- Email: [email protected]
Alginate microparticles loaded with basic fibroblast growth factor induce tissue coverage in a rat model of myelomeningocele
Accepted Manuscript Alginate Microparticles Loaded with Basic Fibroblast Growth Factor Induce Tissue Coverage in a Rat Model of Myelomeningocele
James S Farrelly, Anthony H Bianchi, Adele S Ricciardi, Gina Buzzelli, Samantha L Ahle, Mollie R Freedman-Weiss, Valerie L Luks, W Mark Saltzman, David H Stitelman PII: DOI: Reference:
S0022-3468(18)30649-3 doi:10.1016/j.jpedsurg.2018.10.031 YJPSU 58879
To appear in:
Journal of Pediatric Surgery
Received date: Accepted date:
23 September 2018 1 October 2018
Please cite this article as: James S Farrelly, Anthony H Bianchi, Adele S Ricciardi, Gina Buzzelli, Samantha L Ahle, Mollie R Freedman-Weiss, Valerie L Luks, W Mark Saltzman, David H Stitelman , Alginate Microparticles Loaded with Basic Fibroblast Growth Factor Induce Tissue Coverage in a Rat Model of Myelomeningocele. Yjpsu (2018), doi:10.1016/ j.jpedsurg.2018.10.031
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Title: Alginate microparticles loaded with basic fibroblast growth factor induce tissue coverage in a rat model of myelomeningocele
1
2
12
2
1
Authors: James S Farrelly , Anthony H Bianchi , Adele S Ricciardi , Gina Buzzelli , Samantha L Ahle , 1
1
2
1
T
Mollie R Freedman-Weiss , Valerie L Luks , W Mark Saltzman , and David H Stitelman
Department of Surgery, Yale University School of Medicine, New Haven, Connecticut, USA
2
Department of Biomedical Engineering, School of Engineering and Applied Science, Yale University,
CR
IP
1
US
New Haven, Connecticut, USA
Corresponding Author:
AN
James S Farrelly Yale University School of Medicine
M
Department of Surgery
ED
330 Cedar St, FMB 107 New Have, CT 06510
PT
Telephone: 203-843-4957
CE
E-Mail: [email protected]
Author Contribution:
AC
Study conception and design: JS Farrelly, AH Bianchi, AS Ricciardi, WM Saltzman, DH Stitelman Acquisition of data: JS Farrelly, AH Bianchi, G Buzzelli, AS Ricciardi, SL Ahle, MR Freedman-Weiss, VL Luks
Analysis and interpretation of data: JS Farrelly, AH Bianchi, G Buzzelli Drafting of manuscript: JS Farrelly, DH Stitelman Critical revision of manuscript: AH Bianchi, AS Ricciardi, G Buzzelli, SL Ahle, MR Freedman-Weiss, VL Luks, WM Saltzman, DH Stitelman
ACCEPTED MANUSCRIPT How this paper advances the field: Using a pre-existing drug-induced rat model of myelomeningocele, this work investigates minimally invasive, bioengineered particle therapy as a novel way to treat and potentially cure open human spina bifida in utero.
Abstract
IP
treatment of myelomeningocele (MMC) in an established rat model.
T
Background/Purpose: We sought to develop a minimally invasive intra-amniotic therapy for prenatal
CR
Methods: Time-dated pregnant rats were gavage-fed retinoic acid to induce MMC. Groups received intraamniotic injections at E17.5 with alginate particles loaded with fluorescent dye, basic fibroblast growth
US
factor (Alg-HSA-bFGF), fluorescently tagged albumin (Alginate-BSA-TR), free bFGF, blank alginate particles (Alg-Blank), or PBS. Groups were analyzed at 3 hours for specific particle binding or at term
AN
(E21) to determine MMC coverage.
Results: Alginate microparticles demonstrated robust binding to the MMC defect 3 hours after injection.
M
Of those specimens analyzed at E21, 150 of 239 fetuses (62.8%) were viable. Moreover, 18 of 61 (30%)
ED
treated with Alg-HSA-bFGF showed evidence of soft tissue coverage compared to 0 of 24 non-injected (P=0.0021), 0 of 13 PBS (P=0.0297), and 0 of 42 free bFGF (P= P<0.0001). Scaffolds of aggregated
PT
particles associated with disordered keratinized tissue were observed covering the defect in 2 of 18 (11%) Alg-BSA-TR and 3 of 19 (16%) Alg-Blank specimens.
CE
Conclusions: Injection of microparticles loaded with bFGF resulted in significant soft tissue coverage of the MMC defect compared to controls. Alginate microparticles without growth factors might result in
AC
scaffold development over the fetal MMC.
Keywords: Myelomeningocele; Spina bifida; Fetal therapy; Alginate microparticles; Particle therapy
Type of Study: Basic Science
Level of Evidence: N/A
2
ACCEPTED MANUSCRIPT Myelomeningocele (MMC), an open form of spina bifida, is both the most common and most severe form of neural tube defect associated with long-term survival. Patients with MMC suffer from a “two-hit” disease process, in which an abnormally developed spinal cord gets exposed to secondary toxic and mechanical injury throughout gestation. With no treatment, MMC patients usually have lower extremity paralysis with bowel and bladder dysfunction. Most patients also have an associated Chiari II
T
malformation resulting in hydrocephalus requiring ventriculoperitoneal (VP) shunt placement. Despite
IP
improved multi-disciplinary care, the neonatal mortality rate for MMC patients remains as high as 10%,
CR
and only approximately half of those who survive function independently as adults (1-7). Current options for management of MMC are termination, postnatal repair, or fetal surgery.
US
In a landmark prospective randomized-controlled trial comparing fetal surgery to postnatal MMC
AN
repair, investigators noted that prenatal myocutaneous flap closure of the MMC defect reduced the rate of VP shunt placement from 82% to 40% and doubled the number of patients who could walk independently
M
at 30 months from 21% to 42% (7). Though fetal surgery for MMC has resulted in significantly improved outcomes for some patients, it is expensive, somewhat inaccessible, resource intensive, and potentially
ED
very dangerous for both mother and fetus with a high rate of premature labor, among other serious complications (8, 9). Fetoscopic surgery, though less invasive, has also been associated with a higher
PT
rate of premature rupture of membranes and MMC repair dehiscence requiring postnatal revision (10).
CE
In this study, we explored the feasibility of minimally invasive needle-directed particle therapy for fetal treatment of MMC. In previously presented work, we have shown that intra-amniotic injection of
AC
various biodegradable and biocompatible particles is generally safe (11). Further, certain particles have been successfully used for encapsulation and controlled delivery of protein growth factor protein (12, 13). Therefore, we aimed to deliver biocompatible particles encapsulating basic fibroblast growth factor (bFGF) in utero to induce soft tissue growth over the MMC defect and exposed spinal cord. bFGF was selected due to prior success demonstrated by the Flake laboratory (14-17). Our goal in studying minimally invasive in utero particle therapy for MMC is to develop a safe and inexpensive therapy for spinal cord protection that can be offered to a larger portion of patients earlier in gestation than currently
3
ACCEPTED MANUSCRIPT available treatments. To our knowledge, this is the first study involving the intra-amniotic injection of biocompatible particles for the delivery of growth factors to the MMC defect.
1. Methods
T
1.1 Rat MMC Model
IP
All animal use was in accordance with the guidelines of the Animal Care and Use Committee of
CR
Yale University. Previous studies have established a reproducible MMC model in Sprague-Dawley rats (14-16, 18-22). Time-dated pregnant Sprague-Dawley rats from Charles River Laboratories (Wilmington,
US
MA) were gavage fed 40mg/kg of 97% pure all-trans retinoic acid (Across Organics; Morris Plains, NJ) dissolved in extra virgin olive oil (Whole Foods Market; Austin, TX) on gestation day E10.
AN
1.2 Poly(lactic-co-glycolic acid) (PLGA) Particle Fabrication
M
Poly(lactic-co-glycolic acid) (PLGA) nanoparticles (NPs) were loaded with coumarin-6 (C6)
ED
fluorescent green dye to track the delivery of NPs after intra-amniotic (IA) injection. PLGA nanoparticles encapsulating C6 dye were synthesized using a previously described single-emulsion solvent evaporation
PT
technique (23-25).
CE
1.3 Alginate Microparticle (MP) Fabrication and Characterization
Sterile alginate was purchased from NovaMatrix (Sandvika, Norway) and used in all experiments.
AC
Alginate microparticles (MPs) were fabricated using a previously described method (12). Alginate particles were loaded with recombinant human bFGF stabilized with human serum albumin (HSA), bovine serum albumin (BSA) conjugated to Texas Red® (BSA-TR) (ThermoFisher), or left blank. Once fabricated, the alginate particle pellet was then suspended in ultrapure water, aliquoted, flash frozen, and lyophilized. Goal protein concentrations were 1% BSA-Texas Red (w/w alginate) and 2.0% HSA-bFGF (w/w alginate). All particles were stored at -20°C until use. Particle size (hydrodynamic diameter) was measured using dynamic light scattering and confirmed with electron scanning microscopy. Surface charge (zeta potential) was measured in solution using a Malvern Zetasizer (Malvern Instruments, UK).
4
ACCEPTED MANUSCRIPT 1.5 Intra-amniotic Binding Studies and Fluorescent Photo Analysis
MMC was induced as described above. Control and retinoic acid-exposed dams between 17 and 21 days post-conception were anesthetized with inhaled isoflurane. The gravid uterus was exposed through a midline laparotomy incision. For the fluorescent binding studies, particles were re-suspended
T
by vortex and water bath sonication in PBS. Green fluorescent polystyrene MPs with (Polystyrene-COOH)
IP
and without (Polystyrene-plain) carboxyl surface groups and average diameter of 10 µm were purchased from Phosphorex (Hopkinton, MA). These polystyrene particles were used for injections without any
CR
modification. Surface charge (zeta potential) was measured in solution using a Malvern Zetasizer (Malvern Instruments, UK). All injections were performed with glass micropipettes (tip diameter ~60 µm)
US
and a pneumatic micro-injector (Narishige; Japan). Dams and pups were sacrificed 3 hours after
AN
injections were completed. Specimens were analyzed under a fluorescence microscope (Leica Microsystems, Wetzlar, Germany). Ex vivo imaging of MMC defects was performed on a Leica M80
M
fluorescence stereomicroscope (Wetzlar, Germany). Images were analyzed using Fiji imaging software (NIH website) (26). MMC defect-to-skin mean brightness ratios (MBR) were calculated using raw
ED
integrated pixel densities from image overlays of equal area (400 x 400 pixel squares). MBRs for all groups were compared using one-way analysis of variance (ANOVA) and Tukey multiple comparisons
PT
test.
CE
1.6 Intra-amniotic Alginate-HSA-bFGF MP injections
AC
MMC was induced as described above. Transuterine injections were performed on E17.5 under general isoflurane anesthesia, and groups were sacrificed at E21. Each individual fetal specimen was examined under a dissection stereomicroscope for the presence or absence of tissue covering the MMC defect. The experimental group was treated with alginate MPs loaded with bFGF and human serum albumin for stabilization (Alginate-HSA-bFGF). Control groups were treated with PBS, free bFGF, blank alginate (Alginate-Blank), alginate with fluorescently tagged albumin (Alginate-BSA-TR), or left uninjected. For all injections involving alginate particles, each fetus received 30 µL of 20 mg/mL alginate MPs suspended in PBS (approximately 9 µg bFGF per dose). For the arm treated with bFGF alone, bFGF
5
ACCEPTED MANUSCRIPT in PBS was received from the vendor (Creative Biomart, Shirley, New York) was diluted to match the expected concentration of bFGF in the Alginate-HSA-bFGF.
1.7 Histological Analysis
Specimens were fixed for 2 days in 4% (w/v) paraformaldehyde (Electron Microscopy Sciences;
T
Hartfield, PA) in PBS and transferred at room temperature in 70% (v/v) ethanol in distilled water for
IP
paraffin embedding and sectioning. Cut sections were stained with hematoxylin-eosin, pan cytokeratin
CR
AE1/AE3, or trichrome. The slides were imaged on a Zeiss Axio Scope light microscope (Carl Zeiss
US
Microscopy; Germany).
1.8 Statistical Analysis
AN
Defect-to-skin brightness ratio data for the fluorescence binding study was analyzed using ANOVA with Tukey multiple comparisons test. Again, an adjusted P value < 0.05 was considered
M
significant. Fisher’s exact test was used to analyze the results of the therapeutic growth factor injections.
ED
Statistical analyses were performed using GraphPad Prism (version 7; GraphPad Software; La Jolla, CA).
PT
2. Results
CE
2.1 Fluorescence MP Binding Study
All dams (N=6) and fetuses (n=57) injected with PLGA C6 NPs, Alginate-BSA-TR MPs, and
Figure 1
AC
green fluorescent polystyrene MPs were viable at the time of sacrifice 3 hours after injection. At 3 hours, particles can be seen adherent to the MMC defect (Figure 1). PLGA NPs (p=0.0461), alginate MPs (p<0.0001), and polystyrene-COOH MPs (p<0.0001) had significantly higher defect-to-skin
Figure 2
brightness ratios compared to controls (Figure 1 & 2). The most specific binding was achieved by polystyrene-COOH MPs [mean brightness ratio (MBR) = 2.139, 95% CI 1.768-2.510] and by alginate MPs (MBR=1.957, 95% CI 1.44–2.48) followed by PLGA NPs (MBR =1.61, 95% CI 1.34–1.88). Unmodified polystyrene MPs (Polystyrene-plain) had the weakest (p=0.9796) and least specific binding (MBR=1.18, 95% CI 1.07–1.30) compared to the non-injected controls (MBR=1.09, 95% CI 1.013-1.168) (Figure 2).
6
ACCEPTED MANUSCRIPT Alginate particles were approximately 4.752 +/- 42 microns in diameter with a zeta potential of -35.3 +/2.5 mV. The Polystyrene-COOH particles were approximately 10 microns in diameter with a zeta potential of -47 mV. Polystyrene-plain particles were 10 microns in diameter with a zeta potential of -40 mV.
2.2 Fetal delivery of Alginate-HSA-bFGF MPs and other controls
T
All dams that underwent laparotomy and IA injections were healthy at the time of sacrifice (N =
IP
22). 150 of 239 fetuses (62.8%) that received IA injections at E17.5 were viable at E21. Fluorescent
CR
stereomicroscopy of an E21 MMC specimen treated Alginate-BSA-TR (no growth factor) revealed red fluorescence throughout a layer of tissue over the MMC defect with negligible fluorescent signal from surrounding skin. On histological analysis of the same specimen, the red fluorescent layer appears to
Figure 3
US
Figure 4
be a scaffold of cellularized alginate associated with unorganized keratinized tissue (Figure 3). In all, 2
AN
of 18 (11%) specimens treated with Alginate-BSA-TR and 3 of 19 (16%) treated with AIginate-Blank exhibited similar keratinized tissue coverage. In the Alginate-HSA-bFGF treatment group, 18 of 61 (30%)
M
fetuses had evidence of partial soft tissue coverage compared to 0 of 24 un-injected (P=0.0021), 0 of 13
ED
PBS (P=0.0297), and 0 of 42 bFGF controls (P<0.0001). Histological analysis of Alginate-HSA-bFGF specimens positive for gross tissue coverage confirmed areas of skin-like soft tissue covering the spinal
PT
cord (Figure 4). The skin-like tissue over the MMC defect demonstrated strong staining for pan cytokeratin AE1/AE3 specifically in the area covering the spinal cord, but it had no noticeable hair
CE
follicles. The higher rate of defect coverage in the Alg-HSA-bFGF group failed to reach statistical significance when compared to the Alg-BSA-TR (P=0.3711) and Alginate-Blank (P=0.1364) groups;
AC
however, the histological characteristics of the positive Alg-HSA-bFGF specimens were qualitatively different from the abnormal, unorganized appearance of the Alg-BSA-TR and blank alginate controls.
3. Discussion
This report presents the results of preliminary investigations into a novel, minimally invasive, particle-based approach to the prenatal treatment of myelomeningocele (MMC). Prior published work has already proven the ability of nano- and microparticles to serve as controlled-release delivery vehicles for various molecules from chemotherapy drugs to DNA oligomers (12, 13, 24, 27, 28). Alginate particles
7 Figure 2
ACCEPTED MANUSCRIPT were chosen as the therapeutic vehicle of choice in this study because they are biocompatible, biodegradable, and more efficient than PLGA MPs at encapsulation and controlled delivery of growth factor proteins such as bFGF (12, 13, 29). In the initial fluorescence binding experiment, alginate MPs bound specifically to the MMC defect without any surface modification (Figure 2). To investigate why alginate particles preferentially bind to the MMC defect, fluorescent polystyrene MPs of known size and
T
surface characteristics (carboxyl groups) were tested. Given that the polystyrene-COOH particles with a
IP
more negative zeta potential bound more specifically to the defect than unmodified polystyrene particles,
CR
it is possible that either terminal carboxyl groups, negative zeta potential, or both characteristics contribute to stronger binding to the MMC defect. Of note, alginate MPs are rich in carboxyl groups and
AN
particles contribute to their robust and specific binding.
US
have a negative zeta potential, and therefore, it is possible that these inherent characteristics of alginate
With evidence that alginate MPs bind specifically to the MMC defect at E17.5, we studied both
M
the feasibility and the outcomes of intra-amniotic injection of Alginate-HSA-bFGF MPs. When MMC defects treated with Alginate-HSA-bFGF MPs were compared to untreated fetuses and fetuses treated
ED
with just PBS or free bFGF, a significant number of the fetuses treated with Alginate-HSA-bFGF had coverage of the exposed spinal cord tissue with skin-like tissue. The gross findings were confirmed on histology sections (Figure 4). The immunohistochemistry pan cytokeratin AE1/AE3 stain is a cocktail of
PT
Figure 4
antibodies against cytokeratin found in epithelial cells. In the skin, AE1/AE3 specifically stains the
CE
epidermis, eccrine glands, and folliculosebaceous-apocrine unit (30-32). In addition, the AE1/AE3 stain
AC
was used in a previously published study, which used bFGF-impregnated gelatin sponges to grow new skin-like tissue over the defect (14). Therefore, because the tissue covering the MMC defects in these Alginate-HSA-bFGF stained strongly on AE1/ AE3 sections, it is likely that it is rapidly forming dermal tissue. Moreover, none of the control specimens exhibited similar findings on gross or histological analysis. Of note, several of the specimens appeared to have a thick scaffold overlying the skin-like tissue over the spinal cord (Figure 4). The scaffold appeared to be composed of scattered red blood cells, some nucleated cells, and circular clear structures consistent with alginate particles. The overall structure seemed loosely connected (it washed away during the preparation of pan cytokeratin AE1/AE3 slides. This finding suggests that, after binding to the MMC defect, alginate MPs form a microscopic scaffold
8 Figure 3
ACCEPTED MANUSCRIPT through aggregation of the particles. This hypothesis is supported by the fact that 11-16% of fetuses treated with blank alginate MPs and Alginate-BSA-Texas Red also developed a scaffold and appear to have a partial ingrowth of hyperkeratotic skin underneath the scaffold despite the lack of any encapsulated growth factor (Figure 3).
T
One of the main limitations to this work is the variability in the shape and severity of the MMC
IP
defects in the rat MMC model. Despite the variable nature of the defects, we obtained approximately 90% yield of isolated MMC defects, and none of our uninjected or PBS-injected controls exhibited anything
CR
similar to the grossly apparent skin covering the MMC defect seen in positive experimental specimens. Because of the variability in the dimensions of the MMC defects and the preliminary nature of these
US
investigations, we opted to use a simple binary outcome measure to evaluate for MMC defect coverage.
AN
Partial soft tissue coverage was either present or absent. Our method for assessing the soft tissue coverage did not allow for accurate measurement of the fraction of the defects that were covered.
M
Similarly, although a complete watertight closure is the gold standard MMC repair needed to prevent cerebrospinal fluid leaks and infection, there was no definitive way to prove that our tissue coverings were
ED
watertight or absolutely complete (10). Despite these limitations, however, histology sections confirmed
underlying spinal cord.
PT
the dissection microscope findings and proved that specimens did indeed have soft tissue layers covering
CE
Because intra-amniotic alginate particle therapy for MMC is a novel procedure, there are many aspects of the methods and materials that require optimization. There are multiple technical challenges to
AC
the injections that must be mastered, and the procedures must be performed precisely and efficiently to prevent treatment variability and fetal loss. For example, because of the small size the amniotic space target, it is possible that particles were delivered inadvertently into the extraembryonic coelom, and this mistake would have theoretically resulted in a lower response rate. Ultrasound or microscope-guided injection can be used to ensure proper injection technique in the future. Another limitation to this study is the unknown safety profile of bFGF alginate particle intra-amniotic injections in humans or other animals. It is true that bFGF is a potent mitogen that stimulates the migration, proliferation, and differentiation of cells of mesenchymal and neurectodermal origin, such as keratinocytes, fibroblasts, melanocytes, and
9
ACCEPTED MANUSCRIPT endothelial cells (33). One major concern about intra-amniotic injections of particles delivering growth factors is inadvertent delivery to internal sites such as the gut or lung. In unpublished work from our laboratory, a survey of rat fetal organs during these timepoints revealed no significant particle infiltration of the lung, gut, liver, spleen, and inner dermis after intra-amniotic injection with either PLGA or alginate particles. In order to surpass regulatory barriers to treatment in humans, intra-amniotic alginate-bFGF
T
particle therapy will need to undergo more rigorous testing in successively higher animal models to
IP
establish a safety profile and to prove there are no deleterious teratogenic or carcinogenic effects.
CR
Future experiments will likely include higher growth factor doses and modification of the surface characteristics of the alginate MPs for more specific targeting, increased growth factor delivery to the
US
MMC defect, and increased particle scaffold aggregation. After this alginate particle treatment has been
AN
optimized for the rat MMC model, it can be applied to larger animal models, such as the ovine MMC model.
M
4. Conclusions
ED
Intra-amniotic injection of biocompatible PLGA and alginate particles for the treatment of MMC is feasible. Both PLGA and alginate bind preferentially to the MMC defect compared to the surrounding fetal
PT
skin. When injected into the amniotic space, alginate microparticles are capable of aggregating to form
CE
microscopic scaffolds over the MMC defect. When these particles are loaded with bFGF, 30% of cases resulted in significant soft tissue coverage of the MMC defect compared to controls. This line of therapy
AC
requires further optimization but could represent a broadly applicable minimally invasive adjunct to fetal therapy for MMC.
10
ACCEPTED MANUSCRIPT References
1.
Anglim B, Mandiwanza T, Miletin J, Turner M, Kennelly MM. The natural history of neural
tube defects in the setting of an Irish tertiary referral foetal medicine unit. J Obstet Gynaecol. 2016;36(1):19-23.
T
Davis BE, Daley CM, Shurtleff DB, Duguay S, Seidel K, Loeser JD, et al. Long-term
IP
2.
survival of individuals with myelomeningocele. Pediatr Neurosurg. 2005;41(4):186-91. Manning SM, Jennings R, Madsen JR. Pathophysiology, prevention, and potential
CR
3.
4.
US
treatment of neural tube defects. Ment Retard Dev Disabil Res Rev. 2000;6(1):6-14. Dillon CM, Davis BE, Duguay S, Seidel KD, Shurtleff DB. Longevity of patients born with
5.
AN
myelomeningocele. Eur J Pediatr Surg. 2000;10 Suppl 1:33-4.
Hunt GM. 'The median survival time in open spina bifida'. Dev Med Child Neurol.
Shurtleff DB, Luthy DA, Nyberg DA, Mack LA. The outcome of fetal myelomeningocele
ED
6.
M
1997;39(8):568.
brought to term. Eur J Pediatr Surg. 1994;4 Suppl 1:25-8. Adzick NS, Thom EA, Spong CY, Brock JW, 3rd, Burrows PK, Johnson MP, et al. A
PT
7.
randomized trial of prenatal versus postnatal repair of myelomeningocele. N Engl J Med.
Peranteau WH, Adzick NS. Prenatal surgery for myelomeningocele. Curr Opin Obstet
AC
8.
CE
2011;364(11):993-1004.
Gynecol. 2016;28(2):111-8. 9.
Saadai P, Farmer DL. Fetal surgery for myelomeningocele. Clin Perinatol.
2012;39(2):279-88. 10.
Kabagambe SK, Jensen GW, Chen YJ, Vanover MA, Farmer DL. Fetal Surgery for
Myelomeningocele: A Systematic Review and Meta-Analysis of Outcomes in Fetoscopic versus Open Repair. Fetal Diagn Ther. 2017.
11
ACCEPTED MANUSCRIPT 11.
Farrelly JS, Bianchi AH, Ricciardi AS, Ahle SL, Luks VL, Saltzman WM, Stitelman DH.
Intra-Amniotic Delivery of Biodegradable Microparticles and Nanoparticles to the Spinal Defect in a Rat Myelomeningocele Model. MOLECULAR THERAPY. 2017;25(5):322-3. 12.
Jay SM, Saltzman WM. Controlled delivery of VEGF via modulation of alginate
microparticle ionic crosslinking. J Control Release. 2009;134(1):26-34.
T
Jay SM, Shepherd BR, Andrejecsk JW, Kyriakides TR, Pober JS, Saltzman WM. Dual
IP
13.
vascularization. Biomaterials. 2010;31(11):3054-62.
Watanabe M, Jo J, Radu A, Kaneko M, Tabata Y, Flake AW. A tissue engineering
US
14.
CR
delivery of VEGF and MCP-1 to support endothelial cell transplantation for therapeutic
approach for prenatal closure of myelomeningocele with gelatin sponges incorporating basic
15.
AN
fibroblast growth factor. Tissue Eng Part A. 2010;16(5):1645-55. Watanabe M, Li H, Roybal J, Santore M, Radu A, Jo J, et al. A tissue engineering
M
approach for prenatal closure of myelomeningocele: comparison of gelatin sponge and
ED
microsphere scaffolds and bioactive protein coatings. Tissue Eng Part A. 2011;17(7-8):1099110.
Watanabe M, Kim AG, Flake AW. Tissue Engineering Strategies for Fetal
PT
16.
17.
CE
Myelomeningocele Repair in Animal Models. Fetal Diagn Ther. 2014. Watanabe M, Li H, Kim AG, Weilerstein A, Radu A, Davey M, et al. Complete tissue
AC
coverage achieved by scaffold-based tissue engineering in the fetal sheep model of Myelomeningocele. Biomaterials. 2015;76:133-43. 18.
Mackenzie TC, Flake AW. Human mesenchymal stem cells persist, demonstrate site-
specific multipotential differentiation, and are present in sites of wound healing and tissue regeneration after transplantation into fetal sheep. Blood Cells Mol Dis. 2001;27(3):601-4. 19.
Danzer E, Schwarz U, Wehrli S, Radu A, Adzick NS, Flake AW. Retinoic acid induced
myelomeningocele in fetal rats: characterization by histopathological analysis and magnetic resonance imaging. Exp Neurol. 2005;194(2):467-75.
12
ACCEPTED MANUSCRIPT 20.
Danzer E, Kiddoo DA, Redden RA, Robinson L, Radu A, Zderic SA, et al. Structural and
functional characterization of bladder smooth muscle in fetal rats with retinoic acid-induced myelomeningocele. Am J Physiol Renal Physiol. 2007;292(1):F197-206. 21.
Danzer E, Radu A, Robinson LE, Volpe MV, Adzick NS, Flake AW. Morphologic analysis
of the neuromuscular development of the anorectal unit in fetal rats with retinoic acid induced
IP
Danzer E, Zhang L, Radu A, Bebbington MW, Liechty KW, Adzick NS, et al. Amniotic
CR
22.
T
myelomeningocele. Neurosci Lett. 2008;430(2):157-62.
fluid levels of glial fibrillary acidic protein in fetal rats with retinoic acid induced
US
myelomeningocele: a potential marker for spinal cord injury. American journal of obstetrics and gynecology. 2011;204(2):178 e1-11.
Fahmy TM, Samstein RM, Harness CC, Mark Saltzman W. Surface modification of
AN
23.
biodegradable polyesters with fatty acid conjugates for improved drug targeting. Biomaterials.
McNeer NA, Chin JY, Schleifman EB, Fields RJ, Glazer PM, Saltzman WM.
ED
24.
M
2005;26(28):5727-36.
Nanoparticles deliver triplex-forming PNAs for site-specific genomic recombination in CD34+
Ricciardi AS, Bahal R, Farrelly JS, Quijano E, Bianchi AH, Luks VL, et al. In utero
CE
25.
PT
human hematopoietic progenitors. Mol Ther. 2011;19(1):172-80.
nanoparticle delivery for site-specific genome editing. Nat Commun. 2018;9(1):2481. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: an
AC
26.
open-source platform for biological-image analysis. Nat Methods. 2012;9(7):676-82. 27.
Mahoney MJ, Saltzman WM. Transplantation of brain cells assembled around a
programmable synthetic microenvironment. Nature biotechnology. 2001;19(10):934-9. 28.
Zhou J, Patel TR, Sirianni RW, Strohbehn G, Zheng MQ, Duong N, et al. Highly
penetrative, drug-loaded nanocarriers improve treatment of glioblastoma. Proceedings of the National Academy of Sciences of the United States of America. 2013;110(29):11751-6.
13
ACCEPTED MANUSCRIPT 29.
Jay SM, Shepherd BR, Bertram JP, Pober JS, Saltzman WM. Engineering of
multifunctional gels integrating highly efficient growth factor delivery with endothelial cell transplantation. FASEB J. 2008;22(8):2949-56. 30.
Wallace ML, Smoller BR. Immunohistochemistry in diagnostic dermatopathology. J Am
Acad Dermatol. 1996;34(2 Pt 1):163-83; quiz 84-6.
T
Folpe AL, Cooper K. Best practices in diagnostic immunohistochemistry: pleomorphic
IP
31.
32.
CR
cutaneous spindle cell tumors. Arch Pathol Lab Med. 2007;131(10):1517-24. Fuertes L, Santonja C, Kutzner H, Requena L. Immunohistochemistry in
US
dermatopathology: a review of the most commonly used antibodies (part I). Actas Dermosifiliogr. 2013;104(2):99-127.
Wu J, Ye J, Zhu J, Xiao Z, He C, Shi H, et al. Heparin-Based Coacervate of FGF2
AN
33.
Improves Dermal Regeneration by Asserting a Synergistic Role with Cell Proliferation and
M
Endogenous Facilitated VEGF for Cutaneous Wound Healing. Biomacromolecules.
AC
CE
PT
ED
2016;17(6):2168-77.
14
ACCEPTED MANUSCRIPT Figure Legends Figure 1. Left to right: E17.5 IA PLGA C6 (green dye); E17.5 IA Alginate-BSA-Texas Red™ (Alg-BSATR); E18 Polystyrene-COOH (green dye); E17.5 Polystyrene-plain (greed dye) can be seen bound to the MMC defect at 2.0x magnification Figure 2. Binding specificity of fluorescent particles to the MMC defect compared to adjacent skin.
T
Polystyrene-COOH IA injection resulted in the brightest defect-to-skin ratio which was higher than that of
IP
Polystyrene-plain (p<0.0001), PLGA-C6 (p=0.056) and the non-injected control (p<0.0001). The defect-to-
CR
skin ratio for Alginate-BSA-Texas Red was similar to that of Polystyrene-COOH and was also significantly higher than that of Polystyrene-plain (p=0.0006) and the non-injected control (p<0.0001).
US
Figure 3. Light stereomicroscope image of MMC defect (top left) treated with intra-amniotic injection of Alginate-BSA-Texas Red (Alg-BSA-TR) at E17.5. Scaffold over the MMC defect exhibited fluorescence
AN
under the red channel only, suggesting cellular incorporation of Texas Red tagged albumin still present 3 full days after injection. Histological analysis (bottom) confirmed the presence of alginate scaffold
M
overlying abnormal spinal cord with pan cytokeratin AE1/AE3 positive cells underneath the scaffold.
ED
Figure 4. Gross light images of control (top) and two selected specimens treated with Alg-HSA-bFGF (middle, bottom) along with their associated MMC defect histology sections stained with trichrome (middle
PT
column) and pan cytokeratin AE1/AE3 (right column). The heads of the specimens are located superiorly in the gross photos to the left, and the axial histology sections are oriented with the fetal back superiorly
CE
and vertebral bodies inferiorly. On the axial histology sections through the MMC defects, it is apparent that the specimen treated with Alg-HSA-bFGF has abnormal dermis that stains strongly for pan
AC
cytokeratin AE1/AE3 covering the spinal cords. Of note, the bottom specimen has an extra structure overlying the new skin representative of loosely adherent cellularized alginate scaffold that washed away during the pan cytokeratin stain preparation.
15
Figure 1
Figure 2
Figure 3
Figure 4
James S Farrelly, Anthony H Bianchi, Adele S Ricciardi, Gina Buzzelli, Samantha L Ahle, Mollie R Freedman-Weiss, Valerie L Luks, W Mark Saltzman, David H Stitelman PII: DOI: Reference:
S0022-3468(18)30649-3 doi:10.1016/j.jpedsurg.2018.10.031 YJPSU 58879
To appear in:
Journal of Pediatric Surgery
Received date: Accepted date:
23 September 2018 1 October 2018
Please cite this article as: James S Farrelly, Anthony H Bianchi, Adele S Ricciardi, Gina Buzzelli, Samantha L Ahle, Mollie R Freedman-Weiss, Valerie L Luks, W Mark Saltzman, David H Stitelman , Alginate Microparticles Loaded with Basic Fibroblast Growth Factor Induce Tissue Coverage in a Rat Model of Myelomeningocele. Yjpsu (2018), doi:10.1016/ j.jpedsurg.2018.10.031
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Title: Alginate microparticles loaded with basic fibroblast growth factor induce tissue coverage in a rat model of myelomeningocele
1
2
12
2
1
Authors: James S Farrelly , Anthony H Bianchi , Adele S Ricciardi , Gina Buzzelli , Samantha L Ahle , 1
1
2
1
T
Mollie R Freedman-Weiss , Valerie L Luks , W Mark Saltzman , and David H Stitelman
Department of Surgery, Yale University School of Medicine, New Haven, Connecticut, USA
2
Department of Biomedical Engineering, School of Engineering and Applied Science, Yale University,
CR
IP
1
US
New Haven, Connecticut, USA
Corresponding Author:
AN
James S Farrelly Yale University School of Medicine
M
Department of Surgery
ED
330 Cedar St, FMB 107 New Have, CT 06510
PT
Telephone: 203-843-4957
CE
E-Mail: [email protected]
Author Contribution:
AC
Study conception and design: JS Farrelly, AH Bianchi, AS Ricciardi, WM Saltzman, DH Stitelman Acquisition of data: JS Farrelly, AH Bianchi, G Buzzelli, AS Ricciardi, SL Ahle, MR Freedman-Weiss, VL Luks
Analysis and interpretation of data: JS Farrelly, AH Bianchi, G Buzzelli Drafting of manuscript: JS Farrelly, DH Stitelman Critical revision of manuscript: AH Bianchi, AS Ricciardi, G Buzzelli, SL Ahle, MR Freedman-Weiss, VL Luks, WM Saltzman, DH Stitelman
ACCEPTED MANUSCRIPT How this paper advances the field: Using a pre-existing drug-induced rat model of myelomeningocele, this work investigates minimally invasive, bioengineered particle therapy as a novel way to treat and potentially cure open human spina bifida in utero.
Abstract
IP
treatment of myelomeningocele (MMC) in an established rat model.
T
Background/Purpose: We sought to develop a minimally invasive intra-amniotic therapy for prenatal
CR
Methods: Time-dated pregnant rats were gavage-fed retinoic acid to induce MMC. Groups received intraamniotic injections at E17.5 with alginate particles loaded with fluorescent dye, basic fibroblast growth
US
factor (Alg-HSA-bFGF), fluorescently tagged albumin (Alginate-BSA-TR), free bFGF, blank alginate particles (Alg-Blank), or PBS. Groups were analyzed at 3 hours for specific particle binding or at term
AN
(E21) to determine MMC coverage.
Results: Alginate microparticles demonstrated robust binding to the MMC defect 3 hours after injection.
M
Of those specimens analyzed at E21, 150 of 239 fetuses (62.8%) were viable. Moreover, 18 of 61 (30%)
ED
treated with Alg-HSA-bFGF showed evidence of soft tissue coverage compared to 0 of 24 non-injected (P=0.0021), 0 of 13 PBS (P=0.0297), and 0 of 42 free bFGF (P= P<0.0001). Scaffolds of aggregated
PT
particles associated with disordered keratinized tissue were observed covering the defect in 2 of 18 (11%) Alg-BSA-TR and 3 of 19 (16%) Alg-Blank specimens.
CE
Conclusions: Injection of microparticles loaded with bFGF resulted in significant soft tissue coverage of the MMC defect compared to controls. Alginate microparticles without growth factors might result in
AC
scaffold development over the fetal MMC.
Keywords: Myelomeningocele; Spina bifida; Fetal therapy; Alginate microparticles; Particle therapy
Type of Study: Basic Science
Level of Evidence: N/A
2
ACCEPTED MANUSCRIPT Myelomeningocele (MMC), an open form of spina bifida, is both the most common and most severe form of neural tube defect associated with long-term survival. Patients with MMC suffer from a “two-hit” disease process, in which an abnormally developed spinal cord gets exposed to secondary toxic and mechanical injury throughout gestation. With no treatment, MMC patients usually have lower extremity paralysis with bowel and bladder dysfunction. Most patients also have an associated Chiari II
T
malformation resulting in hydrocephalus requiring ventriculoperitoneal (VP) shunt placement. Despite
IP
improved multi-disciplinary care, the neonatal mortality rate for MMC patients remains as high as 10%,
CR
and only approximately half of those who survive function independently as adults (1-7). Current options for management of MMC are termination, postnatal repair, or fetal surgery.
US
In a landmark prospective randomized-controlled trial comparing fetal surgery to postnatal MMC
AN
repair, investigators noted that prenatal myocutaneous flap closure of the MMC defect reduced the rate of VP shunt placement from 82% to 40% and doubled the number of patients who could walk independently
M
at 30 months from 21% to 42% (7). Though fetal surgery for MMC has resulted in significantly improved outcomes for some patients, it is expensive, somewhat inaccessible, resource intensive, and potentially
ED
very dangerous for both mother and fetus with a high rate of premature labor, among other serious complications (8, 9). Fetoscopic surgery, though less invasive, has also been associated with a higher
PT
rate of premature rupture of membranes and MMC repair dehiscence requiring postnatal revision (10).
CE
In this study, we explored the feasibility of minimally invasive needle-directed particle therapy for fetal treatment of MMC. In previously presented work, we have shown that intra-amniotic injection of
AC
various biodegradable and biocompatible particles is generally safe (11). Further, certain particles have been successfully used for encapsulation and controlled delivery of protein growth factor protein (12, 13). Therefore, we aimed to deliver biocompatible particles encapsulating basic fibroblast growth factor (bFGF) in utero to induce soft tissue growth over the MMC defect and exposed spinal cord. bFGF was selected due to prior success demonstrated by the Flake laboratory (14-17). Our goal in studying minimally invasive in utero particle therapy for MMC is to develop a safe and inexpensive therapy for spinal cord protection that can be offered to a larger portion of patients earlier in gestation than currently
3
ACCEPTED MANUSCRIPT available treatments. To our knowledge, this is the first study involving the intra-amniotic injection of biocompatible particles for the delivery of growth factors to the MMC defect.
1. Methods
T
1.1 Rat MMC Model
IP
All animal use was in accordance with the guidelines of the Animal Care and Use Committee of
CR
Yale University. Previous studies have established a reproducible MMC model in Sprague-Dawley rats (14-16, 18-22). Time-dated pregnant Sprague-Dawley rats from Charles River Laboratories (Wilmington,
US
MA) were gavage fed 40mg/kg of 97% pure all-trans retinoic acid (Across Organics; Morris Plains, NJ) dissolved in extra virgin olive oil (Whole Foods Market; Austin, TX) on gestation day E10.
AN
1.2 Poly(lactic-co-glycolic acid) (PLGA) Particle Fabrication
M
Poly(lactic-co-glycolic acid) (PLGA) nanoparticles (NPs) were loaded with coumarin-6 (C6)
ED
fluorescent green dye to track the delivery of NPs after intra-amniotic (IA) injection. PLGA nanoparticles encapsulating C6 dye were synthesized using a previously described single-emulsion solvent evaporation
PT
technique (23-25).
CE
1.3 Alginate Microparticle (MP) Fabrication and Characterization
Sterile alginate was purchased from NovaMatrix (Sandvika, Norway) and used in all experiments.
AC
Alginate microparticles (MPs) were fabricated using a previously described method (12). Alginate particles were loaded with recombinant human bFGF stabilized with human serum albumin (HSA), bovine serum albumin (BSA) conjugated to Texas Red® (BSA-TR) (ThermoFisher), or left blank. Once fabricated, the alginate particle pellet was then suspended in ultrapure water, aliquoted, flash frozen, and lyophilized. Goal protein concentrations were 1% BSA-Texas Red (w/w alginate) and 2.0% HSA-bFGF (w/w alginate). All particles were stored at -20°C until use. Particle size (hydrodynamic diameter) was measured using dynamic light scattering and confirmed with electron scanning microscopy. Surface charge (zeta potential) was measured in solution using a Malvern Zetasizer (Malvern Instruments, UK).
4
ACCEPTED MANUSCRIPT 1.5 Intra-amniotic Binding Studies and Fluorescent Photo Analysis
MMC was induced as described above. Control and retinoic acid-exposed dams between 17 and 21 days post-conception were anesthetized with inhaled isoflurane. The gravid uterus was exposed through a midline laparotomy incision. For the fluorescent binding studies, particles were re-suspended
T
by vortex and water bath sonication in PBS. Green fluorescent polystyrene MPs with (Polystyrene-COOH)
IP
and without (Polystyrene-plain) carboxyl surface groups and average diameter of 10 µm were purchased from Phosphorex (Hopkinton, MA). These polystyrene particles were used for injections without any
CR
modification. Surface charge (zeta potential) was measured in solution using a Malvern Zetasizer (Malvern Instruments, UK). All injections were performed with glass micropipettes (tip diameter ~60 µm)
US
and a pneumatic micro-injector (Narishige; Japan). Dams and pups were sacrificed 3 hours after
AN
injections were completed. Specimens were analyzed under a fluorescence microscope (Leica Microsystems, Wetzlar, Germany). Ex vivo imaging of MMC defects was performed on a Leica M80
M
fluorescence stereomicroscope (Wetzlar, Germany). Images were analyzed using Fiji imaging software (NIH website) (26). MMC defect-to-skin mean brightness ratios (MBR) were calculated using raw
ED
integrated pixel densities from image overlays of equal area (400 x 400 pixel squares). MBRs for all groups were compared using one-way analysis of variance (ANOVA) and Tukey multiple comparisons
PT
test.
CE
1.6 Intra-amniotic Alginate-HSA-bFGF MP injections
AC
MMC was induced as described above. Transuterine injections were performed on E17.5 under general isoflurane anesthesia, and groups were sacrificed at E21. Each individual fetal specimen was examined under a dissection stereomicroscope for the presence or absence of tissue covering the MMC defect. The experimental group was treated with alginate MPs loaded with bFGF and human serum albumin for stabilization (Alginate-HSA-bFGF). Control groups were treated with PBS, free bFGF, blank alginate (Alginate-Blank), alginate with fluorescently tagged albumin (Alginate-BSA-TR), or left uninjected. For all injections involving alginate particles, each fetus received 30 µL of 20 mg/mL alginate MPs suspended in PBS (approximately 9 µg bFGF per dose). For the arm treated with bFGF alone, bFGF
5
ACCEPTED MANUSCRIPT in PBS was received from the vendor (Creative Biomart, Shirley, New York) was diluted to match the expected concentration of bFGF in the Alginate-HSA-bFGF.
1.7 Histological Analysis
Specimens were fixed for 2 days in 4% (w/v) paraformaldehyde (Electron Microscopy Sciences;
T
Hartfield, PA) in PBS and transferred at room temperature in 70% (v/v) ethanol in distilled water for
IP
paraffin embedding and sectioning. Cut sections were stained with hematoxylin-eosin, pan cytokeratin
CR
AE1/AE3, or trichrome. The slides were imaged on a Zeiss Axio Scope light microscope (Carl Zeiss
US
Microscopy; Germany).
1.8 Statistical Analysis
AN
Defect-to-skin brightness ratio data for the fluorescence binding study was analyzed using ANOVA with Tukey multiple comparisons test. Again, an adjusted P value < 0.05 was considered
M
significant. Fisher’s exact test was used to analyze the results of the therapeutic growth factor injections.
ED
Statistical analyses were performed using GraphPad Prism (version 7; GraphPad Software; La Jolla, CA).
PT
2. Results
CE
2.1 Fluorescence MP Binding Study
All dams (N=6) and fetuses (n=57) injected with PLGA C6 NPs, Alginate-BSA-TR MPs, and
Figure 1
AC
green fluorescent polystyrene MPs were viable at the time of sacrifice 3 hours after injection. At 3 hours, particles can be seen adherent to the MMC defect (Figure 1). PLGA NPs (p=0.0461), alginate MPs (p<0.0001), and polystyrene-COOH MPs (p<0.0001) had significantly higher defect-to-skin
Figure 2
brightness ratios compared to controls (Figure 1 & 2). The most specific binding was achieved by polystyrene-COOH MPs [mean brightness ratio (MBR) = 2.139, 95% CI 1.768-2.510] and by alginate MPs (MBR=1.957, 95% CI 1.44–2.48) followed by PLGA NPs (MBR =1.61, 95% CI 1.34–1.88). Unmodified polystyrene MPs (Polystyrene-plain) had the weakest (p=0.9796) and least specific binding (MBR=1.18, 95% CI 1.07–1.30) compared to the non-injected controls (MBR=1.09, 95% CI 1.013-1.168) (Figure 2).
6
ACCEPTED MANUSCRIPT Alginate particles were approximately 4.752 +/- 42 microns in diameter with a zeta potential of -35.3 +/2.5 mV. The Polystyrene-COOH particles were approximately 10 microns in diameter with a zeta potential of -47 mV. Polystyrene-plain particles were 10 microns in diameter with a zeta potential of -40 mV.
2.2 Fetal delivery of Alginate-HSA-bFGF MPs and other controls
T
All dams that underwent laparotomy and IA injections were healthy at the time of sacrifice (N =
IP
22). 150 of 239 fetuses (62.8%) that received IA injections at E17.5 were viable at E21. Fluorescent
CR
stereomicroscopy of an E21 MMC specimen treated Alginate-BSA-TR (no growth factor) revealed red fluorescence throughout a layer of tissue over the MMC defect with negligible fluorescent signal from surrounding skin. On histological analysis of the same specimen, the red fluorescent layer appears to
Figure 3
US
Figure 4
be a scaffold of cellularized alginate associated with unorganized keratinized tissue (Figure 3). In all, 2
AN
of 18 (11%) specimens treated with Alginate-BSA-TR and 3 of 19 (16%) treated with AIginate-Blank exhibited similar keratinized tissue coverage. In the Alginate-HSA-bFGF treatment group, 18 of 61 (30%)
M
fetuses had evidence of partial soft tissue coverage compared to 0 of 24 un-injected (P=0.0021), 0 of 13
ED
PBS (P=0.0297), and 0 of 42 bFGF controls (P<0.0001). Histological analysis of Alginate-HSA-bFGF specimens positive for gross tissue coverage confirmed areas of skin-like soft tissue covering the spinal
PT
cord (Figure 4). The skin-like tissue over the MMC defect demonstrated strong staining for pan cytokeratin AE1/AE3 specifically in the area covering the spinal cord, but it had no noticeable hair
CE
follicles. The higher rate of defect coverage in the Alg-HSA-bFGF group failed to reach statistical significance when compared to the Alg-BSA-TR (P=0.3711) and Alginate-Blank (P=0.1364) groups;
AC
however, the histological characteristics of the positive Alg-HSA-bFGF specimens were qualitatively different from the abnormal, unorganized appearance of the Alg-BSA-TR and blank alginate controls.
3. Discussion
This report presents the results of preliminary investigations into a novel, minimally invasive, particle-based approach to the prenatal treatment of myelomeningocele (MMC). Prior published work has already proven the ability of nano- and microparticles to serve as controlled-release delivery vehicles for various molecules from chemotherapy drugs to DNA oligomers (12, 13, 24, 27, 28). Alginate particles
7 Figure 2
ACCEPTED MANUSCRIPT were chosen as the therapeutic vehicle of choice in this study because they are biocompatible, biodegradable, and more efficient than PLGA MPs at encapsulation and controlled delivery of growth factor proteins such as bFGF (12, 13, 29). In the initial fluorescence binding experiment, alginate MPs bound specifically to the MMC defect without any surface modification (Figure 2). To investigate why alginate particles preferentially bind to the MMC defect, fluorescent polystyrene MPs of known size and
T
surface characteristics (carboxyl groups) were tested. Given that the polystyrene-COOH particles with a
IP
more negative zeta potential bound more specifically to the defect than unmodified polystyrene particles,
CR
it is possible that either terminal carboxyl groups, negative zeta potential, or both characteristics contribute to stronger binding to the MMC defect. Of note, alginate MPs are rich in carboxyl groups and
AN
particles contribute to their robust and specific binding.
US
have a negative zeta potential, and therefore, it is possible that these inherent characteristics of alginate
With evidence that alginate MPs bind specifically to the MMC defect at E17.5, we studied both
M
the feasibility and the outcomes of intra-amniotic injection of Alginate-HSA-bFGF MPs. When MMC defects treated with Alginate-HSA-bFGF MPs were compared to untreated fetuses and fetuses treated
ED
with just PBS or free bFGF, a significant number of the fetuses treated with Alginate-HSA-bFGF had coverage of the exposed spinal cord tissue with skin-like tissue. The gross findings were confirmed on histology sections (Figure 4). The immunohistochemistry pan cytokeratin AE1/AE3 stain is a cocktail of
PT
Figure 4
antibodies against cytokeratin found in epithelial cells. In the skin, AE1/AE3 specifically stains the
CE
epidermis, eccrine glands, and folliculosebaceous-apocrine unit (30-32). In addition, the AE1/AE3 stain
AC
was used in a previously published study, which used bFGF-impregnated gelatin sponges to grow new skin-like tissue over the defect (14). Therefore, because the tissue covering the MMC defects in these Alginate-HSA-bFGF stained strongly on AE1/ AE3 sections, it is likely that it is rapidly forming dermal tissue. Moreover, none of the control specimens exhibited similar findings on gross or histological analysis. Of note, several of the specimens appeared to have a thick scaffold overlying the skin-like tissue over the spinal cord (Figure 4). The scaffold appeared to be composed of scattered red blood cells, some nucleated cells, and circular clear structures consistent with alginate particles. The overall structure seemed loosely connected (it washed away during the preparation of pan cytokeratin AE1/AE3 slides. This finding suggests that, after binding to the MMC defect, alginate MPs form a microscopic scaffold
8 Figure 3
ACCEPTED MANUSCRIPT through aggregation of the particles. This hypothesis is supported by the fact that 11-16% of fetuses treated with blank alginate MPs and Alginate-BSA-Texas Red also developed a scaffold and appear to have a partial ingrowth of hyperkeratotic skin underneath the scaffold despite the lack of any encapsulated growth factor (Figure 3).
T
One of the main limitations to this work is the variability in the shape and severity of the MMC
IP
defects in the rat MMC model. Despite the variable nature of the defects, we obtained approximately 90% yield of isolated MMC defects, and none of our uninjected or PBS-injected controls exhibited anything
CR
similar to the grossly apparent skin covering the MMC defect seen in positive experimental specimens. Because of the variability in the dimensions of the MMC defects and the preliminary nature of these
US
investigations, we opted to use a simple binary outcome measure to evaluate for MMC defect coverage.
AN
Partial soft tissue coverage was either present or absent. Our method for assessing the soft tissue coverage did not allow for accurate measurement of the fraction of the defects that were covered.
M
Similarly, although a complete watertight closure is the gold standard MMC repair needed to prevent cerebrospinal fluid leaks and infection, there was no definitive way to prove that our tissue coverings were
ED
watertight or absolutely complete (10). Despite these limitations, however, histology sections confirmed
underlying spinal cord.
PT
the dissection microscope findings and proved that specimens did indeed have soft tissue layers covering
CE
Because intra-amniotic alginate particle therapy for MMC is a novel procedure, there are many aspects of the methods and materials that require optimization. There are multiple technical challenges to
AC
the injections that must be mastered, and the procedures must be performed precisely and efficiently to prevent treatment variability and fetal loss. For example, because of the small size the amniotic space target, it is possible that particles were delivered inadvertently into the extraembryonic coelom, and this mistake would have theoretically resulted in a lower response rate. Ultrasound or microscope-guided injection can be used to ensure proper injection technique in the future. Another limitation to this study is the unknown safety profile of bFGF alginate particle intra-amniotic injections in humans or other animals. It is true that bFGF is a potent mitogen that stimulates the migration, proliferation, and differentiation of cells of mesenchymal and neurectodermal origin, such as keratinocytes, fibroblasts, melanocytes, and
9
ACCEPTED MANUSCRIPT endothelial cells (33). One major concern about intra-amniotic injections of particles delivering growth factors is inadvertent delivery to internal sites such as the gut or lung. In unpublished work from our laboratory, a survey of rat fetal organs during these timepoints revealed no significant particle infiltration of the lung, gut, liver, spleen, and inner dermis after intra-amniotic injection with either PLGA or alginate particles. In order to surpass regulatory barriers to treatment in humans, intra-amniotic alginate-bFGF
T
particle therapy will need to undergo more rigorous testing in successively higher animal models to
IP
establish a safety profile and to prove there are no deleterious teratogenic or carcinogenic effects.
CR
Future experiments will likely include higher growth factor doses and modification of the surface characteristics of the alginate MPs for more specific targeting, increased growth factor delivery to the
US
MMC defect, and increased particle scaffold aggregation. After this alginate particle treatment has been
AN
optimized for the rat MMC model, it can be applied to larger animal models, such as the ovine MMC model.
M
4. Conclusions
ED
Intra-amniotic injection of biocompatible PLGA and alginate particles for the treatment of MMC is feasible. Both PLGA and alginate bind preferentially to the MMC defect compared to the surrounding fetal
PT
skin. When injected into the amniotic space, alginate microparticles are capable of aggregating to form
CE
microscopic scaffolds over the MMC defect. When these particles are loaded with bFGF, 30% of cases resulted in significant soft tissue coverage of the MMC defect compared to controls. This line of therapy
AC
requires further optimization but could represent a broadly applicable minimally invasive adjunct to fetal therapy for MMC.
10
ACCEPTED MANUSCRIPT References
1.
Anglim B, Mandiwanza T, Miletin J, Turner M, Kennelly MM. The natural history of neural
tube defects in the setting of an Irish tertiary referral foetal medicine unit. J Obstet Gynaecol. 2016;36(1):19-23.
T
Davis BE, Daley CM, Shurtleff DB, Duguay S, Seidel K, Loeser JD, et al. Long-term
IP
2.
survival of individuals with myelomeningocele. Pediatr Neurosurg. 2005;41(4):186-91. Manning SM, Jennings R, Madsen JR. Pathophysiology, prevention, and potential
CR
3.
4.
US
treatment of neural tube defects. Ment Retard Dev Disabil Res Rev. 2000;6(1):6-14. Dillon CM, Davis BE, Duguay S, Seidel KD, Shurtleff DB. Longevity of patients born with
5.
AN
myelomeningocele. Eur J Pediatr Surg. 2000;10 Suppl 1:33-4.
Hunt GM. 'The median survival time in open spina bifida'. Dev Med Child Neurol.
Shurtleff DB, Luthy DA, Nyberg DA, Mack LA. The outcome of fetal myelomeningocele
ED
6.
M
1997;39(8):568.
brought to term. Eur J Pediatr Surg. 1994;4 Suppl 1:25-8. Adzick NS, Thom EA, Spong CY, Brock JW, 3rd, Burrows PK, Johnson MP, et al. A
PT
7.
randomized trial of prenatal versus postnatal repair of myelomeningocele. N Engl J Med.
Peranteau WH, Adzick NS. Prenatal surgery for myelomeningocele. Curr Opin Obstet
AC
8.
CE
2011;364(11):993-1004.
Gynecol. 2016;28(2):111-8. 9.
Saadai P, Farmer DL. Fetal surgery for myelomeningocele. Clin Perinatol.
2012;39(2):279-88. 10.
Kabagambe SK, Jensen GW, Chen YJ, Vanover MA, Farmer DL. Fetal Surgery for
Myelomeningocele: A Systematic Review and Meta-Analysis of Outcomes in Fetoscopic versus Open Repair. Fetal Diagn Ther. 2017.
11
ACCEPTED MANUSCRIPT 11.
Farrelly JS, Bianchi AH, Ricciardi AS, Ahle SL, Luks VL, Saltzman WM, Stitelman DH.
Intra-Amniotic Delivery of Biodegradable Microparticles and Nanoparticles to the Spinal Defect in a Rat Myelomeningocele Model. MOLECULAR THERAPY. 2017;25(5):322-3. 12.
Jay SM, Saltzman WM. Controlled delivery of VEGF via modulation of alginate
microparticle ionic crosslinking. J Control Release. 2009;134(1):26-34.
T
Jay SM, Shepherd BR, Andrejecsk JW, Kyriakides TR, Pober JS, Saltzman WM. Dual
IP
13.
vascularization. Biomaterials. 2010;31(11):3054-62.
Watanabe M, Jo J, Radu A, Kaneko M, Tabata Y, Flake AW. A tissue engineering
US
14.
CR
delivery of VEGF and MCP-1 to support endothelial cell transplantation for therapeutic
approach for prenatal closure of myelomeningocele with gelatin sponges incorporating basic
15.
AN
fibroblast growth factor. Tissue Eng Part A. 2010;16(5):1645-55. Watanabe M, Li H, Roybal J, Santore M, Radu A, Jo J, et al. A tissue engineering
M
approach for prenatal closure of myelomeningocele: comparison of gelatin sponge and
ED
microsphere scaffolds and bioactive protein coatings. Tissue Eng Part A. 2011;17(7-8):1099110.
Watanabe M, Kim AG, Flake AW. Tissue Engineering Strategies for Fetal
PT
16.
17.
CE
Myelomeningocele Repair in Animal Models. Fetal Diagn Ther. 2014. Watanabe M, Li H, Kim AG, Weilerstein A, Radu A, Davey M, et al. Complete tissue
AC
coverage achieved by scaffold-based tissue engineering in the fetal sheep model of Myelomeningocele. Biomaterials. 2015;76:133-43. 18.
Mackenzie TC, Flake AW. Human mesenchymal stem cells persist, demonstrate site-
specific multipotential differentiation, and are present in sites of wound healing and tissue regeneration after transplantation into fetal sheep. Blood Cells Mol Dis. 2001;27(3):601-4. 19.
Danzer E, Schwarz U, Wehrli S, Radu A, Adzick NS, Flake AW. Retinoic acid induced
myelomeningocele in fetal rats: characterization by histopathological analysis and magnetic resonance imaging. Exp Neurol. 2005;194(2):467-75.
12
ACCEPTED MANUSCRIPT 20.
Danzer E, Kiddoo DA, Redden RA, Robinson L, Radu A, Zderic SA, et al. Structural and
functional characterization of bladder smooth muscle in fetal rats with retinoic acid-induced myelomeningocele. Am J Physiol Renal Physiol. 2007;292(1):F197-206. 21.
Danzer E, Radu A, Robinson LE, Volpe MV, Adzick NS, Flake AW. Morphologic analysis
of the neuromuscular development of the anorectal unit in fetal rats with retinoic acid induced
IP
Danzer E, Zhang L, Radu A, Bebbington MW, Liechty KW, Adzick NS, et al. Amniotic
CR
22.
T
myelomeningocele. Neurosci Lett. 2008;430(2):157-62.
fluid levels of glial fibrillary acidic protein in fetal rats with retinoic acid induced
US
myelomeningocele: a potential marker for spinal cord injury. American journal of obstetrics and gynecology. 2011;204(2):178 e1-11.
Fahmy TM, Samstein RM, Harness CC, Mark Saltzman W. Surface modification of
AN
23.
biodegradable polyesters with fatty acid conjugates for improved drug targeting. Biomaterials.
McNeer NA, Chin JY, Schleifman EB, Fields RJ, Glazer PM, Saltzman WM.
ED
24.
M
2005;26(28):5727-36.
Nanoparticles deliver triplex-forming PNAs for site-specific genomic recombination in CD34+
Ricciardi AS, Bahal R, Farrelly JS, Quijano E, Bianchi AH, Luks VL, et al. In utero
CE
25.
PT
human hematopoietic progenitors. Mol Ther. 2011;19(1):172-80.
nanoparticle delivery for site-specific genome editing. Nat Commun. 2018;9(1):2481. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: an
AC
26.
open-source platform for biological-image analysis. Nat Methods. 2012;9(7):676-82. 27.
Mahoney MJ, Saltzman WM. Transplantation of brain cells assembled around a
programmable synthetic microenvironment. Nature biotechnology. 2001;19(10):934-9. 28.
Zhou J, Patel TR, Sirianni RW, Strohbehn G, Zheng MQ, Duong N, et al. Highly
penetrative, drug-loaded nanocarriers improve treatment of glioblastoma. Proceedings of the National Academy of Sciences of the United States of America. 2013;110(29):11751-6.
13
ACCEPTED MANUSCRIPT 29.
Jay SM, Shepherd BR, Bertram JP, Pober JS, Saltzman WM. Engineering of
multifunctional gels integrating highly efficient growth factor delivery with endothelial cell transplantation. FASEB J. 2008;22(8):2949-56. 30.
Wallace ML, Smoller BR. Immunohistochemistry in diagnostic dermatopathology. J Am
Acad Dermatol. 1996;34(2 Pt 1):163-83; quiz 84-6.
T
Folpe AL, Cooper K. Best practices in diagnostic immunohistochemistry: pleomorphic
IP
31.
32.
CR
cutaneous spindle cell tumors. Arch Pathol Lab Med. 2007;131(10):1517-24. Fuertes L, Santonja C, Kutzner H, Requena L. Immunohistochemistry in
US
dermatopathology: a review of the most commonly used antibodies (part I). Actas Dermosifiliogr. 2013;104(2):99-127.
Wu J, Ye J, Zhu J, Xiao Z, He C, Shi H, et al. Heparin-Based Coacervate of FGF2
AN
33.
Improves Dermal Regeneration by Asserting a Synergistic Role with Cell Proliferation and
M
Endogenous Facilitated VEGF for Cutaneous Wound Healing. Biomacromolecules.
AC
CE
PT
ED
2016;17(6):2168-77.
14
ACCEPTED MANUSCRIPT Figure Legends Figure 1. Left to right: E17.5 IA PLGA C6 (green dye); E17.5 IA Alginate-BSA-Texas Red™ (Alg-BSATR); E18 Polystyrene-COOH (green dye); E17.5 Polystyrene-plain (greed dye) can be seen bound to the MMC defect at 2.0x magnification Figure 2. Binding specificity of fluorescent particles to the MMC defect compared to adjacent skin.
T
Polystyrene-COOH IA injection resulted in the brightest defect-to-skin ratio which was higher than that of
IP
Polystyrene-plain (p<0.0001), PLGA-C6 (p=0.056) and the non-injected control (p<0.0001). The defect-to-
CR
skin ratio for Alginate-BSA-Texas Red was similar to that of Polystyrene-COOH and was also significantly higher than that of Polystyrene-plain (p=0.0006) and the non-injected control (p<0.0001).
US
Figure 3. Light stereomicroscope image of MMC defect (top left) treated with intra-amniotic injection of Alginate-BSA-Texas Red (Alg-BSA-TR) at E17.5. Scaffold over the MMC defect exhibited fluorescence
AN
under the red channel only, suggesting cellular incorporation of Texas Red tagged albumin still present 3 full days after injection. Histological analysis (bottom) confirmed the presence of alginate scaffold
M
overlying abnormal spinal cord with pan cytokeratin AE1/AE3 positive cells underneath the scaffold.
ED
Figure 4. Gross light images of control (top) and two selected specimens treated with Alg-HSA-bFGF (middle, bottom) along with their associated MMC defect histology sections stained with trichrome (middle
PT
column) and pan cytokeratin AE1/AE3 (right column). The heads of the specimens are located superiorly in the gross photos to the left, and the axial histology sections are oriented with the fetal back superiorly
CE
and vertebral bodies inferiorly. On the axial histology sections through the MMC defects, it is apparent that the specimen treated with Alg-HSA-bFGF has abnormal dermis that stains strongly for pan
AC
cytokeratin AE1/AE3 covering the spinal cords. Of note, the bottom specimen has an extra structure overlying the new skin representative of loosely adherent cellularized alginate scaffold that washed away during the pan cytokeratin stain preparation.
15
Figure 1
Figure 2
Figure 3
Figure 4