ALBERT

Description: Key Points CIP4 affects the remodeling of both plasma membrane and cortical cytoskeleton in megakaryocytes. CIP4 in platelet biogenesis involves cortical tension, as does WASP, and WASP-independent plasma membrane reorganization.

Description: Background: Recent clinical trials have demonstrated the efficacy and safety of gene therapy utilizing HIV-derived lentiviral vectors (LVs) for blood disorders. However, the LV requirements and clinical ex vivo cell transduction protocols used in these studies exposes the limitations of the technology and beckons the need for improved LV manufacturing and clinical transduction efficiency. Many methods have been devised to enhance efficiency, although none have circumvented the exorbitant amounts of virus required to achieve therapeutic HSC transduction. Furthermore, prolonged ex vivo cell culture is necessary to achieve sufficient transduction despite exposure to toxic byproducts of LV production. To that end, we developed a novel, scalable microfluidic for clinical LV transduction that leverages mass transfer principles to significantly reduce the amount of LV required to achieve therapeutic levels of gene transfer and transduction time by more efficiently exposing cells to virus. Results: Jurkats were transduced with a GFP-encoding clinical LV in microfluidics with surface areas (SAs) comparable to the bottom surface of 96 and 6-well plates. Microfluidic transductions were compared to well plate transductions with matched SA, cell numbers, viral particles, and incubation times. After LV incubation, cells were removed from the microfluidics and well plates, spun down, re-suspended with fresh media, and cultured for at least 72 hours at 37°C and 5% CO2. Cells were assessed for GFP expression with flow cytometry. Preliminary mouse studies utilized Sca+ cells isolated from CD45.1 donor mice via positive selection. The cells were transduced in the scaled up microfluidic with a bioengineered coagulation factor VIII (fVIII) transgene encoding LV and transplanted into host hemophilia A mice after myeloablative conditioning. Two weeks post-transplantation, blood samples were taken from the recipient animals and assayed for donor cell engraftment by flow cytometry and plasma fVIII activity by chromogenic assay. The high SA:volume ratio of the microfluidic enhances transduction by physically bringing cells and virus into closer proximity and enabling high concentrations of virus to be used without increasing the amount of virus set by the minimum volume requirements of LV transduction platforms (Fig. 1A). The polystyrene bottom of the microfluidic allows for Retronectin coating that immobilizes non-adherent cells on the bottom surface. LV can then be perfused at low concentrations to maintain a constant supply of fresh virus to the cells, increase convective mixing, and to minimize cell exposure to the toxic byproducts of LV production (Fig. 1B). These microfluidics have been scaled up to accommodate 106 cells, with potential to scale up to 107-108 cells (Fig. 1C). Cells transduced in the microfluidics showed 2-6 fold increases in GFP expression over well plates utilizing the same amount of cells, virus, and incubation times (Fig. 2A). The kinetics of LV transduction in the microfluidics also are faster, as seen by the steeper transduction curve. Five hours of incubation in the microfluidic yielded comparable transduction to 24 hours in the 6-well plate (Fig. 2B). Improvements in transduction also were observed by perfusing virus despite using lower virus concentrations (Fig. 2C). Finally, hemophilia A mice transplanted with donor CD45.1 Sca+ cells transduced in the microfluidic have engrafted (Fig. 3A) and produce fVIII (Fig. 3B) after two weeks with similar profiles to control cells transduced in a 6-well plate despite using half the amount of virus and shorter incubation times. Conclusions and ongoing efforts: We describe a novel microfluidic that significantly reduces the amount of virus and ex vivo processing time required for therapeutic levels of transduction in clinical gene therapy. This device is versatile in its compatibility with current transduction strategies such as Retronectin and polybrene in addition to offering new approaches to boosting gene transfer efficiency. Furthermore, we have shown that the device has clinical potential by successfully scaling up cell numbers and transplanting mice with microfluidic transduced cells, of which there is an ongoing effort to monitor fVIII production and determine virus copy number. Future work will involve optimization with transduction-enhancing compounds, further scaling, and continued in vivo experiments. Disclosures Spencer: Expression Therapeutics: Equity Ownership. Doering:Bayer Healthcare: Consultancy, Honoraria, Research Funding; Expression Therapeutics: Equity Ownership.

Description: Background: Disruption of red blood cell (RBC) volume regulation and water homeostasis is a major component of sickle cell pathophysiology. As such, hydration is a mainstay of prevention and treatment of vaso-occlusive crises (VOC) in patients with sickle cell disease (SCD). However, evidence for guiding clinicians' choice for intravenous (IV) fluid hydration in the acute setting is lacking (Okomo, Coch Data Syst Rev, 2012). Pediatric hematologists typically discourage rapid infusion of hypotonic fluid given the risk of hyponatremia, although such fluids as 5% dextrose with 34meq/L or 77meq/L sodium are often used after stabilizing various acute clinical situations relevant to SCD. Although altered sodium concentration and osmolarity have been shown to affect erythrocyte swelling and rheology, dextrose-containing fluids used clinically such as D5 ¼ and D5 ½ normal saline (NS), have not been studied (Reinhart et al., Mic Res, 2015; Hijiya et al., J Lab Clin Med, 1991). To those ends, we sought to investigate the effect of different clinically relevant IV fluid formulations on normal and sickle RBC stiffness and deformability. Methods: Fresh blood was obtained from healthy volunteers and patients with sickle cell anemia (SS) on hydroxyurea and not transfused for at least 100 days. Sterile, clinical grade fluids stored at room temperature were used for the experiments (Baxter, Figure 1A). Our laboratory has previously published a description of a microfluidic device comprised of multiple parallel microchannels 5μm wide recapitulating the in vivo geometry of capillaries (Rosenbluth et al., Lab Chip, 2008). This microvasculature-on-a-chip enables measurements of single-RBC transit times, which correlate with cell stiffness (Figure 1B,C). For the transit time experiments, a master silicon wafer was used to mold the microfluidic channels out of polydimethylsiloxane (PDMS) silicone. Centrifuged RBCs were washed with phosphate-buffered saline (PBS) and then diluted to 0.5% hematocrit (HCT) in the various fluids prior to flow. Cell suspensions were perfused into a microfluidic device pre-coated with 2% bovine serum albumin (BSA) at an average linear flow rate of 0.50 mm s-1 in the smallest channels with a syringe pump (Harvard Apparatus) and then imaged at 20x at 20 frames per second. Images were recorded for future analysis. For the experiments assessing RBC shape, washed RBCs were diluted with PBS to 0.5% HCT and imaged at 40x in plastic wells at time 0. PBS was removed and 100 μL of each clinical fluid formulation was added to the wells and images were then obtained over time. RBC circularity was calculated using custom-written scripts in Matlab (Figure 1D). Results: RBC transit times in both healthy and sickle blood were affected by osmolarity and the various solute concentrations (Figures 2A,B). Transit times of sickle RBCs in all IV fluid formulations were significantly higher, over 10 times, than that of RBCs from healthy donors. Transit times for both healthy and sickle donors were least in the D5 ¼ NS solution. Of note, sickle RBC transit time was greatest in the NS solution. Sickle RBC circularity also changed with solute concentration and osmolarity with statistical significance (Figure 2C). Cells with the highest change in circularity from baseline were also those exposed to the D5 ¼ NS solution. Conclusion: Our results suggest the stiffness of sickle RBCs is affected by different formulations of clinical IV fluids. Increased transit time of sickle RBCs in NS through our device may in part be explained by the decreased circularity, indicating that RBCs adopt more irregular shapes in this fluid. This, in turn, could lead to increased propensity of microchannel obstruction. Although the exact mechanisms are unclear, this begs the question of whether NS is an appropriate choice for initial fluid resuscitation for VOC and other SCD-related complications as it could exacerbate the already high stiffness and shape irregularity of sickle RBCs, further increasing microvascular occlusion. As these in vitro results have significant clinical implications, ongoing experiments involve investigating how these fluid-dependent effects may alter sickle RBC adhesion in 'endothelialized' microfluidic devices, how different oxygen tensions affect these fluid-mediated effects, potential differences on and off of hydroxyurea, and the underlying mechanisms of this IV fluid formulation-dependent effect. Disclosures No relevant conflicts of interest to declare.

Description: Key Points The geometric orientation of the underlying matrix regulates platelet α-granule secretion. On geometrically constrained matrices, platelets self-deposit additional matrix, providing more cell membrane to extend spreading.

Description: Abstract 1071 Introduction: Hemostatic and thrombotic processes are dependent on platelets, coagulation factors, endothelial cells and hemodynamic flow. Current in vitro assays, however, only encompass one or two of these variables, rendering their results difficult to extrapolate to the in vivo setting. To that end, we have further refined our previously published endothelialized microfluidic system for studying thrombotic processes (Tsai et al, JCI, 2012) to specifically incorporate simultaneous differential flow rates spanning venous (10 s−1), capillary (100 s−1), and arterial/arteriolar (1000 s−1) flow conditions. Overall, key advantages of our system include: 1) successful integration of whole blood, an intact endothelium, and hemodynamic flow in a single microfluidic device, 2) simultaneous differential flow rates in a single experiment spanning 3 orders of magnitude, 3) use of corn trypsin inhibitor (CTI) as the sole anticoagulant, enabling calcium dependent processes to occur, and 4) minimal sample volume (1–2 mL) even for high shear conditions. We have applied our microsystem to elucidate some of the underlying mechanistic differences between venous and arterial thrombosis. Methods: We used photolithographic and microfabrication techniques. We previously employed to develop the silicone-based microfluidic device and applied our optimized protocol to culture human endothelial cells (HUVECs) to confluency throughout the entire inner surfaces of the system (Tsai et al, JCI, 2012 and Myers et al, JoVE, 2012) (Figure 1A, 1B). Once HUVECs are successfully cultured, whole blood with 5 % v/v fluorescently-labeled fibrinogen and cell membrane dyes is flowed into the microfluidic system. Our device can then be “activated” to induce 3 different simultaneous shear conditions (10 s−1, 100 s−1 and 1000 s−1) in 3 separate endothelialized microchannels by differentially varying the hydrodynamic resistance in each microchannel. Figure 1C shows a fibrin network that is formed under flow conditions (with whole blood anticoagulated with CTI) in one of the endothelialized microchannels. Results: Elevated levels of prothrombin are known to increase risk of venous, but not arterial thrombosis. We applied our novel system to investigate whether differences between arterial and venous thrombosis risk depend, at least in part, on differences in blood flow/shear in those vessels. By adding 1.38 μM of prothrombin to whole blood with CTI and flowing it into our system, we observed platelet-rich and fibrin rich thrombi form and occlude the 10 s−1 and 100 s−1 microchannels, but not the 1000 s−1 microchannel. In contrast, whole blood with CTI without prothrombin yielded minimal fibrin formation on the endothelial cells with no platelet adhesion/aggregation detected in all 3 shear conditions. Conclusions: Our results suggest that arterial and venous thrombosis risk is due, at least in part, to the differences in shear flow and highlights the utility of our novel endothelialized microfluidic system with simultaneous differential flow rates. Coupled with complementary in vivo experiments, our system is a powerful tool to investigate the underlying mechanisms of hemostatic and thrombotic processes involving flow. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 3300 Background: Hemostasis is an important physiologic process that requires the aggregation of platelets at distinct sites of vascular injury to promote clot formation and prevent blood loss. Platelet response to soluble agonists and shear stress has been studied extensively, but little is known of how microenvironmental geometry affects platelet function. As platelets must quickly adhere to, aggregate, and initiate coagulation only at the affected areas, spatial cues must at some level regulate this process. This aspect of spatial regulation has been investigated under static conditions by our group and others (Kita et al., 2011; Van de Walle et al., 2012). Understanding this aspect of platelet function is vital for better understanding the process of hemostasis and pathophysiological conditions such as thrombosis. Here, we directly examine how spatial cues affect platelet aggregation and physiology under variable shear conditions by flowing heparinized whole blood over micropatterned collagen in a microfluidic channel. This system allows us to assess platelet aggregate morphology under different geometric constraints and shear rates, as well as evaluate platelet physiology at the single cell level by measuring calcium signaling using fluorogenic dyes. Results: A microfluidic channel was bonded to a glass coverslip stamped with FITC-conjugated Type I collagen using a novel technique combining microcontact printing and the stamp-stick bonding technique (Satyanarayana et al., 2005). Before flowing, each chamber was incubated with 1% bovine serum albumin (BSA) blocking solution for 1.5 hours. Whole heparinized blood was then flowed through the chamber at shear rates of 100, 1000, and 10000 s−1. Platelets were labeled with Fura Red, and time lapse confocal imaging was performed for 10 minutes to monitor the aggregation of platelets at the start of flow. The flow chambers were then flushed with Tyrode's buffer with 0.1% BSA using the same experimental shear rates until the chamber was cleared of red blood cells. Image analysis was conducted using ImageJ (to calculate the percentage of platelet coverage on the collagen stamps at different shear rates. Platelets initially adhere to the distal edge of the collagen micropatterns for all shear rates (Fig. 1), indicating that platelets may require a priming region before forming a stable adhesion. As shear rate increased, platelet coverage of the collagen stamps decreased. However, aggregates also grew without conforming to the geometric constraints imposed by the collagen micropatterns more frequently at those higher shear rates (Fig. 1). Though platelet tethers generally aligned in the direction of the flow, increased tether lengths could be seen when platelets were exposed to higher shear, which may explain why platelets were able to span larger gaps and aggregate in a less spatially constrained manner at high shear rates. Image analysis shows that 51.5% of the collagen was covered by platelet aggregates for a shear rate of 100 s−1 with some platelets forming tethers to span gaps (Fig. 2). When the shear rate was increased to 1000 s−1, platelet coverage of the collagen microstamp drastically dropped to 18.5% (Fig. 2). At a pathophysiological shear rate of 10000 s−1, the percentage of collagen covered by platelets dropped further to 12.8% and adopted a linear shape, although a large portion of the aggregate can be seen spanning gaps between the collagen microstamp (Fig. 1 and 2). Conclusions and Ongoing Efforts: Ours is the first reported study of the spatial regulation of platelet aggregation under variable shear in a microfluidic channel. We have found that platelets are more spatially regulated under low shear conditions compared to high shear, which has implications for thrombosis and other clotting disorders. Future studies will incorporate the simultaneous use of ratiometric fluorgenic calcium signaling dyes to investigate the role of spatial regulation in Ca2+ signaling. Finally, we have developed a method to culture endothelial cells around a collagen micropattern to study this spatial regulation of platelet function under more physiological conditions (Fig. 3). Disclosures: No relevant conflicts of interest to declare.