Vascular Endothelial Growth Factor – Releasing Scaffolds Enhance Vascularization and Engraftment of Hepatocytes Transplanted on Liver Lobes

Citation and Paper Link:

Cohen S, Dvir-Ginzberg M, Gamlieli-Bonshtein I, et al. “Vascular Endothelial Growth Factor – Releasing Scaffolds Enhance Vascularization and Engraftment of Hepatocytes Transplanted on Liver Lobes.” Advances in Tissue Engineering: Angiogenesis. 2005 May-Jun; 11(5-6): 715-22.

http://www.ncbi.nlm.nih.gov/pubmed/15998213

Introduction/Motivation

There is currently an interest in developing new treatments for end-stage liver disease and enzyme deficiency by transplanting hepatocytes into affected livers and inducing the regeneration of healthy tissue in vivo. Unfortunately, the two most popular hepatocyte transplantation methods are currently portal vein injection and mesenteric membrane implantation, which have both proven to be unreliable, mainly due to poor cell viability. The extreme vascularization of liver tissue prevents a unique challenge for the engineering of tissue-regenerating scaffolds because implanted cells must have immediate access to oxygen and nutrients in circulation.

The paper by Cohen S, et al, investigates the viability of hepatocytes which have been injected into scaffolds impregnated with VEGF – a known signaling molecule involved with liver regeneration – and implanted into rat livers. The novel approach of this study was to achieve increased cellular engraftment by first inducing the growth of microvessels into the scaffolds before injecting hepatocytes.

Summary of Methods and Results

For the initial stage of the study, two experimental groups were created in order to determine the time dependence of scaffold vascularization: 1) alginate scaffolds implanted into the left liver lobe of rats (control), and 2) alginate scaffolds containing VEGF microspheres implanted into the left liver lobe of rats (experimental). Figure 1 shows one of the implanted scaffolds. Alginate tissue scaffolds were used for this study because of their high porosity (~90%) which would allow ample room for vessel ingrowth and hepatocyte engraftment. To produce the VEGF scaffolds, PLGA microspheres were impregnated with VEGF and incorporated into the scaffolds at the time of fabrication. The microspheres were impregnated with 3.75 μg per 200 mg of PLGA and incorporated into the scaffolds at 40 mg of microspheres per scaffold. The kinetics of VEGF release were measured by ELISA and found to be constant at 8-10 ng/day for two weeks, after an initial burst release of 40%.

Figure 1

Two measurements were made at the time points of 3, 7, and 14 days post-implantation: (Fig 2A) the % cross-sectional area of the scaffold replaced by new vessel ingrowth and (Fig 2B) the capillary density in the scaffold measured in #/mm². All data was collected by harvesting the scaffolds from the rats and taking a histological slice through the approximate middle and staining it with H&E.

Figure 2

Figure 2A shows that there was no statistical difference between the control and experimental groups with regards to the cross-sectional area of the scaffold taken up by new tissue ingrowth. Figure 2B, however, shows that the number of capillaries growing into the scaffolds was affected by VEGF and that the experimental scaffold group had a higher blood vessel density. In order to inject hepatocytes into a pre-vascularized scaffold while still allowing for ample space for engraftment, the 7 day mark was chosen as the optimal injection date.

For the hepatocyte engraftment phase of the study, three experimental groups were used: 1) scaffolds implanted into rats with no VEGF and no hepatocyte injection (sham-operated), 2) scaffolds implanted without VEGF, but hepatocytes were injected on day 7 (control), and 3) scaffolds implanted with VEGF microspheres and hepatocytes were injected also on day 7 (experimental). Histological samples of the scaffolds were taken at time points of 1, 3, 7, and 12 days post-injection. The sham-operated group showed no hepatocytes, indicating that any hepatocyte present in the other groups came from the injection. The histological H&E stains from each time point and experimental group were taken and the cross-sectional area of hepatocytes was computed in μm² per slide, as shown in Figure 3.

Figure 3

For this study hepatocyte engraftment was measured as a function of the cross-sectional area of hepatocytes in the histological slides. Figure 3 shows the decrease in hepatocyte area over time in both the experimental and control groups, however the VEGF was found to have an important effect. The VEGF-impregnated scaffold had higher hepatocyte engraftment at each time point and by day 12 had over a 4 times greater area of intact hepatocytes. This data suggests that the VEGF-induced vascularization served to encourage the attachment of injected hepatocytes.

Overall, this study showed that the vascularization of alginate liver scaffolds can be induced by VEGF and that the enhanced ingrowth of vessels has a positive effect on hepatocyte engraftment.

Shortfalls and Possible Improvements

The main drawback of this study is that it does not directly address its motivation: the restoration of liver function through the transplantation of hepatocytes into a host. In order to investigate liver function, they would have had to remove the scaffolds and analyze the liver-specific gene expression of the engrafted hepatocytes through an experimental method such as PCR. By comparing the gene expression of the implanted cells with normal liver tissue expression, the researchers could have measured how much liver function, if any, was restored. Additionally, using healthy livers in the experiment most likely provided results that would be different than if they had used diseased livers. It is possible that healthy livers are more or less likely to vascularize the scaffolds than diseased livers.

A shortfall of the data analysis was the failure of the researchers to report the vascularization of their scaffolds in comparison with normal anatomical values. Again, it is difficult to assess the efficacy of alginate scaffold vascularization if the normal liver vascularization values are not known. In order to improve the study, I would have proposed taking similar cross-sections of liver tissue from the lobe in which the implant was made and comparing vascularization amounts with the scaffold.

Finally, the measurement of hepatocyte cross-sectional area (which was the primary indicator of cell engraftment) was highly dependent on their injection process which was not rigidly standardized across all of the scaffolds. The cells were injected into the scaffolds at “different sites” in order to have a better distribution of cells, but then it was reported that cell engraftment was greater at the periphery of the scaffolds. If some scaffolds received injections closer to areas of higher engraftment rates than others, then the cross-sectional area alone can not be used as a reliable indicator of cell engraftment. An investigation of cell engraftment as a function of scaffold injection site must be conducted before making conclusions about the effect of VEGF on engraftment.

Microfluidic environment for high density hepatocyte culture

Introduction:

Upon isolation, monolayer hepatocytes lost liver-specific function rapidly; this limits their application in drug screening and as a dependable cell source for bioartificial liver support. Hepatocytes function is influenced by the microenvironment such as flow rate, cell density, media composition, geometry configuration, extracellular matrix, and cell-cell contacts. Microfludic technology has recently offered the advantages in controlling fluid flows and in improving cell culture microenvironments.

Summary:

The objective of the study is to “controllably maintain cells in a “tissue-like” state by using a microfluidics mass transport design that mimics the tissue microvasculature.” The fluidic device, in short, was a PDMS based device containing one inlet, one outlet, and one cell loading area (Figure 1).

Fig. 1 (a) Microfluidic culture unit design. The microfluidic unit consisted of three parts: a 150 μm wide by 440 μm long cell culture area (blue), a microfluidic perfusion barrier (gray), and a medium flow channel (red). Cells were introduced from the top port, and localized into the cell culture area. The perfusion barrier consisted of a grid of channels 5 μm wide and 2 μm tall, serving to prevent cells from passing through, while enabling nutrient exchange from the flow channel on the opposite side.Inset shows SEM micrograph of perfusion channels. Scale bar represents 5 μm. (b) Glucose concentration at the center of cell culture mass was determined using finite element analysis software and plotted against medium flow rate. (c) Surface plot of glucose concentration throughout one culture unit with an inlet glucose concentration of 25 μmM. At a flow rate of 10μl/day, glucose concentration at the center of the cell mass after consumption was theoretically determined to be 22 mM.

The hepatocytes (HepG2/G3A) cells were seeded in the device (incubated on a 45 degree slope rack to maintain continous gravity driven flow) and in the standards tissue culture 12-well plates (control). The flow rate for the device was measured by collecting the outlet media (10.3±1.6 μl/day/well).

Characterization of different cell cultures:

Cell viability and proliferation were determined daily using Trypan Blue and Live/Dead fluoresence assay. The cell viability in the device after one week is 80%. The cells in the control appeared strongly attached and more spread, with distinct membrane The cells in the device exhibited a “tissue-like” morphology (i.e. dense packing, cuboidal geomery, and indistinguishable fused membrane (Figure 3).

Fig. 2 HepG2/C3A human hepatoma cell growth inside the microfluidic cell culture array. Scale bar represents 100 μm

Fig. 3 High density cell culture. Phase contrast micrographs of confluent cell morphology on a standard 12-well tissue culture plate (a) and inside one sinusoid of the microfluidic cell culture array (b) after 8 days of culture. Images are at equal magnification, with scale bar representing 50 μm. (c) Cells were seeded at 10^{5 cells/well for controls (filled circles) and about 500 cells/well for devices (filled triangles) and measured for cell density over 1 week.}

Liver-specific synthetic function of albumin was examined via a quantitative dot-blot assay (a method using immunostaining on nitrocellulose membrane). The result (Figure 4) shows that the hepatocytes on the device were able to produce three times albumin on per-cell basis compared to the control culture.

Fig. 4 Albumin secretion was characterized in TC dish and the microfluidic device after 4 days of culture. Albumin secretion from cells cultured in the microfluidic device increased threefold compared with on a dish.

Comment:

Overall, this work is impressive because the design took into account the physiological relevant mass transport condition; therefore, sustaining a high cell density (>2000 cells/mm2) without nutrient depletion over one week is possible. Additionally, the design enables the cells to assembly in close proximity thus might enhance cell-cell contacts normally found in native liver tissue. Furthermore, this work also shows how the mass transport can affect maintenance of differentiated phenotype and biological function of hepatocytes. Finally, the configuration (i.e. flow rate, geometry) also enhanced the production of albumin in the cells by threefold comparing to a 12-well plate.

The researchers suggest that the close proximity assembly of hepatocytes cell in the device might enhance cell-cell contacts such as tight junctions and desmosome, which normally found in native liver tissue. However, no chemical assay (i.e. Cadherin immunostaining assay or Cadherin gene expression assay) was used to test their suggestion. Meanwhile the end goal is to use the device in drug screening, drug metabolism, fundamental physiology studies; the detoxification activity mediated by cytochrome p450 (via CYP2B6, CYP2E1, CYP1A2 gene expression) of the hepatocytes cells was not examined.

Reference:
Mimi Y. Zhang & Philip J. Lee & Paul J. Hung &Terry Johnson & Luke P. Lee & Mohammed R. K. Mofrad, “Microfluidic Environment for High Density Hepatocyte Culture”, Biomed Microdevices (2008) 10:117–121

Link: http://biomechanics.berkeley.edu/assets/papers/LiverChip_Mimi2007.pdf

Apatite content of collagen materials dose-dependently increases pre-osteoblastic cell deposition of a cement line-like matrix.

Reference

Perrier A, Dumas V, Linossier MT, Fournier C, Jurdic P, Rattner A, Vico L, Guignandon A. Apatite content of collagen materials dose-dependently increases pre-osteoblastic cell deposition of a cement line-like matrix. Bone 2010.

Summary

The authors showed the increase in osteoblast activity during bone remodeling process is attributed to pre-osteoblastic secretion of VEGF-A and expression of RGD-containing proteins, osteopontin and fibronectin. Bone breakdown exposes varying degree of mineralization (dom) within a collagen I matrix; pre-osteoblast cells encounter such structure and release the above factors to promote cell viability and adhere to the collagen I-hydroxyapatite composite. The researchers evaluated this by observing cell morphology, mRNA production, adhesion rate, and cell migration. With increasing dom, the cells took on a more unique morphology, produced more mRNA for VEGF-A and RGD-containing proteins, displayed higher adherence rate, and migrated larger distances.

Results

The Apatite Collagen Complex (ACC) was made in 3 groups: 0%, 25%, and 50% (as the human bone is max 42%) apatite degree of mineralization. The ACC was first characterized as a substrate under light microscopy to quantify (via ImageJ) the amount of collagen and degree of mineralization was easily distinguishable. The ACC substrates were then tested for biocompatibility by seeding MC3T3-E1 cells; the light microscopy images are shown below. As the dom increases, the cells take on a more elongated, ciliated, stretched morphology showing adhesion and migration.

mRNA of VEGF-A was quantified using ELISA which as the figure below indicates, the 25% and 50% show marked, dose-dependent increases from 0% to 25% to 50% both in soluble VEGF-A and matrix-bound. The RGD-containing proteins OPN & FN were also shown to increase in mRNA content by RT-PCR; OPN increased 50-fold for 25% & 40-fold for 50% whereas FN increased by 7-fold for 25% and 8.5-fold for 50%. B3 integrin, the RGD-containing protein cell surface receptors, were also upregulated by 3-fold to 4-fold 25%-50%, respectively.

From protein expression, the authors moved on to the manifestation of increased expression by mechanically testing cell adhesion. A vertical sinusoidal vibration was applied to the cells; collagen I indeed did facilitate better adhesion and as the dom increased, the adhesion also increased dramatically: 4.59+0.82% cells resisted (TCPS), 29.00+5.77% (ACC0), 56.17+11.84% (ACC25) and 75.14+6.44 (ACC50). Finally, cell migration was observed using time-lapse microscopy in which the cells’ migration was recorded to move greater distances with increasing dom; it was also found that cells in unmineralized substrates (TCPS & ACC0) had low speed profile while those in 25% and 50% had alternating low and high speed profiles.

Shortfalls

Overall, this paper presents relatively minute shortfalls.

Although the paper used several methods to prove the changing profile of VEGF and OPN/FN, the most basic information of experimental setup was omitted: the authors did not provide a total cell count in the 4 categories they were comparing. Hence, there is no way to tell if the results are statistically significant except relative to each other.

The ACC substrate system was imported from those used in previous osteoclast experiments; osteoclasts, though derived from the same precursor, have ruffled border morphology to aid in adhesion and bone breakdown. The ACC was thus prepared in a way to aid adhesion which can contribute to the increasing cell adhesion and morphological changes to osteoblast with increasing dom, thus presenting a confounding factor.

A type of protein quantification for the RGD-containing proteins OPN, FN, and B3 integrin would have complemented the mRNA results as they had done for VEGF-A. Finally, the choice of 32hr post seeding to conduct experiments seem arbitrary given a doubling time of 27 hours for MC3T3-E1 cells; an scientific rationale would have strengthen the results.

Transplantation of Insulin-Producing Clusters Derived From Adult Bone Marrow Stem Cells to Treat Diabetes in Rats

Janna Serbo

Paper Summary and Review

Reference:

Gabr, Mahmoud M, Mohamed M Sobh, et al. "Transplantation of Insulin-Producing Clusters Derived From Adult Bone Marrow Stem Cells to Treat Diabetes in Rats." Experimental and Clinical Transplantation 6, no. 3 (2008).

http://www.ectrx.org/forms/ectrxcontentshow.php?year=2008&volume=6&issue=3&supplement=0&makale_no=0&spage_number=236&content_type=FULL%20TEXT

Introduction:

Islets of Langerhans are cell clusters in the pancreas that contain many types of cells, including alpha, beta, and delta cells. Specifically, beta cells produce the hormone insulin and directly secrete insulin into the blood. Insulin is a hormone that controls blood glucose levels, allowing cells to absorb glucose. The cells store glucose as glycogen, a polysaccharide of glucose, to use as a short term energy supply. This way, cells do not use fat stores as an energy source until it is necessary. Diabetes Mellitus Type One, or juvenile diabetes, is an autoimmune disease in which the body attacks the patient’s pancreatic, insulin-producing beta cells.

Summary:

Transplantation of Insulin-Producing Clusters Derived From Adult Bone Marrow Stem Cells to Treat Diabetes in Rats from Experimental and Clinical Transplantation describes adult bone marrow stem cell experiments in rats as a mean to explore Diabetes treatment (Gabr, et al. 2008). Adult bone marrow stem cells derived from Sprague-Dawley rats where isolated and differentiated into glucose regulating, insulin producing cells. These cells were grown into clusters to mimics the Islets of Langerhans (beta cells) that exist in a normal, healthy pancreas. Diabetes was induced in rats by destroying their pancreatic beta cells with streptozotocin, a toxic chemical to beta cells. The differentiate bone marrow cells were transplanted into the diabetic rats by injection into the gonads. The gonads were pulled into the abdominal cavity and sutured onto the abdominal wall to create a “pancreatic like” sack. The injected cells were tested and observed for insulin production and glucose regulation. Insulin was indeed produced, and cells absorbed glucose from the blood to store as an energy source, as would happen in a non-diabetic individual.

Figure 5. (A) Changes in the glucose concentrations among the 4 studied groups: group control (sham experiment), group acellular (diabetic rats treated by culture media), group undifferentiated (diabetic rats treated by undifferentiated hematopoietic-rich stem cells), and group differentiated (diabetic rates treated by differentiated clusters). (B) Changes in the insulin levels in clusters of rats that had undergone a transplant. Blood glucose began to decrease within 2 to 3 days. Throughout the observation period, the determined blood glucose levelswere significantly lower (P < .001) and serum insulin levels were significantly higher (P < .001) in animals in group differentiated than they were in animals in group acellular and group undifferentiated. Following orchidectomy, these rats became hyperglycemic again with a steep reduction in the serum insulin levels.

This paper used a number of experimental methods to test the differentiation and insulin production ability of bone marrow derived cells (see paper for all images). Cells were characterized through protein imaging by using immunostaining. Monoclonal, rat, anti-insulin antibody was used to target and stain insulin-producing differentiate cells. The immunoreactive cells were observed with a Vectastain Elite ABC Kit using tetrachloride as a chromogen. This selected the bone marrow cells that differentiated properly and confirmed that the differentiated cells were producing insulin.

Figure 1. Morphologic changes of hematopoietic-rich stem cells during differentiation. (A) Undifferentiated hematopoietic-rich stem cells, 1 day after isolation (magnification ×200). (B) Hematopoietic-rich stem cells formed isletlike clusters by dimethyl sulfoxide treatment for 3 days; this was followed by treatment with high-glucose and pancreatic extract for 7 days (magnification ×40). (C) Collected isletlike clusters after treatment with nicotinamide and exendin-4 for 7 days (magnification ×100). D) Ditizone staining of isletlike clusters. The clusters distinctly stained crimson red by DTZ (magnification ×400).

Also, differentiated cells were characterized through the analysis of biochemical components using ELISA (Enzyme Linked Immunosorbent Assay). Intracellular C-peptide is a byproduct of the insulin process, showing that cells are producing insulin. An ELISA kit was used to see if cells were secreting Rat C-peptide. This confirmed that undifferentiated cells were not producing insulin and differentiated cells were producing insulin. Furthermore, cells were characterized by gene expression using reverse transcription polymerase chain reaction (RT-PCR). RT-PCR is when RNA is “reverse” transcribed into DNA. Then, the DNA is replicated and amplified by PCR heating and cooling cycles. This was done to check that islet related genes were expressed in differentiated cells, proving that the function of the differentiated cells was similar to the Islet of Langerhans cells. An additional test to check for differentiation was done by staining the cells with DTZ (diphenylthiocarbazone). The cells that show up red will be cells containing zinc, which plays an important role in storage, secretion, synthesis of insulin. This confirms which cells have differentiated correctly.

Shortfalls:

While this paper is very promising, there are a few general problems that were not address. The cells tested were rat adult bone marrow stem cells, making this an impractical treatment alternative. Xenogeneic treatments involve risks of pathogen transfer and immune compatibility complications. This paper was good for a proof of concept, but further research would have to be done specifically for human stem cells and for clinical trials. Also, this treatment is not fully effective for Diabetes Type one patients. The implanted cells would be autologous because they were derived from the adult subject. However in the case of diabetes type one, an autologous source of stem cells still has immune compatibility issues. The body attacks both autologous and allogeneic sources because the disease is autoimmune. Autologous beta-like cells would still be recognized as foreign material by the body. The gonads were used as an encasing, but an impractical one used to keep the cell clusters together for easy removal. Human patients would likely be unwilling to have their gonads injected and sutured into the abdomen. Other protective encapsulation methods would need to be explored to immunoisolate injected cells. Also, future research must still explore problems such as the ability to successfully culture large numbers of islet-like cells in vitro, to provide stable and consistent culture conditions, and to optimize an area for injection and transplantation of differentiated cells.