therefore, the objective of this study was to
evaluate the effect of hypoxia on the fate of SCAP.
Methods: SCAP were cultured under normoxia (21%
O2) or hypoxia (1% O2) in basal or differentiation media.
Cellular proliferation, gene expression, differentiation,
and protein secretion were analyzed by live imaging,
quantitative reverse-transcriptase polymerase chain reaction, cellular staining, and enzyme-linked immunosorbent assay, respectively. Results: Hypoxia had no effect
on SCAP proliferation, but it evoked the up-regulation of
genes specific for osteogenic differentiation (runtrelated transcription factor 2, alkaline phosphatase,
and transforming growth factor-b1), neuronal differentiation ( 20 -30 -cyclic nucleotide 30 phosphodiesterase,
SNAIL, neuronspecific enolase, glial cell-derived neurotrophic factor and neurotrophin 3), and angiogenesis
(vascular endothelial growth factor A and B). Hypoxia
also increased the sustained production of VEGFa by
SCAP. Moreover, hypoxia augmented the neuronal differentiation of SCAP in the presence of differentiation
exogenous factors as detected by the up-regulation of
NSE, VEGFB, and GDNF and the expression of neuronal
markers (PanF and NeuN). Conclusions: This study
shows that hypoxia induces spontaneous differentiation
of SCAP into osteogenic and neurogenic lineages while
maintaining the release of the proangiogenic factor
VEGFa. This highlights the potential of SCAP to promote
pulp-dentin regeneration. Moreover, SCAP may represent potential therapeutic agents
for neurodegenerative conditions because of their robust differentiation potential. (J Endod 2014;40:1410–1418)
Key Words
Angiogenic factor, apical papilla, dental stem cells, gene expression, hypoxia, neurodifferentiation, regenerative endodontics, stem cells from the apical papilla
I
n the past decade, regenerative endodontic procedures have emerged as a treatment
alternative for immature teeth with pulp necrosis with the goal of pulp tissue regeneration (1, 2). This new treatment modality typically consists of chemical
debridement and disinfection of the root canal followed by placement of an
intracanal medication (3). On a subsequent appointment, the intracanal medicament
is removed, and stem cells are delivered from the apical region by evoked bleeding after
instrumentation beyond the apex (4, 5). The delivered stem cells are located in the
surroundings of the blood clot, often throughout the whole extent of the root canal
to the cementum-enamel junction. Thus, a substantial number of stem cells are
placed several millimeters away from nearby apically positioned blood vessels. In the
case of immature teeth, these procedures aim at restarting the development of the
immature root, possibly increasing its resistance to fracture and survival (6). Many
case reports have been published showing that such procedures are able to lead to healing of the periapical lesion and in a certain proportion of cases to an increase of root
thickness and/or length (7). Thus, these procedures provided unprecedented clinical
benefit in the treatment of immature teeth with pulpal necrosis. Nonetheless, it is unclear whether these procedures are able to recapitulate the once lost pulp-dentin complex.
Multiple animal models of regenerative endodontics have been used to evaluate
the nature of the tissue formed. It was found that, in most of these studies, the tissue
formed consisted of loose fibrous connective tissue with cementumlike tissues, osteodentin, and bonelike tissue (8, 9). It is important to note that, in certain animal studies
using tissue-engineering approaches, complete regeneration of the pulp-dentin complex has been achieved (10, 11). Despite these encouraging reports, histologic
examination of human teeth treated with revascularization procedures suggest that
the tissues formed lack organization with mineralized tissues that resemble
cementum or osteodentin along the dentinal walls and islands of dystrophic
calcifications within the canal lumen (12). Therefore, it appears that several factors
From the *Pharmaceutics and Drug Delivery Unit, Louvain Drug Research Institute, Universite Catholique de Louvain, Brussels, Belgium; †Metabolism and Nutrition
Research Group, WELBIO (Walloon Excellence in Life Sciences and BIOtechnology), Louvain Drug Research Institute, Universite catholique de Louvain, Brussels, Belgium;
‡
Pole of Pharmacology, Institute of Experimental and Clinical Research, Universite Catholique de Louvain, Brussels, Belgium; and §University of Texas Health Science
Center at San Antonio, San Antonio, Texas.
Address requests for reprints to Dr Julie Vanacker, Pharmaceutics and Drug Delivery Unit, Louvain Drug Research Institute, Universite Catholique de Louvain, Avenue
Mounier 73, Box B1.73.12, 1200 Brussels, Belgium. E-mail address: [email protected]
0099-2399/$ - see front matter
Copyright ª 2014 American Association of Endodontists.
http://dx.doi.org/10.1016/j.joen.2014.04.008
1410
Vanacker et al.
JOE — Volume 40, Number 9, September 2014
Basic Research—Biology
need to be optimized in order to achieve better special and temporal
control of tissue regeneration.
For a more orchestrated regenerative process, one of the main factors that needs to be better controlled is mesenchymal stem cell fate. The
cells present in the root canal after induced bleeding were shown to present the characteristics of undifferentiated mesenchymal stem cells
(expression of CD73+, CD105+, or Stro-1+) (5). Whether these cells
are stem cells from the apical papilla (SCAP) or cells of other periradicular sources remains unclear despite the fact that the apical papilla was
clearly shown to possess the ability to remain vital for some time even in
the presence of periapical radiolucency (4, 13). Shortly after blood clot
formation, SCAP are very likely to be in hypoxic conditions until blood
vessels grow into the blood clot. Therefore, they are initially subjected to
very different conditions compared with those used for their in vitro
characterization (21% O2). Culture in hypoxic conditions of dental
pulp stem cells (DPSCs) was shown to affect their proliferation (14,
15), migration ability (16), differentiation (14), and angiogenic potential (17). However, DPSCs are not expected to participate in the current
regenerative endodontic procedures, and the effect of hypoxic conditions on SCAP has never been previously evaluated. Thus, the objective
of the present work was to study the influence of hypoxia (1% O2) on the
proliferation and differentiation potential compared with normoxia
(21% O2).
RP-89 Cell Proliferation
RP-89 cell proliferation was quantified by life imaging. Images
(n = 3–5 zones/well) of 6-well plates seeded with 220,000 RP-89 cells
were acquired with an Axio Observer microscope (Zeiss, Zaventem,
Belgium) either in normoxia (n = 9) or hypoxia (n = 6) every hour
for 72 hours. Six pictures per condition every 24 hours were selected,
and the surface occupied by RP-89 cells was quantified using ImageJ
(National Institutes of Health, Besthesda, MD). Results were expressed
as a percentage of the total analyzed surface. Cell confluence was set as
100% of the analyzed surface.
Osteo-, Adipo-, and Neurodifferentiation of RP-89 cells
Osteogenesis. RP-89 cells were seeded (12,000 cells/cm2) in 6- or
12-well plates (n = 3), incubated 24 hours in normoxia, and grown
either in hypoxia or normoxia until reaching 80% confluence. Cultures
were then induced (day 0) to undergo osteogenesis by replacing the
basal medium with the differentiation medium (StemPro Osteogenesis
Differentiation Kit, Life Technologies). Cells were maintained in osteodifferentiation medium for 4 weeks, and medium was replaced twice a
week. Cultures were stained for calcium phosphate by alizarin red
(Sigma-Aldrich) or treated for quantitative reverse-transcriptase polymerase chain reaction (qRT-PCR) analysis (discussed later). RP-89
cells cultured in basal medium were used as the negative control.
Adipogenesis. RP-89 cells were seeded (12,000 cells/cm2) in 6- or
12-well plates (n = 3), incubated 24 hours in normoxia, and grown
either in hypoxia or in normoxia until reaching 80% confluence. Cultures were then induced (day 0) to undergo adipogenesis by replacing
the basal medium with the differentiation medium (StemPro Adipogenesis Differentiation Kit, Life Technologies). Cells were maintained in adipodifferentiation medium for 4 weeks, and medium was replaced twice
a week. Cultures were stained for neutral lipids with oil red O (SigmaAldrich) or treated for qRT-PCR analysis. RP-89 cells cultured in basal
medium were used as the negative control.
Neurogenesis. RP-89 cells were seeded (3,000 cells/cm2) in
4-well Nunc LabTek II chamber slides (Thermo Fischer Scientific, Waltham, MA) (n = 2) or 6-well plates (n = 3), incubated 24 hours in normoxia, and grown either in hypoxia or normoxia until reaching 50%
confluence. Cultures were then induced (day 0) to undergo neurogenesis by replacing the basal medium with the differentiation medium
adapted from Abe et al (19) and composed of Neurobasal A (Life Technologies), B27 (1/50, Life Technologies), epidermal growth factor
(EGF) (20 ng/mL, Life Technologies), nerve growth factor (NGF)
(50 ng/mL, Life Technologies), basic fibroblast growth factor (bFGF)
(20 ng/mL, Life Technologies), cyclic adenosine monophosphate
(AMPc) (1 mmol/L, Sigma-Aldrich), 3-isobutyl-1-methylxanthine
(IBMX) (0.5 mmol/L, Sigma-Aldrich), retinoic acid (2 mmol/L,
Sigma-Aldrich), and penicillin/streptomycin (PEST) (Life Technologies). Cells were maintained in neurodifferentiation medium for
1 week, and medium was replaced twice a week. Cultures were rinsed
with phosphate-buffered saline, and neurogenic differentiation was assessed by immunofluorescence for Pan neurofilaments (Covance, Brussels, Belgium, SMI-32R, 1/1000) and NeuN (Millipore; Billerica, MA,
MAB377, 1/100). Anti–immunoglobulin G Alexa488 (1/500) was
used as a secondary antibody. Actin was stained with rhodamine phalloidin (Life Technologies, R415, 1/40) and the nucleus with 40 ,6-diamidino-2-phenylindole (DAPI). Images were acquired with a confocal
microscope (LSM700, Zeiss). PanNF and NeuN staining quantifications
were realized with an RGB quantification script (Zen, Zeiss). The percentage of green pixels (positive staining) for a known analyzed surface
(1.44 mm2) was measured and divided by the cell number for the same
surface (n = 10 picture per condition and staining). The number of
cells positive for PanNF and NeuN was also quantified and expressed
as a percentage of the total number of cells. RP-89 cells cultured in basal
medium were used as the negative control.
Vascular Endothelial Growth Factor A Secretion
Supernatants of RP-89 cells cultured in basal medium for 4 weeks
in normoxic and hypoxic conditions (n = 3) were collected and stored
at 20 C for human VEGFa quantification using an enzyme-linked
immunosorbent assay (PeproTech, Rocky Hill, NJ). Samples were not
diluted, and the supplier’s instructions were followed. Results were expressed in ng/mL as cumulated VEGFa concentration over time.
qRT-PCR Analysis
RP-89 cells were cultured as described previously (n = 3) either
in basal or differentiation media in normoxic or hypoxic conditions
during 1 or 4 weeks. Cells were treated with Trizol (Life Technologies),
and cell pellets were frozen at 80 C. Total messenger RNA was extracted by the chloroform extraction method and stored at 80 C.
Approximately 1 mg of messenger RNA was reverse transcribed into
Materials and Methods
SCAP Culture
SCAP previously isolated and characterized from the apical papilla
of immature permanent teeth of a 17-year-old girl (RP-89 cells) were
used (18). RP-89 cells between the 3rd and 6th passage were grown at
37 C in 21% O2 and 5% CO2 in basal culture medium composed of minimum essential medium (Sigma-Aldrich, St Louis, MO) supplemented
with 10% fetal bovine serum (Gemini, Sacramento, CA), 1% L-glutamine (Gemini), and 1% antibiotics (Gemini). RP-89 cell passages
were performed at 80% confluence using Accutase (Life Technologies,
Gent, Belgium). Cell culture in hypoxic conditions (1% O2) was performed as followed. Cells were allowed to attach for 24 hours after seeding in basal culture conditions, referred to as normoxia (humidified cell
incubator, 21% O2, and 5% CO2), before being transferred to hypoxic
conditions (humidified InVivo2 400 Hypoxia Workstation, Ruskinn,
Bridgend, UK; 1% O2). Medium was replaced every other day by fresh
cell culture media previously placed in vented tubes in the hypoxia
chamber to allow media equilibration at 1% O2.
JOE — Volume 40, Number 9, September 2014
SCAP Differentiation in Hypoxia
1411
Basic Research—Biology
complementary DNA (cDNA) (20 mL) (Reverse Transcription System
Kit; Promega, Madison, WI) and stored at 20 C. Primers were designed using the Primer 3 software (Whitehead Institute for Biomedical
Research, Cambridge, MA) against SURVIVIN, CLUSTER DIFFERENTIATION (CD) 90 and CD105, Runt-related transcription factor 2
(RUNX2), alkaline phosphatase (ALP), fatty acid binding protein 4
(ALBP), 20 -30 -cyclic nucleotide 30 phosphodiesterase (CNP), neuronspecific enolase (NSE), SNAIL, vascular endothelial growth factor a
and b (VEGFA and VEGFB), basic fibroblast growth factor (BFGF),
transforming growth factor beta 1 (TGF-b1), glial cell–derived neurotrophic factor (GDNF), neurotrophin 3 (NT3), and actin beta (ACTB)
(Supplemental Figure S1 is available online at www.jendodon.com).
Primer efficiency was calculated, and the most appropriate cDNA concentration was selected to perform qRT-PCR analysis. SYBR green
real-time qRT-PCR analysis (GoTaq qPCR MasterMix Kit, Promega)
was performed on a StepOne Plus Real-Time PCR System (Applied Biosystems, Life Technologies). Each sample was tested in duplicate.
Melting curves were analyzed to guaranty the absence of unspecific
PCR products. A negative control (RNAse-free water) was included in
each run. ACTB was used as a housekeeping gene. The results were
analyzed by StepOne Software V2.1 (Applied Biosystems). The relative
gene expression was calculated according to the delta-delta Ct method
(2 –DDCt). Gene expression in RP-89 cells cultured in hypoxia was
compared with gene expression in RP-89 cells cultured in normoxia after 1 and 4 weeks. Gene expression in cells grown in differentiation media was compared with gene expression in cells grown in basal medium.
Statistical Analysis
Statistical analyses were performed using PRISM (GraphPad Software, San Diego, CA) and JMP10 (SAS Institute, Cary, NC). One- and
2-way analysis of variance with post hoc Bonferroni multiple comparison tests were performed with a threshold significance level set at
alpha = 0.05. Differences were considered significant for a P value
<.05, very significant for a P value < .01, and extremely significant
for a P value < .001. Error bars represent the standard error of the
mean in all figures.
Results
SCAP Proliferation in Hypoxia
Hypoxia did not influence the SCAP proliferation rate; the growth
curves of RP-89 cells cultivated in hypoxia and normoxia were similar,
with comparable slopes (Fig. 1A). The RP-89 cell doubling time was
30.7 hours in hypoxia and 32.2 hours in normoxia.
Effect of Hypoxia on SCAP Phenotype
in Basal Culture Medium
After 1 week in hypoxia, most of the genes were significantly upregulated compared with normoxia (Fig. 1B and C). Neuro- (CNP,
SNAIL, NSE, GDNF, and NT3) and osteo- (RUNX2, ALP, and TGF b1)
specific genes, as well as VEGFA and VEGFB, were particularly upregulated (from 2- to 8-fold). The expression of SURVIVIN, CD105,
and BFGF was not influenced by hypoxia, whereas ALBP expression
was completely suppressed by hypoxia.
After 4 weeks in hypoxia, most of the gene expression decreased
compared with 1 week (from 1.4- to 9.7-folds) (Fig. 1B and C), with the
exception of SURVIVIN and SNAIL, which were expressed more significantly after 4 weeks than after 1 week in hypoxia (2-folds). RUNX2,
SNAIL, NSE, and VEGFA were still expressed more significantly after
4 weeks in hypoxia compared with normoxia (from 2- to 5-fold).
VEGFA expression remained stable over time (Fig. 1C), whereas
CD90, ALP, BFGF, and GDNF were significantly down-regulated
1412
Vanacker et al.
compared with normoxia (from 2- to 5-folds). The expression of
CD105, CNP, VEGFB, TGF b1, and NT3 was not influenced by O2 concentration at this time. ALBP expression was still suppressed in hypoxia.
Moreover, because the VEGFA gene was consistently up-regulated
in hypoxic conditions, VEGFa concentration in RP-89 cell supernatants
was quantified over 1 month. It was significantly higher (P < .001) as
early as after the first 24 hours of incubation (2.13 ng/mL and
0.73 ng/mL, respectively) and then at every time point in hypoxia
compared with normoxia (Fig. 1D). VEGFa production was linear
over time in hypoxia and normoxia (R2 = 0.9894 and R2 = 09657,
respectively), and the 2 linear regressions were significantly different
(P < .001). RP-89 cells produced 0.9 ng/mL VEGFa per day in normoxia, whereas they produced 1.4 ng/mL per day in hypoxia.
Effect of Hypoxia on Induced SCAP
Differentiation Potential
SCAP Differentiate in Osteogenic, Adipogenic, and
Neurogenic Lineages in Normoxia. After 4 weeks of culture in
osteodifferentiation medium in normoxia, SURVIVIN, RUNX2, and TGF
b1 genes were expressed more significantly compared with RP-89 cells
cultured in basal medium (from 1.7- to 2.8-fold) (P < .05), whereas
ALP gene expression decreased (3.8-fold) (P < .001) (Fig. 2A). Alizarin red staining confirmed RP-89 cell differentiation in osteogenic lineage (Fig. 2D).
After 4 weeks in adipodifferentiation medium in normoxia, RP89 cell expression of SURVIVIN and ALBP significantly increased
(1.8- and 1,189-fold, respectively) (P < .05 and P < .001, respectively)
(Fig. 2B), whereas CD90 gene expression decreased (3.2-fold)
(P < .01). Oil red O staining was negative (Fig. 2E), but the 1,189fold increased expression of ALBP confirmed RP-89 cell differentiation
into adipogenic lineage.
After 1 week in neurodifferentiation medium in normoxia,
RP-89 cells overexpressed CNP, SNAIL, NSE, VEGFA, BFGF, and GDNF
(from 1.4- to 8.7-fold, P < .001) (Fig. 2C). SURVIVIN (P < .01),
CD90, and VEGFB (P < .001) were down-regulated (2-fold), whereas
CD105 and NT3 were not affected by the differentiation medium
(Fig. 2C). RP-89 cell proliferation and morphology were influenced
by culture in neurodifferentiation medium. Cell proliferation was significantly lower in neurodifferentiation medium compared with RP-89 in
basal medium (85 cells/1.44 mm2 vs 360 cells/1.44 mm2, respectively)
(Fig. 2F and G). PanNF staining was denser and more defined (Fig. 2F).
Cells developed neurites, easily identified by morphology and PanNF
staining, which confirmed RP-89 cell differentiation into neuronallike cells. Although the level of neurofilament (PanNF) and neuron
(NeuN) marker expression (Fig. 3D and E) was not significantly
different in RP-89 cells in neurodifferentiation medium compared
with RP-89 cells in basal medium, the percentage of positive cells
cultured in neurodifferentiation medium was increased (Fig. 3F and G).
SCAP Differentiate into Osteogenic and Neurogenic
Lineages but Not into Adipogenic Lineage in Hypoxia. After 4 weeks in osteodifferentiation medium in hypoxia, BFGF
(P < .001), VEGFA, and SURVIVIN (P < .01) were up-regulated
compared with normoxia (from 3- to 5-fold), whereas ALP was
down-regulated (20-fold) (P < .05) (Fig. 4A). The other tested genes,
especially RUNX2 and TGF b1, were not affected by hypoxia, neither was
the alizarin red staining (Fig. 4C).
VEGFA and BFGF (P < .001), CD105 and VEGFB (P < .05), and
SURVIVIN (P < .01) were up-regulated when RP-89 cells were grown in
adipodifferentiation medium in hypoxia (from 3.25- to 13.45-fold)
(Fig. 4B). ALBP expression was totally suppressed compared with normoxia. As in normoxia, no oil red O staining was observed (Fig. 4D).
JOE — Volume 40, Number 9, September 2014
Basic Research—Biology
Figure 1. Hypoxia influences SCAP cultured in basal medium. (A) RP-89 cell proliferation was assessed by in vitro culture in normoxia or hypoxia in basal
medium during 72 hours in the chamber of an Axio Observer microscope. Pictures were taken every hour, and around 6 pictures/condition every 24 hours
were analyzed with ImageJ program. Results are expressed as a percentage of the total analyzed surface. R2 indicates the correlation between the obtained values
and the estimated exponential growth. (B) RP-89 cells were cultured in vitro in normoxia and hypoxia in basal medium during 1 or 4 weeks (n = 3). Stemness
genes (SURVIVIN, CD90, and CD105), osteogenic (RUNX2 and ALP), adipogenic (ALBP), neurogenic (CNP, NSE, and SNAIL) lineages genes and (C) growth factor
genes (VEGFA, VEGFB, BFGF, TGF-b1, GDNF, and NT3) were studied. ACTB was used as a housekeeping gene. The gene expression in hypoxia was compared with
normoxia (normoxia = 1). a and ***: P < .001, b and **: P < .01, and c and *: P < .05. Lowercase letters indicate significance to normoxia. Stars indicate
significance between 1 and 4 weeks. (D) Supernatants of RP-89 cells cultured for 4 weeks in normoxia and hypoxia in basal medium were used to analyze human
VEGFa secretion. The cumulated secretion over time was calculated, and results are presented in ng/mL. R2 indicates the correlation between the obtained values
and the estimated linear regression.
When RP-89 cells were cultured in hypoxia in neurodifferentiation medium, NSE, VEGFB, and GDNF gene expression was upregulated (from 3- to 4-fold) compared with normoxia
(P < .001) (Fig. 3A). SURVIVIN and BFGF were down-regulated
(2-fold) (P < .001 and P < .01, respectively), whereas the other
genes were not influenced by hypoxia. Neurites were detected in normoxia and hypoxia (Fig. 3B), but more PanNF staining was visible in
cells cultured in hypoxia compared with normoxia although the difference (2-fold) was not significant (Fig. 3D). Weak NeuN staining
was observed in normoxia, whereas NeuN staining was more intense
(100-fold) when cells were grown in hypoxia (P < .05) (Fig. 3C
and E). However, the percentage of positive cells for PanNF and
NeuN stainings was similar in hypoxia and normoxia (Fig. 3F and G).
Discussion
The present study highlights the significant influence of hypoxic
conditions on the differentiation potential of SCAP. It notably led to
an up-regulation of neurogenic and osteogenic differentiation genes
as well as secretion of proangiogenic factors. Hypoxia had no influence
JOE — Volume 40, Number 9, September 2014
on SCAP proliferation. Once in differentiation medium, SCAP differentiated in neurogenic, osteogenic, and adipogenic lineages in normoxia
but only in neurogenic and osteogenic lineages in hypoxia. Moreover,
hypoxia promoted neurodifferentiation.
SCAP have been described as an interesting cell source in regenerative endodontics and could also be used in other regenerative procedures. In regenerative strategies, SCAP will be grafted in a hydrogel
or another type of scaffold to deliver them locally and, in a similar way
as in regenerative endodontic procedures, are likely to be exposed to
hypoxic conditions until blood vessels grow in the scaffold material.
Hence, like for regenerative endodontic procedures, it is important to
study the impact of hypoxia on SCAP characteristics because cell cultures are commonly performed under atmospheric oxygen tension
(21% O2), which does not correspond to physiological conditions
(20). O2 concentration was indeed identified in a recent review as
a key environmental factor that might play a vital role on mesenchymal
stem cell (MSC) fate and function (21). Hypoxia allows hypoxia
inducible factor 1a (HIF-1a) to translocate to the nucleus and to
induce changes on the metabolism of MSCs (lactate dehydrogenase,
SCAP Differentiation in Hypoxia
1413
Basic Research—Biology
Figure 2. Influence of differentiation on SCAP phenotype in normoxia. (A–C) Gene expression was evaluated on RP-89 cells cultured in vitro in normoxia for
4 weeks in (A) osteo- or (B) adipodifferentiation medium and on (C) RP-89 cells cultured in vitro in normoxia during 1 week in neurodifferentiation medium.
Stemness genes (SURVIVIN, CD90, and CD105), osteogenic (RUNX2 and ALP), adipogenic (ALBP), neurogenic (CNP, NSE, and SNAIL) lineages genes and growth
factor genes (VEGFA, VEGFB, BFGF, TGF-b1, GDNF, and NT3) were studied. ACTB was used as a housekeeping gene. The gene expression in RP-89 cells cultured in
differentiation medium was compared with RP-89 cells cultured in basal medium. a: P < .001, b: P < .01, and c: P < .05. (D–G) RP-89 cells were cultured in
differentiation medium in (right) normoxia and (left) RP-89 cells cultured in basal medium were used as negative controls. (D) RP-89 cells cultured in basal or
osteodifferentiation medium were stained with alizarin red. (E) RP-89 cells cultured in basal or adipodifferentiation medium were stained with oil red O. (F and G)
RP-89 cells cultured in basal or neurodifferentiation medium were stained with an (F) anti-PanNeuF or (G) anti-NeuN antibody. Actin was stained with rhodamine
phalloidin and nucleus with DAPI.
glucose transporter, and so on); on trafficking, homing, and engraftment thanks to the expression of chemokine receptors at the cell surface; and on angiogenesis via VEGF and VEGF receptors (21).
Furthermore, hypoxia could also decrease the production of reactive
oxygen species production in the MSCs and, therefore, decrease the
risk of aneuploidy, DNA damage, and telomere shortening, improving
their biosafety (21).
The proliferation of SCAP in culture was not modulated by hypoxia.
To our knowledge, no previous studies compared the SCAP proliferation rate in hypoxia (1% O2) and normoxia (21% O2). Agata et al
(22) compared the living porcine DPSC number in vitro cultured during 24 hours in normoxia or hypoxia. They showed no difference at 5%
1414
Vanacker et al.
O2 but an important decrease under 0.1% O2. Sakdee et al (15) showed
that human DPSCs proliferated 2 times faster in hypoxia (3% O2). Similarly, Iida et al (14) pointed out an increased proliferation rate after
7 days at 3% O2 compared with normoxia, but they also showed no difference between 1% and 21% O2. Another study using stem cells from
deciduous teeth claimed that they exhibited a higher proliferation in
terms of cell yield compared with DPSCs in hypoxia (23). Thus, it seems
that the impact of hypoxia on the dental stem cell proliferation rate depends on the stem cell type, the percentage of O2 used in the hypoxic
model, the time spend in hypoxia, and maybe also the cell counting
technique. Nonetheless, hypoxia had no effect on SCAP proliferation
in the conditions tested in this study.
JOE — Volume 40, Number 9, September 2014
Basic Research—Biology
Figure 3. Hypoxia induces neurogenic differentiation of SCAP. (A) Gene expression was evaluated on RP-89 cells cultured in vitro in normoxia and hypoxia
during 1 week in neurodifferentiation medium. Stemness genes (SURVIVIN, CD90, and CD105), neurogenic (CNP, NSE, and SNAIL) lineage genes and growth
factor genes (VEGFA, VEGFB, BFGF, GDNF, and NT3) were studied. ACTB was used as a housekeeping gene. The gene expression in hypoxia was compared
with normoxia (normoxia = 1). a: P < .001, b: P < .01, and c: P < .05. (B and C) RP-89 cells were cultured in differentiation medium in (left) normoxia
and (right) hypoxia. RP-89 cells cultured in neurodifferentiation medium were stained with an (B) anti-PanNeuF or (C) anti-NeuN antibody. Actin was stained
with rhodamine phalloidin and nucleus with DAPI. Pictures were obtained with a confocal microscope. Quantification of confocal microscopy PanNF and NeuN
images using an RGB quantification script. (D and E) The percentage of green pixels (positive staining) for a known analyzed surface (1.44 mm2) was measured
and divided by the cell number quantified for the same surface. Nondifferentiated cells were used as negative control. (F and G) The number of cells positive for
PanNF and NeuN was also quantified and expressed as a percentage of the total number of cells.
Hypoxia induced the expression of SURVIVIN, a member of the
inhibitor of apoptosis family (24). In SCAP grown in basal medium,
SURVIVIN was expressed more significantly after 4 weeks in hypoxia
compared with 1 week. SURVIVIN has been shown to be highly expressed in human (25) and mouse embryonic stem cells (26) as
well as in several somatic stem cell types (27). Interestingly, SURVIVIN is induced by HIF-1a in cells cultured in hypoxia (28). An
increased expression of SURVIVIN results in decreased apoptosis
and higher cell proliferation. Thus, the hypoxic hostile environment
leads to the release of this supportive growth factor minimizing cell
JOE — Volume 40, Number 9, September 2014
death possibly through an autocrine, paracrine function on nearby
cells.
Hypoxia resulted in the long-term up-regulation of VEGFA and
VEGFB. In SCAP grown in basal medium, VEGFA and VEGFB were upregulated, which was correlated to a steady higher VEGFa production
by SCAP incubated in hypoxia compared with normoxia. To our knowledge, this is the first study in which the VEGFa secretion by SCAP was
monitored for 4 weeks, showing that SCAP provide constant proangiogenic signals, in hypoxia but also, to a lesser extend, in normoxia. Aranha et al (17) also described that hypoxia enhanced VEGF secretion of
SCAP Differentiation in Hypoxia
1415
Basic Research—Biology
Figure 4. Hypoxia increases induced differentiation of SCAP into osteogenic lineage but not into adipogenic. (A and B) Gene expression was evaluated on RP89 cells cultured in vitro in normoxia and hypoxia in (A) osteo- or (B) adipodifferentiation medium for 4 weeks. Stemness genes (SURVIVIN, CD90, and CD105),
osteogenic (RUNX2 and ALP) and adipogenic (ALBP) lineages genes, and growth factor genes (VEGFA, VEGFB, BFGF, and TGF-b1) were studied. ACTB was used as
a housekeeping gene. The gene expression in hypoxia was compared with normoxia (normoxia = 1). a: P < .001, b: P < .01, and c: P < .05. (C and D) RP-89 cells
were cultured in differentiation medium in (left) normoxia and (right) hypoxia. (C) RP-89 cells cultured in osteodifferentiation medium were stained with alizarin
red. (D) RP-89 cells cultured in adipodifferentiation medium were stained with oil red O.
human DPSCs. Nondifferentiated adipose-derived MSCs were also able
to secrete more VEGF after 3 days in hypoxia (0,1% O2) compared with
normoxia (10 ng/mL and 6 ng/mL, respectively) (29). VEGFa is wellknown to be largely involved in angiogenesis, but it has recently been
described to have neurogenic effects (30) and to be involved in dental
stem cell recruitment (31) and proliferation (32). VEGFb is involved in
neuron protection (33) and neurogenesis (34). These observations
support the potential of SCAP to stimulate revascularization of the
root canal space and to participate in dental pulp regeneration when
placed in hypoxia.
Hypoxia significantly increased SCAP osteogenic differentiation.
Its impact was first evaluated on SCAP grown in basal medium, and it
was observed that RUNX2, ALP, and TGF-b1 were up-regulated after
1 week in hypoxia. TGF-b1 was described as a key factor for dental
pulp reparative and regenerative events (35). Notably, it has been
shown to be involved in dental stem cell recruitment (36), proliferation (37), and odontoblastic differentiation (38). This supports
that the hypoxic environment is not the limiting factor in the regeneration of the dentin-pulp complex. However, the up-regulation of
RUNX2, which is an osteoblastic marker and which was also
described in preodontoblasts and immature odontoblast cells but
not in the terminal differentiation phase (39), could potentially
explain the formation of bone observed in regenerative endodontic
clinical cases. The up-regulation of ALP after 1 week in hypoxia
may also partially explain the mineralization observed in the root canal because this specific gene is involved in the initiation of mineralization. Then, the influence of hypoxia was evaluated on SCAP grown
in osteodifferentiation medium, and it was shown that ALP was down1416
Vanacker et al.
regulated compared with normoxia after 4 weeks. This suggests its
expression is time dependent. Nonetheless, osteogenic differentiation
was evidenced by the higher expression of osteogenesis-specific
genes (RUNX2 and TGF-b1). The mineral deposition, evidenced by
alizarin red staining, did not seem to be significantly different after
4 weeks in normoxia and hypoxia. Thus, it seems that ALP was expressed before 4 weeks, suggesting that ALP might be involved in
the early stages of the mineralization process. This hypothesis is supported by Iida et al (14), who described a decrease of ALP activity after 14 days of osteodifferentiation in hypoxia (3% O2). However, it is
very delicate to directly relate the early genetic events observed at 1
and 4 weeks to the clinical outcome observed in regenerative endodontic procedures. Besides hypoxia, cells are indeed subjected to
other environmental stimulations, such as the release of TGF-b1
from the dentin because of EDTA irrigation (40). Nevertheless, it
can be speculated from our results that hypoxia may be one of the factors triggering a chain of events leading to root width and length gain
through the up-regulation of ALP, RUNX2, and TGF-b.
Hypoxia had a significant effect on adipogenic differentiation.
When SCAP were grown in basal medium, no ALBP expression was
observed in hypoxia although ALBP was highly up-regulated in normoxia when SCAP were cultured in adipogenic differentiation medium. This confirmed the RP-89 cell ability to differentiate into
adipogenic lineage and the complete inhibition of ALBP expression
by hypoxic conditions. The deleterious effect of hypoxia on the adipogenic differentiation of human MSCs was recently described (41, 42).
Zhang et al (41) used hypoxia (1% O2) as a physiological approach to
inhibit mitochondrial activity of human MSCs. They found that
JOE — Volume 40, Number 9, September 2014
Basic Research—Biology
reducing mitochondrial respiration by hypoxia inhibited adipogenic
differentiation. Moreover, the lack of a clear positive oil red O staining
is in accordance with the original work of Sonoyama et al (43), which
reported only very few positive cells in SCAP and DPSC cultures,
whereas more than one third of cells stained positive in BMMSC cultures.
Hypoxia significantly increased the neurogenic differentiation of
SCAP. SCAP were shown to express notably a wide variety of neurogenic
markers such as nestin, neurofilament, bIII tubulin, or NeuN (43),
which makes them particularly interesting candidates for neural tissue
regeneration (eg, in case of spinal cord injury [SCI] for which treatments in clinics are currently quite limited) (44). A novel approach
involving DPSC injection at the level of the lesion has recently been proposed to promote recovery after SCI (45). Encouraging results were reported. Hence, SCAP could also be a promising source of stem cells for
such SCI treatments, during which they would probably be subjected to
hypoxia. In this perspective, the influence of hypoxia on SCAP neurodifferentiation would have a great impact on the success of such a strategy.
When SCAP were grown in basal medium, the expression of
neurogenesis-related genes (CNP, SNAIL, NSE, GDNF, and NT3), was
highly up-regulated after 1 week in hypoxia, whereas SNAIL had greater
expression after 4 weeks in hypoxia. The most highly expressed gene
was NSE, coding an enzyme present in neurons and cells of neuronal
origin. NSE is highly expressed in neurogenesis (46). The upregulation of CNP, coding a myelin-associated enzyme, is one of the
earliest events in oligodendrocyte differentiation (47). SNAIL is implicated in neural crest formation (from where most of the dental organ
originated, including SCAP) and in neuronal differentiation (48). Moreover, GDNF and NT3 are neurotrophic factors that play an important
role in the survival of neural populations in the central nervous system
by inhibiting apoptosis and mediating axonal growth (49, 50). The
hypoxic influence was also evaluated on SCAP grown in neurodifferentiation medium. NSE, VEGFB, and GDNF were significantly
up-regulated compared with normoxia. In addition, SCAP developed
a neuronlike morphology with angle-shaped cell bodies and neurites
that could play the role of dendrites or axons. Some cells were also connected to each other, possibly via synapses. Thus, SCAP have some potential to differentiate into de novo neurons that could reconnect with
the nerve tract beyond the tooth apex. This axonal targeting could
also be mediated by the hypoxia-induced production of neurotrophic
factors (GDNF and NT3). This may explain the recovery of vital pulp responses reported in a certain proportion of regenerative endodontic
cases (4). Regarding nondental applications, functional nerve recovery
should be evaluated through electrophysiological tests (51). SCAP also
presented a decreased proliferation rate, which is a sign of neurodifferentiation. In hypoxia, PanNF and NeuN expressions were 2- and
100-fold higher, respectively, compared with normoxia although the
number of positive cells was not different. Thus, hypoxia seems to be
very favorable for neuronal differentiation.
Taken together, our results showed that hypoxia induced an upregulation of neuro- and osteospecific genes and proangiogenic factor
(VEGFa) in SCAP, meaning that the cellular microenvironment is playing
a major role in their fate. A recent review highlighted the role of stem
cells and the microenvironment in regenerative endodontics (52). It reported the importance of cytokines and growth factors, adhesion molecules, and the extracellular matrix on stem cells during regeneration of
the dentin-pulp complex. In the current study, hypoxic conditions present in root canals devoid of adequate circulation are likely modulating
stem cell fate. Moreover, in the context of regenerative strategies, SCAP
appear as an important stem cell population because they respond with
desirable pro-osteogenic, neurogenic, and angiogenic functions when
placed in the root canal hypoxic environment.
JOE — Volume 40, Number 9, September 2014
Conclusion
This study shows a clear impact of hypoxic culture conditions on
the differentiation potential of SCAP. In particular, the up-regulation of
neuro- and osteospecific genes and the proangiogenic factor in SCAP
cultured in basal medium supports the potential of SCAP to promote
pulp-dentin regeneration. Hypoxia was also particularly favorable for
neurodifferentiation, which is promising for neuroregenerative events,
in regenerative endodontic procedures as well as in other applications,
such as nerve regeneration.
Acknowledgments
Julian G. Leprince and Anne des Rieux equally contributed to
the work.
The authors deny any conflicts of interest related to this study.
Supplementary Material
Supplementary material associated with this article can be
found in the online version www.jendodn.com (http://dx.doi.org/
10.1016/j.joen.2014.04.008).
References
1. Iwaya SI, Ikawa M, Kubota M. Revascularization of an immature permanent tooth
with apical periodontitis and sinus tract. Dent Traumatol 2001;17:185–7.
2. Banchs F, Trope M. Revascularization of immature permanent teeth with apical periodontitis: new treatment protocol? J Endod 2004;30:196–200.
3. Ruparel NB, Teixeira FB, Ferraz CC, Diogenes A. Direct effect of intracanal medicaments on survival of stem cells of the apical papilla. J Endod 2012;38:1372–5.
4. Diogenes A, Henry MA, Teixeira FB, Hargreaves KM. An update on clinical regenerative endodontics. Endod Topics 2013;28:2–23.
5. Lovelace TW, Henry MA, Hargreaves KM, Diogenes A. Evaluation of the delivery of
mesenchymal stem cells into the root canal space of necrotic immature teeth after
clinical regenerative endodontic procedure. J Endod 2011;37:133–8.
6. Jeeruphan T, Jantarat J, Yanpiset K, et al. Mahidol study 1: comparison of radiographic and survival outcomes of immature teeth treated with either regenerative
endodontic or apexification methods: a retrospective study. J Endod 2012;38:
1330–6.
7. Hargreaves KM, Diogenes A, Teixeira FB. Treatment options: biological basis of
regenerative endodontic procedures. J Endod 2013;39(3 suppl):S30–43.
8. da Silva LA, Nelson-Filho P, da Silva RA, et al. Revascularization and periapical repair
after endodontic treatment using apical negative pressure irrigation versus conventional irrigation plus triantibiotic intracanal dressing in dogs’ teeth with apical periodontitis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2010;109:779–87.
9. Yamauchi N, Nagaoka H, Yamauchi S, et al. Immunohistological characterization of
newly formed tissues after regenerative procedure in immature dog teeth. J Endod
2011;37:1636–41.
10. Rosa V, Zhang Z, Grande RH, N€or JE. Dental pulp tissue engineering in full-length
human root canals. J Dent Res 2013;92:970–5.
11. Iohara K, Imabayashi K, Ishizaka R, et al. Complete pulp regeneration after pulpectomy by transplantation of CD105+ stem cells with stromal cell-derived factor-1.
Tissue Eng Part A 2011;17:1911–20.
12. Shimizu E, Ricucci D, Albert J, et al. Clinical, radiographic, and histological observation of a human immature permanent tooth with chronic apical abscess after revitalization treatment. J Endod 2013;39:1078–83.
13. Tobias Duarte PC, Gomes-Filho JE, Ervolino E, et al. Histopathological condition of
the remaining tissues after endodontic infection of rat immature teeth. J Endod
2014;40:538–42.
14. Iida K, Takeda-Kawaguchi T, Tezuka Y, et al. Hypoxia enhances colony formation
and proliferation but inhibits differentiation of human dental pulp cells. Arch
Oral Biol 2010;55:648–54.
15. Sakdee JB, White RR, Pagonis TC, Hauschka PV. Hypoxia-amplified proliferation of
human dental pulp cells. J Endod 2009;35:818–23.
16. Gong QM, Quan JJ, Jiang HW, Ling JQ. Regulation of the stromal cell-derived factor1alpha-CXCR4 axis in human dental pulp cells. J Endod 2010;36:1499–503.
17. Aranha AM, Zhang Z, Neiva KG, et al. Hypoxia enhances the angiogenic potential of
human dental pulp cells. J Endod 2010;36:1633–7.
18. Ruparel NB, de Almeida JFA, Henry MA, Diogenes A. Characterization of a stem cell
of apical papilla cell line: effect of passage on cellular phenotype. J Endod 2013;39:
357–63.
SCAP Differentiation in Hypoxia
1417
Basic Research—Biology
19. Abe S, Hamada K, Miura M, Yamaguchi S. Neural crest stem cell property of apical
pulp cells derived from human developing tooth. Cell Biol Int 2012;36:927–36.
20. Sitkovsky M, Lukashev D. Regulation of immune cells by local-tissue oxygen tension:
HIF1 alpha and adenosine receptors. Nat Rev Immunol 2005;5:712–21.
21. Haque N, Rahman MT, Abu Kasim NH, Alabsi AM. Hypoxic culture conditions as a
solution for mesenchymal stem cell based regenerative therapy. ScientificWorldJournal 2013 Aug 27;2013:632972. http://dx.doi.org/10.1155/2013/632972.
22. Agata H, Kagami H, Watanabe N, Ueda M. Effect of ischemic culture conditions on the
survival and differentiation of porcine dental pulp-derived cells. Differentiation
2008;76:981–93.
23. Kanafi MM, Ramesh A, Gupta PK, Bhonde RR. Influence of hypoxia, high glucose,
and low serum on the growth kinetics of mesenchymal stem cells from deciduous
and permanent teeth. Cells Tissues Organs 2013;198:198–208.
24. Ambrosini G, Adida C, Altieri DC. A novel anti-apoptosis gene, survivin, expressed in
cancer and lymphoma. Nat Med 1997;3:917–21.
25. Filion TM, Qiao M, Ghule PN, et al. Survival responses of human embryonic stem
cells to DNA damage. J Cell Physiol 2009;220:586–92.
26. Coumoul X, Li W, Wang RH, Deng C. Inducible suppression of Fgfr2 and Survivin in
ES cells using a combination of the RNA interference (RNAi) and the Cre-LoxP
system. Nucleic Acids Res 2004;32:e85.
27. Gheisari Y, Soleimani M, Zeinali S, et al. Isolation of stem cells from adult rat kidneys. Biocell 2009;33:33–8.
28. Bai H, Ge S, Lu J, et al. Hypoxia inducible factor-1a-mediated activation of survivin
in cervical cancer cells. J Obstet Gynaecol Res 2013;39:555–63.
29. Schubert T, Xhema D, Veriter S, et al. The enhanced performance of bone allografts
using osteogenic-differentiated adipose-derived mesenchymal stem cells. Biomaterials 2011;32:8880–91.
30. Ruiz de Almodovar C, Lambrechts D, Mazzone M, Carmeliet P. Role and therapeutic
potential of VEGF in the nervous system. Physiol Rev 2009;89:607–48.
31. Matsushita K, Motani R, Sakuta T, et al. The role of vascular endothelial growth
factor in human dental pulp cells: induction of chemotaxis, proliferation, and differentiation and activation of the AP-1-dependent signaling pathway. J Dent Res
2000;79:1596–603.
32. D’Alimonte I, Nargi E, Mastrangelo F, et al. Vascular endothelial growth factor enhances in vitro proliferation and osteogenic differentiation of human dental pulp
stem cells. J Biol Regul Homeost Agents 2011;25:57–69.
33. Falk T, Yue X, Zhang S, et al. Vascular endothelial growth factor-B is neuroprotective
in an in vivo rat model of Parkinson’s disease. Neurosci Lett 2011;496:43–7.
34. Sun Y, Jin K, Childs JT, et al. Vascular endothelial growth factor-B (VEGFB) stimulates neurogenesis: evidence from knockout mice and growth factor administration.
Dev Biol 2006;289:329–35.
35. Leprince JG, Zeitlin BD, Tolar M, Peters OA. Interactions between immune system
and mesenchymal stem cells in dental pulp and periapical tissues. Int Endod J
2012;45:689–701.
1418
Vanacker et al.
36. Howard C, Murray PE, Namerow KN. Dental pulp stem cell migration. J Endod 2010;
36:1963–6.
37. Sant’Ana AC, Marques MM, Barroso TE, et al. Effects of TGF-beta1, PDGF-BB, and
IGF-1 on the rate of proliferation and adhesion of a periodontal ligament cell lineage
in vitro. J Periodontol 2007;78:2007–17.
38. He H, Yu J, Liu Y, et al. Effects of FGF2 and TGFbeta1 on the differentiation of human
dental pulp stem cells in vitro. Cell Biol Int 2008;32:827–34.
39. Zanini M, Sautier JM, Berdal A, Simon S. Biodentine induces immortalized murine
pulp cell differentiation into odontoblast-like cells and stimulates biomineralization.
J Endod 2012;38:1220–6.
40. Graham L, Cooper PR, Cassidy N, et al. The effect of calcium hydroxide on solubilisation of bio-active dentine matrix components. Biomaterials 2006;27:2865–73.
41. Zhang Y, Marsboom G, Toth PT, Rehman J. Mitochondrial respiration regulates adipogenic differentiation of human mesenchymal stem cells. PLoS One 2013;8:e77077.
42. Boyette LB, Creasey OA, Guzik L, et al. Human bone marrow-derived mesenchymal
stem cells display enhanced clonogenicity but impaired differentiation with hypoxic
preconditioning. Stem Cells Transl Med 2014;3:241–54.
43. Sonoyama W, Liu Y, Yamaza T, et al. Characterization of the apical papilla and its
residing stem cells from human immature permanent teeth: a pilot study.
J Endod 2008;34:166–71.
44. Dietz V, Harkema SJ. Locomotor activity in spinal cord-injured persons. J Appl
Physiol 2004;96:1954–60.
45. Sakai K, Yamamoto A, Matsubara K, et al. Human dental pulp-derived stem cells promote locomotor recovery after complete transection of the rat spinal cord by multiple neuro-regenerative mechanisms. J Clin Invest 2012;122:80–90.
46. Cooper EH. Neuron specific enolase: a marker of (small cell) cancers of neuronal
and neuroendocrine origin. Biomed Pharmacother 1985;39:165–6.
47. Wang SZ, Dulin J, Wu H, et al. An oligodendrocyte-specific zinc-finger transcription
regulator cooperates with Olig2 to promote oligodendrocyte differentiation. Development 2006;133:3389–98.
48. Nieto MA. The snail superfamily of zinc-finger transcription factors. Nat Rev Mol Cell
Biol 2002;3:155–66.
49. Clarkson ED, Zawada WM, Freed CR. GDNF improves survival and reduces apoptosis
in human embryonic dopaminergic neurons in vitro. Cell Tissue Res 1997;289:
207–10.
50. Kusano K, Enomoto M, Hirai T, et al. Transplanted neural progenitor cells expressing mutant NT3 promote myelination and partial hindlimb recovery in the chronic
phase after spinal cord injury. Biochem Biophys Res Commun 2010;393:812–7.
51. Barth L, S€utterlin R, Nenniger M, Vogt KE. Reduced synaptic activity in neuronal networks derived from embryonic stem cells of murine Rett syndrome model. Front
Cell Neurosci 2014;26:79.
52. Lin LM, Ricucci D, Huang GT. Regeneration of the dentine-pulp complex with revitalization/revascularization therapy: challenges and hopes. Int Endod J 2014;47:
713–24.
JOE — Volume 40, Number 9, September 2014