ARTICLE Using an Aerosol Deposition Model to Increase Hairy Root Growth in a Mist Reactor Melissa J. Towler, 1 Barbara E. Wyslouzil, 2 Pamela J. Weathers 1 1 Department of Biology/Biotechnology, Worcester Polytechnic Institute, Worcester, Massachusetts 01609; telephone: 508-831-5196; fax: 508-831-5936; e-mail: weathers@wpi.edu 2 Department of Chemical and Biomolecular Engineering, The Ohio State University, Columbus, Ohio 43210 Received 15 September 2005; accepted 13 July 2006 Published online 28 July 2006 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/bit.21143 ABSTRACT: Gas-phase reactors, including the mist reactor, have distinct advantages over liquid-phase reactors includ- ing the ability to manipulate the gas composition, to allow effective gas exchange in a densely growing biomass, and to affect secondary metabolite production. Mathematical mod- eling suggested that roots in a mist reactor are often too sparsely packed to capture mist particles efficiently and cannot, therefore, meet the nutrient demands required to maintain high growth rates. Indeed, growth rates of Arte- misia annua hairy roots increased significantly when the initial packing density increased or when a higher sucrose concentration was used in the medium. Growth kinetics for 2, 4, and 6 days, however, showed a decrease or stationary growth rate after only 4 days for both 3 and 5% sucrose feeds. Residual medium analyses indicated that carbon was not exhausted, nor were any of the other major nutrients including phosphate. Increasing the mist duty cycle at constant carbon flux through the reactor reduced the growth rates slightly. In general, the aerosol deposition model correctly predicted how to optimize hairy root growth in the mist reactor. Biotechnol. Bioeng. 2007;96: 881–891. ß 2006 Wiley Periodicals, Inc. KEYWORDS: bioreactor; aeroponics; Artemisia annua Introduction Plants produce many useful and commercially interesting secondary metabolites, and in vitro culture of transformed (hairy) roots has been proposed as a potential source of these important compounds (Flores and Curtis, 1992). Bioreactor design for transformed root culture must, however, consider the morphological characteristics of the tissue—in particular the shear sensitivity and the high oxygen demand of the rapidly growing root tips (Ramakrishnan and Curtis, 1995). The work in our research group has focused on mist reactors, because this environment reduces the gas-exchange limita- tions and shear conditions normally found in liquid-phase reactors. Furthermore, establishing a hairy root culture on solid medium does not necessarily guarantee that these roots will continue to grow when they are transferred to liquid medium (Hallard et al., 1997) and gas-phase reactors may be the only option. High biomass density is often required for an economic- ally viable reactor system, and the maximum root tissue concentration that can be achieved in a bioreactor depends on the delivery of oxygen and other nutrients into the dense root matrix (Curtis, 2000). McKelvey et al. (1993) showed that roots are more able to compensate for poor liquid dispersion than for poor gas dispersion within a reactor (Curtis, 1993). In our earlier work (Weathers et al., 1999) we used alcohol dehydrogenase (ADH) mRNA levels in A. annua hairy roots as a marker of oxygen deprivation and found that roots grown in bubble columns or shake flasks showed ADH mRNA expression at packing densities as low as 6% (v/v; a ¼ 0.06). In contrast, roots grown in mist reactors were not oxygen limited at root packing fractions as high as a ¼ 0.37 (Kim, 2001). The packing fraction a is the volume fraction, and equals the fresh weight concentration (g L 1 )  0.001 in the growth chamber since the mass density of roots is 1 g mL 1 . As anaerobic respiration increases, the respiratory quotient—the ratio of CO 2 produced to O 2 consumed— also increases (Bordonaro and Curtis, 2000). When Bordonaro and Curtis (2000) compared the respiratory quotients of Hyoscyamus muticus hairy roots grown in either a bubble column reactor or a trickle bed reactor, they noted Correspondence to: P.J. Weathers Contract grant sponsor: DOE Contract grant number: 1R21AI3917-01, 1R15GM069562-01 Contract grant sponsor: NIH Contract grant number: P200A50010-95 ß 2006 Wiley Periodicals, Inc. Biotechnology and Bioengineering, Vol. 96, No. 5, April 1, 2007 881