Page 91 - IJB-2-1
P. 91

Wen Shing Leong, Shu Cheng Wu, Kee Woei Ng, et al.

                                                               examined to study the effect of pore size of electros-

                                                               pun scaffold on cellular  distribution  in  scaffold  after
                                                               cell seeding. Both  types of scaffolds were surface
                                                               modified with the chosen surface modification method
                                                               demonstrated in Section 3.1, and  compared to that
                                                               without surface  modification. As discussed  earlier,
                                                               surface modification significantly enhanced hydrophi-
                                                               licity of the PCL scaffold  without deteriorating  the
                                                               architectural properties. As shown in  Figure 5, 3D
            Figure 3. Mercury porosimeter revealed a four-fold increment
            in  pore size in 3D  scaffold  in  comparison  with  2D  scaffold,   multi-scale electrospun PCL improved in water ab-
            with little change in porosity.                    sorption  and  expanded in  phosphate  buffer  solution
                                                               only after effective surface modification.
            slanted needle opening instead of the needle tip.   Cell culture results as shown in  Figure 6  revealed
            Eventually, newly spun fibers would wrap around the   the difficulty for HDFs to  be seeded  into  traditional
            needle tip and develop into a cotton-like scaffold. The
            scaffold thickness collected within 30  minutes was   2D  electrospun  scaffold.  Traditional  2D  PCL  elec-
            more than 6 mm and this is about 75 times more than   trospun scaffold had desired porosity for tissue engi-
            that in 2D electrospun PCL scaffold. On the other   neering but the dense fiber packing resulted in small
            hand, existence of a mixture of micro- and nano- fiber   pore size which restricted cell to be seeded throughout
            could play a part in disrupting fibers packing, result-  the whole scaffold (Figure 6A). Even with the aid of
            ing in an increase of the pore size of the scaffold [19–22] .   gelatin  grafting  to  improve  wettability  and  cell-sca-
            Larger pore size offers higher opportunities for cell   ffold interaction, no significant improvement in cellu-
            infiltration  and  mass transfer without  sacrificing  the   lar distribution was observed. HDFs seeded on gelatin
            ECM mimicry nanofeatures.                          grafted 2D electrospun scaffold were found to adhere
                                                               on the top surface only, despite the enhanced wettabil-
            3.4 Comparison of HDFs Distribution Between 2D     ity (Figure 6B). This is a common issue that has re-
            Electrospun and 3D Multi-scale Scaffold                                                   [9,10]
                                                               stricted the application of electrospun scaffold  . By
            Both 2D electrospun and 3D multi-scale scaffold were   increasing the pore  size of the scaffold using needle

                                                                                 100 m/s





























            Figure 4. Electrospinning setup. (A) Collection of 2D electrospun scaffold on plate collector; (B) Collection of 3D electrospun scaf-
            fold on needle collector.
                                        International Journal of Bioprinting (2016)–Volume 2, Issue 1      87
   86   87   88   89   90   91   92   93   94   95   96