A collagen network phase improves cell seeding of open-pore structure scaffolds under perfusion

Scaffolds with open-pore morphologies offer several advantages in cell-based tissue engineering, but their use is limited by a low cell seeding efficiency. We hypothesized that inclusion of a collagen network as filling material within the open-pore architecture of polycaprolactone-tricalcium phosphate (PCL-TCP) scaffolds increases human bone marrow stromal cells (hBMSC) seeding efficiency under perfusion and in vivo osteogenic capacity of the resulting constructs. PCL-TCP scaffolds, rapid prototyped with a honeycomb-like architecture, were filled with a collagen gel and subsequently lyophilized, with or without final crosslinking. Collagen-free scaffolds were used as controls. The seeding efficiency was assessed after overnight perfusion of expanded hBMSC directly through the scaffold pores using a bioreactor system. By seeding and culturing freshly harvested hBMSC under perfusion for 3 weeks, the osteogenic capacity of generated constructs was tested by ectopic implantation in nude mice. The presence of the collagen network, independently of the crosslinking process, significantly increased the cell seeding efficiency (2.5-fold), and reduced the loss of clonogenic cells in the supernatant. Although no implant generated frank bone tissue, possibly due to the mineral distribution within the scaffold polymer phase, the presence of a non crosslinked collagen phase led to in vivo formation of scattered structures of dense osteoids. Our findings verify that the inclusion of a collagen network within open morphology porous scaffolds improves cell retention under perfusion seeding. In the context of cell-based therapies, collagen-filled porous scaffolds are expected to yield superior cell utilization, and could be combined with perfusion-based bioreactor devices to streamline graft manufacture.

Molecular Structure of the mineral woodhouseite CaAl3(PO4,SO4)2(OH)6

The mineral woodhouseite CaAl3(PO4,SO4)2(OH)6 is a hydroxy phosphate-sulphate mineral belonging to the beudantite subgroup of alunites, and has been characterised by Raman spectroscopy, complimented with infrared spectroscopy. Bands at various wavenumbers were assigned to the different vibrational modes of woodhouseite, which were then associated to the molecular structure of the mineral. Bands were primarily assigned to phosphate and sulphate stretching and bending modes. Two symmetric stretching modes for both phosphate and sulphate supported the concept of non-equivalent phosphate and sulphate units in the mineral structure. Bands in the OH stretching region enabled hydrogen bond distances to be calculated.

Infinite-horizon policy-gradient estimation

Gradient-based approaches to direct policy search in reinforcement learning have received much recent attention as a means to solve problems of partial observability and to avoid some of the problems associated with policy degradation in value-function methods. In this paper we introduce GPOMDP, a simulation-based algorithm for generating a biased estimate of the gradient of the average reward in Partially Observable Markov Decision Processes (POMDPs) controlled by parameterized stochastic policies. A similar algorithm was proposed by Kimura, Yamamura, and Kobayashi (1995). The algorithm's chief advantages are that it requires storage of only twice the number of policy parameters, uses one free parameter β ∈ [0,1) (which has a natural interpretation in terms of bias-variance trade-off), and requires no knowledge of the underlying state. We prove convergence of GPOMDP, and show how the correct choice of the parameter β is related to the mixing time of the controlled POMDP. We briefly describe extensions of GPOMDP to controlled Markov chains, continuous state, observation and control spaces, multiple-agents, higher-order derivatives, and a version for training stochastic policies with internal states. In a companion paper (Baxter, Bartlett, & Weaver, 2001) we show how the gradient estimates generated by GPOMDP can be used in both a traditional stochastic gradient algorithm and a conjugate-gradient procedure to find local optima of the average reward. ©2001 AI Access Foundation and Morgan Kaufmann Publishers. All rights reserved.