Bioengineering and Cell-derived Strategies for Salivary Gland Regeneration

概要

学位論文の要旨

Bioengineering and Cell-derived Strategies
for Salivary Gland Regeneration
唾液腺再生のためのバイオエンジニアリングと
無細胞化ストラテジー

名古屋大学・ルンド大学国際連携総合医学専攻
頭頸部・感覚器外科学講座 顎顔面外科学分野
（指導：日比 英晴
董 娇

教授）

【Introduction】
A physiologically balanced oral cavity is the foundation to support multiple basic and vital
human activities such as breathing, digestion, chewing, and speech. These vital functions are
critically supported by an aqueous and protein secretion mixture call ed saliva covering the
entire surface of the oral, which is produced by the salivary gland (SG). SG are organs that
anatomically and physiologically composed of acinar, ductal, and myoepithelial cells. Acinar
secretory cells have a central lumen into which proteins and fluid are secreted, while ductal
cells form tubular conduits to deliver it into mouth opening under innervation stimulation.
Daily around 1.0 – 1.5 L of saliva are delivered to the mouth from the SGs and a less production
of saliva is a disease called dry mouth symptoms. It can have multiple reasons from medication
usages, salivary gland dysfunction after irradiation (IR) therapy and chemotherapy for head
and neck cancer patients, other systematic diseases like Sjögren's syndrome. These patien ts
experience burning sensations in mouth and have difficulty eating or drinking (dysphagia),
negatively impacting a person's quality of life and interfering with daily activities, which can
accompany social marginalization. Other symptoms include sticky saliva, and oral fungal
infections, sore throats, cracked lips, and, which all cause various life -impacting pathological
events that compromise both dental and general health. This condition affects 630,000 patients
globally. The common believed mechanism for salivary function continuing to decline is due
to inflammation and atrophy in salivary secretory units, then irreversible fibrosis occurs.
The current treatment is supportive therapy, including increased water intake, chewing gum,
using pharmaceutical therapies such as oral sialagogues pilocarpine and radioprotectors (like
reactive oxygen species scavengers), oral rinses or moisturizer gels, but they only provide
temporary relief and cannot stop this irreversible process. These years, extracellular vesi cles
(EVs) from stem cells and tissue bioengineering are also emerging as therapies. But alternative
therapies are highly sought after because there is a general lack of methodologies and models
which can be used to study the complexity of salivary gland damage and potential regeneration.
New ways to study salivary gland damage and potential repair are urgently needed to evaluate
potential therapies which can be administered to patients.
Research subject and hypothesis: As a whole, the objective of this thesis was to regenerate
the salivary glands function with cell-free therapy as well as tissue-engineering strategies for
modeling. We firstly evaluate whether cell-derived strategies (extracellular vesicles, EVs)
could be a potential new therapy to ameliorate salivary gland injury and restore function. So,
we established two animal models of salivary gland hypofunction to mimic the two main
salivary gland diseases in irradiation damage and autoimmune diseases ( Paper I & III).
Furthermore, we also explored the possibility of implementing an ex vivo organotypic
precision-cut salivary gland slices (PCSS) model for a diverse modeling and evaluating option
for salivary gland disease and function (Paper II). We finally pursue to examine the feasibility
of using sustainable tissue processing method without xylene and histological scoring methods

-1-

for a systematic assessment and classification of salivary gland’s tissue and diseases ( Paper IV).
【Material and Methods】
Firstly, we constructed an in vivo murine model of 25 Gy irradiation -induced salivary gland
damage model and a Sjögren’s syndrome model using non-obese diabetic (NOD) mice which
spontaneously develop SS-like pathologies.
We aim to evaluate the potential of human dental pulp stem cell (hDPSCs)-derived EVs.
In radiation model, EVs were injected 3x weekly via tail vein, beginning immediately after
irradiation. Salivary gland function was evaluated 18 days after irradiation using salivary gland
flow rate (SFR), gene expression (by qRT-PCR) and histopathology.
In NOD model, mice in the hDPSC-EVs group received 10 g hDPSC-EVs in 200 µl PBS
once a week for 14 weeks. Mice in the other group received the same volume of PBS as a
negative control. And also test salivary gland flow rate (SFR), gene expression (by qRT-PCR)
and histopathology.
Next, we tested different methods to generate PCSS using a vibratome and evaluated the
slices in terms of viability (by WST-1), gene expression (by qRT-PCR), secreted α-amylase
activity (by α-amylase assay kit) and histological/light sheet fluorescence microscopy (LSFM)
three-dimensional imaging.
In the end, for Paper IV, we use a previously established porcine model of ALI via
intratracheal and intravascular lipopolysaccharide (LPS) administration. Additionally, we
gather the higher number of samples generated from these large animal models, and we worked
to implement a more sustainable and greener histopathological workflow throughout the entire
process to do H&E staining. And by reviewing previously published papers about scoring lung
injury and also by carefully observing and summarizing the H&E staining images that we
produced in our lab with our wide samples, we developed and validated this histological
scoring system for lung acute injury in bigger animals.
【Results】
The results show that following irradiation, SFR decreased while senescence -associated βgalactosidase-positive cells (via immunofluorescence) and senescence -related genes and
secretory-phenotypes (e.g., p21 and MMP3 in qRT-PCR) increased. SFR was unchanged
following EV treatment, but senescence-associated genes and secretory-phenotypes decreased.
We also demonstrated that in an animal model of Sjögren's syndrome patients, which exhibit
dry mouth symptoms, that hDPSCs-EVs could inhibit acquisition of the senescent phenotype
in salivary gland epithelial cells and alleviate the loss of glandular function. EVs were also
found to perform these effects through an underlying immunomodulatory mechanis m. For
PCSS, we developed protocols to produce viable slices of different thicknesses which secreted
functional α-amylase over three days. Phenotypic salivary gland cell epithelial markers (e.g.,

-2-

Keratin 5 and Aquaporin 5) increased over time in PCSS (by qRT-PCR). We developed
workflows to perform LSFM 3D visualization in whole salivary glands as well as the PCSS
model.
【Discussion】
My thesis has focused on better understanding and developing new therapies for patients
that suffer from dry-mouth (clinically referred to as xerostomia, characterized by
compromised saliva flow). In particular, the main portion of my thesis has focused on
understanding xerostomia caused by the necessary radiotherapy for head and neck cancer
patients whereby irradiation passes through the salivary gland in order to irradiate the head
and/or neck tumor. In order to study this, I established a murine model of submandibular gland
damage via irradiation in Nagoya and examined potential molecular mechanisms underlying
the loss of salivary gland function. I first characterized the model with regard to salivary gland
function and cellular level changes via immunofluorescence and qRT -PCR for genes and
proteins known to be altered following irradiation. My work demonstrated that saliva flo w
rate (SFR) decreased after irradiation (confirming the efficacy of the model), and that there
was an increase in senescence-associated cells (via immunofluorescence) and transcriptional
assessment via qRT-PCR (p21). Next, I studied the potential therapeutic role of human dental
pulp stem cell-derived small extracellular vesicles (hDPSC-sEV) in the mouse model I
established for irradiation-induced damage (focused on submandibular gland damage) and
their potential mechanism. As a result of the sEV treatment, while SFR remained unchanged,
inflammatory cytokines and senescence-related gene expression were reduced. Furthermore,
sEVs reversed oxidative stress in submandibular cells. Taken together, this research
demonstrates that hDPSC-sEV could prevent irradiation-induced cellular senescence, thereby
helping to contribute to the development of a potential stem-cell-free therapy for effectively
treating salivary gland dysfunction that develops after radiotherapy for head and neck cancer.
(Paper I)
In the second part of my thesis, I pursued the possibility of advancing a precision -cut
salivary gland slices (PCSS) ex vivo model aiming at salivary gland disease modeling and in
the future, evaluating potential therapies (e.g., cell-derived). In order to do this, I tested
multiple different methods to generate PCSS tissue pieces using a vibratome (e.g., method of
agarose embedding, cutting bath medium, orientation of the organs, and culture media). Next,
I assessed their metabolic activity (WST-1) and cellular function over time (e.g., gene
expression via qRT-PCR), functional α-amylase secretory, and immunofluorescence confocal
imaging/light sheet fluorescence microscopy (LSFM) (Paper II). I have successfully
developed optimized workflows to generate PCSS tissue pieces and perform the above assays,
several of which have thus far not been reported in the literature (e.g., qRT -PCR for murine
PCSS and light sheet imaging for either mouse or human salivary gland tissue). I demonstrated

-3-

that PCSS are metabolically active and functional α-amylase is present for up to three days of
ex vivo culture. My results demonstrated how phenotypic salivary gland cellular components
(e.g., keratin 5 to detect ductal epithelia cells and aquaporin 5 to detect functional acinar
epithelial cells) change dynamically over time and how to analyze the morphology and
histology of whole salivary glands and PCSS in 3D visualization. Taken together, I developed
techniques to reliably generate viable murine PCSS tissue pieces which retain function . Also,
I developed new state-of-the-art analysis methods for assessing possible potential therapies
(e.g., gene or cell therapy or pharmacological agents). In the future, the PCSS model could
serve as a disease model which is more similar to the native sa livary gland architectures and
physiology and future work will also work to apply this to human tissue. ( Paper II)
In addition to irradiation-induced damage, xerostomia can be caused by other conditions
such as immunological disorders. Sjögren's syndrome (SS) is one such disorder that impairs
the function of the salivary and lacrimal glands and significantly reduces th e quality of life
for these patients. I collaborated on work with my colleagues in Nagoya on a murine model
of Sjögren’s syndrome which similarly evaluated the therapeutic potential of hDPSC -EVs for
these patients. Their data supports that administration of hDPSC-EVs could prevent reduction
in glandular function and confers resistance to disease-induced cellular senescence of salivary
gland epithelial cells (SGECs). (Paper III)
Finally, I contributed as a co-author on work from the Lund lab which sought to establish a
‘greener’ method of histology. As traditional histological processing is toxic, new histological
workflows have been developed by others previously to replace xylene with isopropanol in the
tissue-clearing step to dramatically reduce the overall consumption of xylene. In addition to
the reduction in chemical waste, this can also improve the workplace environment for
histopathologist's by reducing their exposure to xylene. I helped develop and validate
techniques for xylene-free histology of multiple organs, including salivary glands and
performed semi-quantitative scoring in animal models of tissue injury. Such approaches
overcome the disadvantages of subjective histological analysis and the huge challenges of
objectively comparing histological injuries. In my previous work in Nagoya, they only
described histology and there were no attempts at quantification. I have now been exposed to
new techniques which can be incorporated in histological evaluation of salivary gland
dysfunction and is one of several aspects and novel methodologies I have taken back to Nagoya
for future incorporation in their work. (Paper IV)
For Paper I and Paper III, whether EVs influence directly or indirectly need further
assessment. It also remains to be explored in detail which segment of microRNA inside EVs
carry influences organ function. So, doing RNA sequencing is necessary to screen targeted
segments of microRNAs from hDPSCs-EVs library accompanied by adenovirus transfection
technology. If combining large animal models such as dogs and minipigs, it will be much
suitable to simulate the long-term clinical scenarios of intravenous or in situ injections of EVs

-4-

derived from human cells. In addition, For Paper I, hope to study the long-term mechanism of
how submandibular gland responses to radiation as well.
In Paper II, several fundamental questions and research gaps must be addressed firstly.
PCSS tissue pieces are not produced with consistent precision or thickness in the current
workflow, which prevents us from performing more efficient quantitative studies, which can
be improved by using rabbit or human samples. Additionally, we do not yet know what the
best culture system for PCSS tissue pieces is, which affects our ability to assess whether this
model can be used to study senescence and how fibrosis occurs. We will therefore study the
DMEM/F12 medium in the future, which has been previously published, and we will compare
the effects of three media on PCSS ex vivo (McCoy, RPMI, DMEM/F12). The transcriptional
expression of four genes is insufficient to explain the changes in cell expression. Thus, we will
perform RNA sequencing in order to gain a general understanding of the changes that occur at
cellular and molecular levels during the ex vivo period of 3 -5 days in this model. Therefore,
we are extremely dependent on the extraction of sufficient high -quality RNA from PCSS
models with small tissue, which is also one of the reasons we put in many efforts into optimize
and validate the PCSS RNA extraction workflow. We will promote three-dimensional imaging
analysis methods as part of light-sheet fluorescence microscopy technology. The development
of a coded analysis will greatly expedite our work. Later antibody staining will also be
performed, which will require us to develop and optimize a protocol for antibody penetrating
on cleared tissues for light sheet microscopy. The antibody incubation period may need to be
extended to three days or longer, and a pump may be required to ensure constant circulation
of the antibody. As we gain a deeper understanding of the PCSS model, it will be used to
compare FLASH radiation therapy with traditional radiation therapy (FLASH radiotherapy is
the delivery of ultra-high dose rate radiation several orders of magnitude higher than wh at is
currently used in conventional clinical radiotherapy and has the potential to revolutionize the
future of cancer treatment). A variety of other applications of PCSS modelling will be opened
as a result of this. Human salivary glands can be used for PCSS models, which can then be
inoculated with salivary gland epidermoid carcinoma spheroids, then radiation therapy or EV
therapy can be tested for clinical scenarios.
For Paper IV, as outlined above, I intend to apply sustainable histology and histologica l
scoring system into the salivary gland research field evaluation and to further broaden its
application to salivary gland disease category classification and grading. In salivary gland
diseases, fibrosis is a very devastating end. By developing a tissue scoring system for this
condition, we would be able to gain a better understanding of how salivary gland pathology
transfers and outcomes occur.
【Conclusion】
In conclusion, hDPSCs-EVs reduced senescence of salivary gland epithelial cells in

-5-

both murine irradiation and Sjögren's syndrome models and is a promising future for
xerostomia patients. For the murine PCSS, we successfully established an executable
operating procedure at the methodological level to reliably generate viable and
functional murine PCSS and developed new state-of-the-art analytical methods (such as
LFSM 3D imaging and qRT-PCR) to increase the diversity of objective tools to evaluate
PCSS. Therefore, this work laid the foundation for the future application of other
therapies to the PCSS model (such as irradiation therapy or EVs therapy), including
drug screening or mechanism of injury study. At the same time, we developed a
sustainable histology process to reduce xylene utilization in histological processing for
salivary gland tissue processing and also used histologically scoring method to evaluate
salivary glands’ diseases. Therefore, this work has developed a set of in vitro and in
vivo experiments and methods to do advanced modelling and systematic diagnosis, thus
contributing a better salivary gland research and therapies evaluation (cell -free and
tissue engineering treatments) in the future.

-6-

Paper I

Figure 1
Identification of small extracellular vesicles (sEV) and their uptake by cells.
(A) Immunogold labeling of human dental pulp stem cell-derived sEV (hDPSCs-sEV) with CD9
antibody visualized by transmission electron microscopy. Scale bar: 100 nm. (B) hDPSCs-sEV
particle size distribution was measured. (C) Exosomal marker expression in sEV was determined by
western blotting, with α-tubulin expression as a reference. (D) Analysis of cellular uptake of sEV.
Hoechst 33258 was used to stain cell nuclei (blue). Scale bar: 20 µm.

-7-

Figure 2
Establishment of in vivo model.
(A) Overview of experimental design. (B) Gross visual observation of the submandibular gland
specimens. Scale bar: 1 cm (n = 4 per group, 1 image per sample). (C) Changes in the size of
submandibular glands at 18 days after IR (*P <0.05). (D) Histological evaluation was performed using
HE staining. (Scale bar, 20 µm) (40 x magnification). Black arrows depict enlarged cells with
hyperchromatic nuclei. (E) Immunohistochemistry was used to assess the expression of acinar
epithelial cells positive for AQP5 expression. Scale bar: 20 µm. (F) Percentage of AQP5-positive cells
(40 x magnification) in different groups (**P <0.01). (G) SFR was analyzed for each mouse (n = 4
per group, 1 data per sample) (**P <0.01). (H) RT-qPCR analysis of AQP5 expression. All data are
presented as the mean ± standard deviation (n = 4 per group, 3 replicates per group) (**P <0.01).

-8-

Figure 3
sEV's role in cellular senescence through protecting ductal epithelial cells.
(A) Senescence-associated (SA)-β-galactosidase-positive cells (green) were double-stained with Ecadherin (red) and Hoechst 33258 (blue) for immunofluorescence staining. Scale bar: 20 µm (n = 4
per group, 3 images per sample). (B) Percentage of SA-β-galactosidase-positive cells in the PBS and
sEV groups are depicted (n = 4 per group, 3 replicates per sample) (**P <0.01). All data are presented
as the mean ± standard deviation and analyzed by two independent examiners blindly. (C)
Immunofluorescence co-staining of SA-β-galactosidase (green) and pH2A.X (red). Scale bar: 20 µm.
(D) Number of pH2A.X and SA-β-galactosidase double-positive cells was calculated (**P < 0.01).
(E) Senescence-associated gene expression was analyzed at 18 days after IR (*P < 0.05, **P < 0.01).
(F) Immunofluorescence for AQP5 (red) and SA-β-galactosidase (green) expression at 18 days after
IR. Blue, Hoechst33258; white arrow, senescent cells that are not positive to AQP5. Scale bar: 20 µm.
(n = 4 per group, 3 images per sample). (G) Krt5 (red) and SA-β-galactosidase (green) immunostaining.
Scale bar: 20 µm.

-9-

Figure. 4.
Effect of sEV on oxidative stress in the epithelium of the submandibular glands.
All data represent the mean ±standard deviation from three independent experiments (**P <0.01).

- 10 -

Paper II

Figure 1.
Murine PCSS model retains metabolic activity initially after vibratome slicing.
(A) Schematic overview of the working flow to generate PCSS from murine salivary glands. Agarose
embedded to encapsulate mouse salivary glands and sectioned using vibratome to 200 µm thickness.
Additional details described in the Materials and Methods. (B) Freshly generated PCSS demonstrating
structural integrity. (C) Brightfield microscopy of freshly generated PCSS. (D) Overview of culturing
PCSS in McCoy and RPMI+β-mercaptoethanol medium (Scale bar: 500 µm) and metabolic activity
assessment (by WST-1 assay) immediately after vibratome sectioning. (E) Metabolic activity of
freshly generated slices via WST-1 assay. The optical density value (OD450 nm) of WST-1 that minus
the background blank control are measured immediately in two mediums. Data are presented as Box
and Whiskers. n = 5 mice, n = 3 for McCoy and RPMI+β-Mer medium respectively. The dotted line
marks the OD450 nm of the blank group without tissue or FBS.

- 11 -

Figure 2.
McCoy and RPMI+β-mercaptoethanol medium retain cellular metabolism of PCSS culturing
over five-days.
(A) Schematic overview of experimental setup for metabolic activity. (B) Analysis on cellular
metabolism were conducted by WST-1 in two medium comparisons (for McCoy medium, n = 8(24h,
48h), n = 5(72 h, 96 h); for RPMI+β-mercaptoethanol medium, n = 5(24 h - 96 h), n = 3(120 h)). Each dot
represents a single mouse where individual mice are color coded.

- 12 -

Figure 3.
PCSS secrete functional α-amylase up to 48 hours ex vivo.
(A) Schematic overview of experimental setup for protein secretory function. (B) Biological αamylase protein secretory responses of PCSS model during 48 hours of culturing were validated in
two groups of supernatants. One dot of measurements was conducted in at least duplicate. Data were
presented as mean, n = 2 mice. (C) α-amylase activity (mU/ml) of PCSS model was calculated
according to provided formular. Data are presented as Box and Whiskers (n = 2).

- 13 -

Figure 4.
Nanodrop detection confirms the A260/A280 purity of PCSS samples' RNA is qualified.
(A) Spectrometric assessing the purity of RNA extracted from native salivary gland with and without
agarose following modified protocol using TRIzol. (B) Bubble plot of PCSS samples cultured 3 days
ex vivo. RPIM+β-mercaptoethanol medium (n=2 mice). McCoy medium (n=5 mice). (C) A260/A280
purity and (D) A260/A230 purity of RNA extracted from PCSS samples cultured three days in McCoy
and RPMI+β-mercaptoethanol mediums. Thick dashed line represents the median and thin dotted line
the quartiles. From B to E, each dot represents an individual RNA isolation.

- 14 -

Figure 5.
Low A260/A230 purity RNA isolated from PCSS is applied to qRT-PCR with acceptable primer
efficiency for the common housekeeping gene (mouse HPRT).
(A) The schematic workflow to extract RNA and make a series of dilution and run qRT-PCR after
cDNA synthesis. (B) PCSS-derived RNA samples with low A260/A230 and high A260/A230 purities are
applicable in transferring to cDNA and downstream qRT-PCT. (C) Corresponding melt curves for
each replicate mean in B. Ct, cycle threshold; RFU, relative fluorescence units. n = 1 independent
isolations from PCSS.

- 15 -

Figure 6.
qRT-PCR is feasible for PCSS and suggests epithelial cell-type changes and ECM remodeling
activated.
The PCR results of Aquaporin 5 (Aqp5), Keratin 5 (Krt5), Collagen 1a1, and Fibronectin of PCSS
model in three days’ culture in McCoy medium (McCoy medium: n = 5, each line represent one
mouse). Hprt (hypoxanthine-guanine phosphoribosyl transferase) was used as a reference gene and
counting ΔΔCt relative to Hprt = Ct (Target gene) - Ct (Hprt). Red and Green dots represent PCSS
prepared in different buffer systems when collecting and keeping organs before vibratome sectioning.
PBS: phosphate-buffered saline; PSS (physiological salt solution) contained (in mM) 140 NaCl, 5 KCl,
1 MgCl2, 1 CaCl2, and 10 Na-HEPES, 10 glucose, 0.8 thioureas, 0.4 ascorbic acid, pH 7 warmed to
37 ℃.

- 16 -

Figure 7.
Microscopy and immunohistology represented that PCSS in McCoy medium persistent
maintaining morphological integrity, cell-ECM interactions and innervation for section.
(A) A gross visual overview to observe 200 μm thickness of PCSS in whole structure by bright field
microscopic images taken in McCoy at serial time points. (B) Immunofluorescence staining of β-IVtubulin (green) and E-cadherin (red) was performed to evaluate the apical-basolateral polarity
architecture under neuro-innervation controls for PCSS after 24 hours and 48 hours of cultivation,
with comparable native salivary glands as a positive control in (C). The nucleus is stained by DAPI in
blue. Scale bar: 20 μm (20x magnification) (n = 2 per group, 2 images per sample).

- 17 -

Figure 8.
Using light sheet microscopy to image transparent salivary gland tissue after optical tissue
clearing.
(A) A schematic workflow of how to clear and fluorescently visualize salivary glands by light sheet
microscope. (B) Comparing uncleared and cleared salivary glands, submerging in DiBenzyl Ether
solution showed the best transparency clearing effect. (C) The visualization by light sheet microscopy
images by autofluorescence (maximum intensity projections) show tissue architecture in native
salivary gland and PCSS tissue pieces from mouse (24 hour) (Ex/Em: 480/520nm), with HE staining
from sustainable histology as a positive control (D).

- 18 -

Figure 9.
A series of irradiation exposure induced physiological responses related to cytotoxicity and
apoptosis alterations in PCSS.
(A) The schematic layout of experimental setups. (B) The measurement of lactate dehydrogenase
(LDH) level from PCSS tissue pieces in three days post-irradiation. LDH is an indicator of cytotoxicity
after cell injury. (C) Detection of the apoptosis pathway-related genes (p16, p21) was carried out on
the third day of cultivating PCSS (n=1 mouse, and n=4 biological replicates of PCSS per dose group).

- 19 -

Paper III

Figure 1.
Characterization of human dental pulp stem cell-derived extracellular vesicles (hDPSC-EVs).
(A) Particle size distribution of hDPSC-EVs evaluated using a Nanosight system. The EVs collected
by enrichment with phosphate-buffered saline were 50 to 150 nm in diameter. (B) Western blot
analysis of specific exosomal markers, CD9, CD63, and CD81. (C) Representative image of purified
EVs obtained by transmission electron microscopy. The EVs, labeled for CD9 antigen with gold
particles, exhibited a spherical shape; scale bar: 100 nm.

Figure 2.
Treatment with dental pulp stem cell-derived extracellular vesicles (DPSC-EVs) prevented the
decline of salivary gland function in NOD mice.
(A) Schematic representation of the experimental design for DPSC-EV therapy of NOD mouse. (B)
Salivary flow rate (SFR) was determined by volume of saliva/min/gm body weight. The SFR of
DPSC-EV group was significantly higher than that of phosphate-buffered saline (PBS) group at 22
weeks of age (n = 9). (C) Hematoxylin and eosin (H & E) -stained images of lymphocytic infiltration
in the submandibular glands. The bottom panels are high-resolution images of the black square in the
upper panel. Scale bar in the upper panel: 300 µm. Scale bar in the bottom pane: 100 µm. (D) Focus
score analysis (number of lymphocytic infiltrates/4 mm2) using H&E-stained sections. The score of
DPSC-EV group was significantly lower than that of PBS group (n = 9). (E) Tracking of PKH-26labeled DPSC-EVs in NOD mouse. PKH-26-labeled DPSC-EVs were observed in the submandibular
glands tissue of NOD mouse. Scale bar: 50 µm.

- 20 -

Figure 3.
Senescent cells were detected in the submandibular gland epithelial cells of NOD mice at 22
weeks of age, and treatment with dental pulp stem cell-derived extracellular vesicles (DPSCEVs) reduced the expression of senescent cell markers.
(A), (B) Sudan black B and senescence-associated β-galactosidase stain-positive cells, markers of
cellular senescence, were detected in salivary gland cells in the phosphate-buffered saline (PBS) group
but not in the ICR group or DPSC-EV group. Scale bar: 50 µm. (B) EpCAM-positive cells were
isolated from submandibular gland tissue and relative expression of key genes for salivary gland
function (AQP5), cellular senescence, and senescence-associated secreted phenotype (SASP) were
analyzed by quantitative RT-PCR. Gene expression levels in the EpCAM-positive cells were higher
than those in the PBS group and the control groups for cellular senescence and SASP genes. Y-axis
represents relative expression of the gene compared to that of GAPDH. Three experimental replicates
were used for each sample. *P < 0.05, **P < 0.01; ***P < 0.001, ****P < 0.0001. n= 3–6. All data
were presented as mean ± standard deviation (S.D.).

- 21 -

Figure 4.
Sphere-forming ability of the salivary gland epithelial cells in 22-week-old NOD mice was
reduced and cellular senescence occurred, while individual treatment with dental pulp stem cellderived extracellular vesicles (DPSC-EVs) improved forming ability and prevented cellular
senescence.
(A) EpCAM-positive cells were cultured as salispheres and observed in a bright field 10 days later.
Scale bar: 100 µm. (B) The number of spheres per 1 mm2 in three randomly determined locations was
significantly lower in the phosphate-buffered saline (PBS) group (n = 3). (C) Senescence-associated
β-galactosidase positively stained cells were detected in the spheres of the PBS group. Scale bar: 100
µm. (D) Relative expression of key genes for salivary gland function (AQP5) and cellular senescence
was analyzed by quantitative RT-PCR. Y-axis shows relative expression of the gene compared to that
of GAPDH. Three experimental replicates were used for each sample. *P < 0.05, **P < 0.01; ***P
< 0.001. n = 3. All data were presented as mean ± standard deviation (S.D.).

- 22 -

Paper IV

Figure 1
Spin Tissue Processor (STP 120) showing individual containers numbered 1-12 and the
programmable interface.

- 23 -

Figure 2.
Histopathology area, developed with revitalized, refurbished, and repurposed equipment.
A) Manual microtome (revitalized); B) Block heater (refurbished); C) Water bath (revitalized); D)
Cooling tray (repurposed); E) Paraffin tank dispenser (water boiler retrofitted with a stainlesssteel spigot).

Figure 3.
Tissue sectioning from biopsies prior to tissue processing.
A) Preparation for tissue sectioning; B) Sectioning of the tissue with a scalpel; C) Transfer of the
tissue sections into a pre-labeled cassette.

- 24 -

Figure 4.
Paraffin embedding of the tissue.
A) Pour molten paraffin into the pre-warmed mold; B) Transfer the specimens and orient in the
appropriate direction, considering that the blocks should be cut parallel to the base of the mold; C)
Place the bottom of the cassette on top of the mold filled with paraffin, ensuring that the paraffin fills
the grates in the cassette; D) Representative image of the process to cool down and solidify the
paraffin-infiltrated tissue and liquid paraffin on the pre-chilled cooling tray; E) Representative image
of solidified, paraffin-infiltrated tissue after cooling.

- 25 -

Figure 5.
Sectioning of the paraffin block.
A) Microtome set up for sectioning paraffin blocks; B) Sectioning of the tissue; C) Transfer of the
sectioned tissue with wrinkles to a water bath; D) Relaxation of the paraffin ribbon in the water bath
prior to mounting on microscope slides.

Figure 6.
Deparaffinization and H&E staining.
A) Example slide placed horizontally into the oven for paraffin removal overnight; B) Slide placed
into Coplin jar; C) Wash in running tap water, ensuring that the water flow is gentle enough to slowly
exchange the water without disturbing tissue sections on the slides; D) Stained and mounted slide.

- 26 -

Figure 7.
Representative examples of three different scoring sheets supplied to reviewers for the
histological scoring system.
Three different examples of the photomicrographs at three digital magnifications (4×, 10×, and 20×
to provide representative views across magnifications). The score sheet uses a modified version of
Table 1 on each page to collect information.

- 27 -

Figure 8.
Representative photomicrographs of H&E stained sections demonstrating the different features
selected for development of the lung tissue injury score system in low and high magnification.
Black arrowheads demonstrate examples of features.

- 28 -

Figure 9.
Scoring outcomes and relative reproducibility across different independent scorers with
different levels of experience (novice, moderate, and expert).
Representative total scores from the three example sheets shown in Figure 7 and transformed to a
scale of 0–100. The total score and the ability of users to distinguish between different severities of
injuries did not differ dramatically based on experience (see Note 17 for classification based on
experience). These samples were selected to represent a range of the scores our system detected (low,
mild, and extensive level of lung injury), as judged by the blinded scorers. See Figure 7 (score sheet
34 as an example for low level of lung injury, sheet 22 for mild level of injury, and sheet 24 for high
level of lung injury). The horizonal line represents the median of all scorers for each slide.

- 29 -

Figure 10.
Photomicrographs of sections stained with H&E demonstrating a range of scores across all
features.
Score ‘Minimum: absence of injury’; ‘Score 2/3’: mild injury; ‘Score 4’: moderate injury; ‘Score 5/6’:
pronounced injury; and ‘Maximum’: extensive damage and highest score given. Arrows demonstrate
examples of features.

- 30 -

Figure 11.
Example of H&E staining of porcine lung tissue using xylene-free tissue processing,

- 31 -