High Cell Density Propionic Acid Fermentation with an Acid Tolerant Strain of Propionibacterium acidipropionici Zhongqiang Wang, Ying Jin, Shang-Tian Yang William G. Lowrie Department of Chemical & Biomolecular Engineering, The Ohio State University, 140W. 19th Ave, Columbus, OH 43210; telephone: þ1-614-292-6611; fax: þ1-614-292-3769; e-mail: yang.15@osu.edu ABSTRACT: Propionic acid is an important chemical with wide applications and its production via fermentation is of great interest. However, economic production of bio-based propionic acid requires high product titer, yield, and productivity in the fermentation. A highly efficient and stable high cell density (HCD) fermentation process with cell recycle by centrifugation was developed for propionic acid production from glucose using an acid-tolerant strain of Propionibacterium acidipropionici, which had a higher specific growth rate, productivity, and acid tolerance compared to the wild type ATCC 4875. The sequential batch HCD fermenta- tion at pH 6.5 produced propionic acid at a high titer of 40 g/L and productivity of 2.98 g/L h, with a yield of 0.44 g/g. The product yield increased to 0.53–0.62 g/g at a lower pH of 5.0–5.5, which, however, decreased the productivity to 1.28 g/L h. A higher final propionic acid titer of >55 g/L with a productivity of 2.23 g/L h was obtained in fed-batch HCD fermentation at pH 6.5. A 3-stage simulated fed-batch process in serum bottles produced 49.2 g/L propionic acid with a yield of 0.53 g/g and productivity of 0.66 g/L h. These productivities, yields and propionic acid titers were among the highest ever obtained in free-cell propionic acid fermentation. Biotechnol. Bioeng. 2015;112: 502–511. ß 2014 Wiley Periodicals, Inc. KEYWORDS: acid tolerant strain; high cell density fermentation; propionic acid; Propionibacterium acidipropionici; sequential batch fermentation Introduction Propionic acid is a 3-carbon fatty acid with various industrial applications, including uses as preservatives in animal feed and dairy and bakery products, and in manufacturing herbicide intermediate and cellulose acetate propionate. Currently, propionic acid is mainly produced by petrochemi- cal-based processes, such as the oxo process with ethylene and carbon monoxide as feedstocks. With the depletion of petroleum, rising oil price, environmental burden of fossil fuel production and public’ s preference for biobased chemicals, the production of propionic acid from renewable biomass via fermentation has gained large attention in recent years. Many bacteria are capable of producing propionic acid. Particularly, the species in the genus of Propionibacterium, including P. acidipropionici, P. freudenreichii, and P. shermanii that have been widely used in industrial production of Swiss cheese and vitamin B 12 (Gardner and Champagne, 2005), have been intensively studied for propionic acid fermentation (Barbirato et al., 1997; Dishisha et al., 2012; Huang et al., 2002; Liang et al., 2012; Liu et al., 2011; Paik and Glatz, 1994; Wang and Yang, 2013; Yanget al., 1994, 1995; Zhu et al., 2012). However, propionic acid fermentation usually suffers from low productivity, yield and titer due to the strong inhibition of propionic acid and the coproduction of acetate and succinate, making it difficult to compete with petrochemical processes. Numerous research efforts have thus focused on increasing the reactor productivity by elevating viable cell density (see Table I) or removing propionic acid in situ to alleviate product inhibition (Jin and Yang, 1998; Lewis and Yang, 1992; Yang et al., 2006). Cell recycle, retention and immobilization are three efficient ways to achieve high density of active cells. Continuous fermentation processes with cell retention using an in situ spin filter (Gupta and Srivastava, 2001) or cell recycle via an external ultrafiltration unit have been used to increase cell density and reactor productivity, achieving the highest volumetric productivity of 14.3 g/L h ever reported (Boyaval and Corre, 1987). However, these continuous fermentation processes were not stable, because of membrane fouling, and suffered from a low final product titer, usually less than 20 g/L, which would greatly increase the cost for downstream product recovery. The high equipment and operation costs with ultrafiltration and spin filter for continuous cell recycle or retention also limit their Zhongqiang Wang and Ying Jin contributed equally to this paper. Correspondence to: S.-T. Yang Contract grant sponsor: The Dow Chemical Company Received 16 May 2014; Revision received 28 July 2014; Accepted 16 September 2014 Accepted manuscript online 25 September 2014; Article first published online 21 October 2014 in Wiley Online Library (http://onlinelibrary.wiley.com/doi/10.1002/bit.25466/abstract). DOI 10.1002/bit.25466 ARTICLE 502 Biotechnology and Bioengineering, Vol. 112, No. 3, March, 2015 ß 2014 Wiley Periodicals, Inc.