Tech Science Daily News

In the last decade cell therapy of limb ischemia has been extensively developed exploiting
different cell types and a variety of delivery methods. Still, the transplantation
procedure and its therapeutic efficacy remain issues despite feasible access to skeletal
muscle (compared to myocardium or brain) and sophisticated guidance and application
techniques we dispose at the moment 35]. Our work reports a rapid and reproducible method for generation of CS from ADSC
and suggests aprocedure for baculoviral expression of VEGF to enhance their angiogenic
efficacy.

Formation of CS requires optimal conditions partially depending on species due to
cell size differences between mammals (e.g., human ADSC are nearly three times as
big as mouse) and matrix synthesis rate. Our data confirmed the role of optimal seeding
density (Fig. 1a and b): formation of CS from mADSC was observed at 1.0-1.5?×?10
⁶
cells/well of a 12-well plate (?250-375?×?10
³
cells/cm
²
), which allowed the generation of a transplant with a surface of 0.8-1.0 cm
²
. The latter was smaller than the well’s bottom area due to surface tension-induced
retraction of CS. At lower densities formation of CS from mADSC may require up to
five days 24] and after seeding of??500?×?10
³
cells/well in a 12-well plate we incubated them for at least three days to obtain
a detachable structure (data not shown). Thus, both approaches led to generation of
CS from mADSC, but a higher seeding density allowed reduction of time required for
CS to form and become detachable. However, rapid generation relies on a large number
of cells, which is feasible for ADSC, but for other cell types, which are not available
in abundance or have lower proliferation rate, this can be an obstacle and turn attention
to long-term incubation.

Cell sheets from mADSC were detached using short-term trypsinization without loss
of integrity and comprised a solid structure of approximately 50–70 ?m thickness,
which accounts for three to six layers of cells (Fig. 1). Most studies in the field exploit commercial or in-house made thermoresponsive
culture dishes 36], 37] where detachment of CS occurs due to polymer coating that changes its hydrophilic
properties depending on temperature. We had to omit use of thermoresponsive dishes
because the protocol for BV transduction included 6 hrs of incubation at 27 °C, which
resulted in loss of adherence. This methodical hurdle was circumvented by application
of short-term (10–15 sec) trypsinization and also reduced the procedure’s cost by
using routine laboratory equipment to detach and manipulate the CS (Fig. 1 and Additional file 1: Video 1).

We established a protocol for effective BV-transduction of mADSC in HBSS with subsequent
treatment by sodium butyrate. Transduction of mammalian cells by BV is influenced
by a number of factors including temperature, presence of serum or other supplements
and transduction medium buffer system.

FACS-analysis of eGFP expression showed that BV-transduction in PBS or HBSS resulted
in up to 90 % efficacy after 6 hrs of incubation at MOI?=?150 (Fig. 2a). Our data support the use of low-NaHCO
₃
media for viral transduction of mADSC in the same manner as it was observed by Shen
et al. 32]. In our study use of ?-MEM and DMEM for viral stock dilution led to reduced transduction
efficacy, which correlated with higher NaHCO
₃
content in these media known to hinder BV transduction of mammalian cells. The negative
influence of NaHCO
₃
is attributed to its ability to reduce the amount of BV entering the cell without
significant influence on viral particles stability or binding to the cell membrane
32].

Cell survival after BV transduction was assessed using eGFP-expressing vector to avoid
mitogenic or pro-survival effects of VEGF165 and showed that use of HBSS is preferable
to PBS. Despite containing only 1.0 g/L D-glucose HBSS significantly attenuated toxicity
of NaBu added to the cells to boost protein production (Fig. 3). Thus, we suggest that even minimally nourished HBSS mitigated stress that mADSC
underwent during transduction and NaBu treatment so we used it in further procedures.

Expression of human VEGF165 in rodent ADSC using the Bac-FCVW/Bac-FLPo system has
showed its efficacy in a rabbit model of myocardial infarction with a prolonged (1 month)
expression period in vitro24] and minimal immunogenicity in vivo38]. In the current work this approach has been successfully applied for modification
of high-density CS from 1.0?×?10
⁶
mADSC. Our experiments suggest that maximum human VEGF165 output was achieved after
transduction by Bac-FCVW/Bac-FLPo MOI of 150/15 with subsequent NaBu treatment resulting
in VEGF165 content up to 25–27 ng/ml/10
⁵
cells (Fig. 4). Still, protocols resembling ours are not limited to application of BV for ADSC
– due to the mechanism of BV entry after certain tests it can be adapted for modification
of other cell types. BV-mediated transduction of chondrocytes 39], endothelial, cancer and epithelial cells 40] has been reported to result in significant expression of the delivered gene. During
development of new approaches, NaBu treatment can be omitted in case of prominent
deteriorating effects on cell survival yet in ADSC we managed to establish a “tempered”
protocol when application of appropriate medium (HBSS) and other transduction conditions
resulted in acceptable cell survival along with peak VEGF165 production.

The method of transplantation we used requires routine laboratory equipment for detachment
and transfer of CS and manipulations within the surgical wound were performed in a
drop of saline to reduce mechanical damage. High adherence of CS to skeletal muscle
allowed omitting any additional material (e.g., fibrin glue). Furthermore, the stickiness
of CS was supported by histology findings at day 14 that showed close contact of dye-labeled
CS with skeletal muscle (Fig. 6 and Additional file 2: Figure S1). Interestingly, in CS-covered muscle we observed a number of young myofibers
underlying the mADSC mass, which raises a question whether myogenesis has been supported
by CS or was induced by surgical and ischemic injury to the muscle.

The conventional method for ADSC delivery is injection of cells suspended in an isotonic
vehicle solution. Evaluation of dispersed mADSC engraftment after intramuscular injection
revealed that dye-labeled cells resided between myofibers at day 14 (Fig. 6). Labeled mADSC located within the injection site were numerous indicating engraftment
and survival, which is in accordance with previous findings in BALB-C nude mice 8].

In the described case, the putative advantage of CS besides transplantation of ADSC
with matrix proteins is the allocation of cells to the peripheral region of the muscle
(Fig. 5 and Additional file 2: Figure S1) where the impact of inflammation and ischemia is not so prominent. Indeed,
in examined muscle specimens the majority of viable tissue was located at the periphery,
whereas infiltration and necrosis occurred in the middle portion (Fig. 8). In our experiments with BALB-C mice suffering from extreme ischemia after surgery
due to weak collateral vascularization we found all central parts of the section to
consist of necrotic tissue in different stages of disruption 8]. Thus, we may speculate about a more “tranquil” environment for ADSC delivered on
the muscle surface compared to cells injected inside the ischemic tissue. However,
injection of dispersed cells is feasible in certain cases and can be easily adjusted
for cell number, injection count, and location and can be repeated over time.

Doppler measurements of limb perfusion showed that transplantation of mock-treated
CS led to a significant improvement of blood flow compared to untreated controls at
day 14 and we also found ADSC CS to be superior to injection of an equivalent dose
of cells in terms of limb perfusion (Fig. 5). Even higher perfusion was observed in BV-transduced mADSC expressing VEGF165 suggesting
that modification of CS improved its therapeutic efficacy and resulted in a better
functional outcome with relative perfusion reaching???60 %. To our surprise, we found
no significant improvement of limb perfusion in animals injected with ADSC despite
histology studies revealing engraftment of injected cells. In this case, a possible
reason underlying this discrepancy may be duration of follow-up period. In our previous
studies with injection of cell suspension to ischemic limbs we found maximum difference
between control and treated mice at 20–21 days of experiment. At this time point,
spontaneous reperfusion responsible for restoration of blood supply in untreated animals
was at the plateau while treated animals showed improvement 8], 29].

Overall, period of observation confines us to certain boundaries and we cannot make
claims about the inefficiency of injection delivery. The conclusion that can be drawn
from our study is that delivery of ADSC CS to ischemic limbs induces a faster perfusion
rise compared to dispersed ADSC. Still, further observation time points are required
to fully characterize the efficacy of both methods.

In limb ischemia models perfusion correlates with blood vessel density and we assessed
the influence of CS transplantation on capillary and arteriole counts. We found capillary-sized
vessels increased in numbers after transplantation of mock-transduced mADSC CS compared
to untreated controls (Fig. 7). Delivery of mADSC by injection showed no significant effect on capillary density
and was inferior to CS delivery. This finding was quite surprising as paracrine effects
of ADSC are known to induce angiogenesis within tissue, matrix, and other animal models
of vessel growth.

Maximum increase in CD31+ capillary counts was found in animals that received transplantation
of VEGF165-expressing CS, which may be accounted for production of angiogenic growth
factor known to induce sprouting and vascularization 8], 41]. In these specimens capillary counts were significantly higher than in mice that
received mock-transduced CS subcutaneously. Observed changes in capillary density
are classic evidence of angiogenesis stimulation, accepted since early work in this
field 42]. Still, we did not find any evidence for enhancement of arteriogenesis assessed by
?-SMA+ blood vessels in any study group (Fig. 7). This finding emphasizes importance of expanded observation time-frame as far as
stabilization of largerblood vessels may take up to three to four weeks in rodents
43].

Despite the obvious role of larger blood vessels for perfusion, capillaries are known
to be of crucial importance for tissue nutrition 44] and do have an impact on blood flow measured by Doppler-based methods 45]. Another mechanism of blood flow restoration relies on collateral formation/remodeling,
which may occur at the proximal part of the limb and cannot be evaluated using laser
Doppler.

Observed improvement of blood flow was expected to enhance muscle nutrition, which
prompted histological evaluation of the anterior tibia muscle (Fig. 8). ADSC CS group animals showed reduced necrosis and infiltration compared to untreated
controls, whereas mice that received VEGF-ADSC CS showed significant improvement compared
to untreated CS. Attenuation of necrosis is an important endpoint in experimental
8], 29], 46] and clinical studies on limb ischemia therapies 47], 48], thus, this piece of evidence obviously supported the efficacy of the developed technique.
Conventionally used injection of ADSC significantly reduced necrosis as well as muscle
infiltration (Fig. 8), which indicates that the antiapoptotic effect of this cell type takes place early
after delivery and may prevent necrosis and tissue disruption even when vascularization
has not yet been established. The latter statement relies on lack of significant changes
in perfusion and vessel density in ADSC injected animals. Still, it seems that when
the cells are injected and, thus, delivered to the epicenter of ischemia – the central
portion of the muscle – they manage to induce a paracrine-mediated response and reduce
the number of necrotic fibers.

Transplantation of cellular grafts may induce an immune response by the host and we
evaluated infiltration of CS by monocytes using immunohistochemical staining for CD68.
Monocytes (and their mature forms – macrophages and dendritic cells) are known to
play a pivotal role in resorption/phagocytosis, but also drive regeneration and control
inflammation 49], 50]. Moreover, despite that the stromal-vascular fraction may contain 10-15 % of the
monocytes 51], passaged ADSC are CD68-negative and, thus, evaluation of monocyte infiltration would
characterize host response to delivered CS. We found up to 25 % of cells within the
CS were CD68-positive (Fig. 9c-d) indicating their monocyte lineage and this figure was similar in both BV-treated
and mock-transduced constructs. This was an encouraging finding showing that treatment
of cells by BV using our procedure did not enhance the monocyte-mediated immune response
to transplanted cells. Moreover, VEGF165 itself may be characterized as a cytokine
with certain inflammation-driving modality 52], yet we found no evidence for increased monocyte invasion in VEGF-ADSC CS. Of course,
immune response to a graft may involve other cell types, especially CD4+ and CD8+
T-cells, but our previous data in a model of bone defect treated by BV-transduced
ADSC showed no significant difference in infiltration by these immune cells 38]. Overall, these findings require additional investigation in a more appropriate model
as far as use of inbred strain mice, which are close to the sibling state, does not
imitate allogeneic transplantation.

Fig. 9. Immunostaining for vascularization and monocyte invasion to transplanted CS. Images
were acquired from CS and skeletal muscle sections extracted at day 14 of the experiment
and stained for blood vessel and monocyte antigens. a Representative image of blood vessels within CS and corresponding H E staining.
b Representative image of larger blood vessels within transplanted CS (green arrow – ?-SMA-positive blood vessels; yellow arrow – ?-SMA-negative). c Microphotographs illustrating CD68+ monocyte infiltration (DAB-visualized) to mock-transduced
of VEGF-expressing CS. d Graphic presentation of monocyte counts within CS from corresponding study groups;
Mann–Whitney U test. CS cell sheet, H E hematoxylin and eosin, ?-SMA ?-smooth muscle actin, DAB diaminobenzidine, VEGF vascular endothelial growth factor

Retention of CS was found at days 7 and 14 and was accompanied by vascularization
of transplanted cell mass by capillary (some had visible lumen) and sporadic ?-SMA-positive
blood vessels in most specimens analyzed (Fig. 9). Formation of functional blood vessels in CS has been previously observed in constructs
from endothelial cells 53], 54] and cardiac primitive cells 55]. To our knowledge, this is the first report of blood vessel formation within non
pre-vascularized ADSC-based CS in a model of ischemic pathology. Moreover, the short-term
CS formation used in our protocol and use of non-angiogenic conditions (DMEM/10 %
FBS, normal O
₂
pressure, etc.) marginally excludes the possibility of spontaneous pre-vascularization
of CS during in vitro preparation. The question to be addressed is the origin of the CD31-positive endothelial
cells in the CS. Host blood vessel in-growth is the most obvious answer, while ADSC
are known to have a limited capacity for endothelial differentiation 56] or can contain a small fraction of endothelial progenitor cells that may contribute
to vessel formation 57], 58].

We also obtained evidence for limited (6-7 % of total number) proliferation of cells
within CS, which was not confined to transplanted mADSC. We found a number of CMFDA-negative
proliferating cells, which are likely to be dividing host cells that invaded CS during
engraftment (Fig. 10). As for apoptosis prevalence, it was approximately 10-12 % in both – mock-transduced
and VEGF165-expressing CS (Fig. 10). To explain lack of difference between mock-transduced and VEGF-expressing CS we
may refer to our published in vitro data showing VEGF165 production reduces 12–14 days after transduction, 24] so growth factor may not have had an impact on cell fate.

Fig. 10. Proliferation and apoptosis of cells within subcutaneously transplanted CS. CS were
stained by CMFDA and transplanted to mice subcutaneously. At day 14, CS adherent to
the femoral quadriceps muscle was extracted and stained for Ki-67 and cleaved caspase
3. Upper row left and middle – CMFDA-positive (red arrows) and negative (yellow arrows) cells with Ki-67+ nuclei indicating proliferation. Magnification 400×. Lower row left and middle – representative images of cells positive for cleaved caspase 3 (white arrows). Magnification 630×. Right column graphs illustrate results of manual count of proliferating and apoptotic cells in
mock-transduced and VEGF-expressing CS; Mann–Whitney U test. CS cell sheet, CMFDA 5-chloromethylfluorescein diacetate, VEGF vascular endothelial growth factor

Observed vascularization of CS stressed in our study provides mechanistic support
for the efficacy of CS. We can propose that paracrine activity of ADSC, which is considered
to be a driver of its angiogenic and regenerative capacity 40], 59], may have an extended impact area due to growth factor uptake through formed vessels.
This may explain stimulation of angiogenesis and reduction of necrotic alterations
in the distal part of the limb we used for histology evaluation. Another point supporting
a paracrine mode of action in the CS-based protocol is increment of its efficacy after
VEGF165 expression in the absence of reduced apoptosis or increased proliferation
of transplanted mADSC. Because cell secretome was the only parameter manipulated in
this case, it is very likely that growth factors and cytokines could be the main mediators
of efficacy.

Subcutaneous implantation of modified or xenogenic CS has been utilized for creating
cellular factories producing FVIII in hemophiliac mice 54] or insulin in diabetic SCID mice 60], 61]. In these works the authors observed a significant improvement of blood clotting
(with FVIII appearance) and euglycemia (with xenogenic C-peptide detectable) suggesting
systemic uptake of secreted factors from CS.

On the one hand, the abovementioned results provide support for the possible concept
of uptake of proteins secreted by CS and, thus, may explain the efficacy of distant
transplantation of CS. On the other hand, for VEGF165 this could be considered a side
effect due to its tumor-activating potential and possible influence on pathologic
angiogenesis in retina, skin, etc. 62]. We performed ELISA of plasma from mice at day 7 post-transplantation of VEGF-expressing
CS and did not detect systemic circulation of human VEGF165.

The preliminary format of this research has also revealed certain limitations listed
below and hinting at further study directions:

1) host/donor origin of vascular cells within CS – the ADSC population mainly contains
cells that carry CD31 or may undergo endothelial differentiation 63]; this requires clarification. This point can be addressed using a sex mismatch approach
in further studies. Our data also indicate that in angiogenesis models even CD31-negative
ADSC may incorporate (unpublished) or locate adjacent to vascular structures acting
as pericytes 63];

2) animal study duration – chosen terms of 14 days were optimal for early stage evaluation
of necrosis and CS retention, yet extension of the experiment’s time-frame would expand
our knowledge of both cell fate and CS effects on perfusion values and vessel density;

3) syngeneic transplantation applied in the study is a widely used model for autologous-based
procedures in recent years. However, allogeneic transplantation is easier to scale-up
and, thus, properties of allogeneic CS may be of interest for development of clinically
relevant procedures. These studies would also provide more valuable data on graft-host
interaction;

4) ADSC grafting is donor-dependent so it could of interest to evaluate the influence
of aging, chronic diseases and other factors on CS formation and their regenerative
properties 64].