http://biotech.nature.com • MAY 2001 • VOLUME 19 • nature biotechnology Transgenic mice expressing bacterial phytase as a model for phosphorus pollution control Serguei P. Golovan 1,2 , M. Anthony Hayes 3 , John P. Phillips 2 *, and Cecil W. Forsberg 1 * We have developed transgenic mouse models to determine whether endogenous expression of phytase transgenes in the digestive tract of monogastric animals can increase the bioavailability of dietary phytate, a major but indigestible form of dietary phosphorus. We constructed phytase transgenes composed of the appA phytase gene from Escherichia coli regulated for expression in salivary glands by the rat R15 proline- rich protein promoter or by the mouse parotid secretory protein promoter. Transgenic phytase is highly expressed in the parotid salivary glands and secreted in saliva as an enzymatically active 55 kDa glycosylat- ed protein. Expression of salivary phytase reduces fecal phosphorus by 11%. These results suggest that the introduction of salivary phytase transgenes into monogastric farm animals offers a promising biological approach to relieving the requirement for dietary phosphate supplements and to reducing phosphorus pollu- tion from animal agriculture. RESEARCH ARTICLE Animal waste is a leading source of phosphorus pollution from agri- culture 1 . Manure from monogastric animals such as poultry and swine is high in phosphorus, and when manure is repeatedly applied as fertilizer, phosphates can pollute surface and groundwater with severe biological consequences. Because phosphate is a limiting nutrient that restricts microbial populations in many freshwater environments 2 , the influx of phosphorus can lead to eutrophication. The results are cyanobacterial blooms, hypoxia and death of fish and aquatic animals 3 , and the production of nitrous oxide, a potent greenhouse gas 4 . The projected growth of the livestock industry 5 is expected to accelerate such environmental problems on a global scale. It is critical that agricultural practices be modified to reduce such environmental impacts 6 . The high phosphorus content of manure from monogastric ani- mals arises from the inability of these animals to hydrolyze phytate 1 , the major form of organic phosphate present in the typical plant- based diet 7 . The nutritional requirements for phosphorus needed to attain optimal growth in swine and poultry have traditionally been met through dietary supplementation with inorganic phosphate. This approach has been nutritionally successful but environmentally coun- terproductive. A more environmentally sound but far less common approach is the use of a microbial phytase as a feed additive. Phytase hydrolyzes phytate, and the addition of phytase to feed (250–1000 U kg -1 ) can fully replace phosphorus supplements at all stages of pig production 8,9 . However, the use of phytase as a feed addi- tive is limited by cost, by inactivation at the high temperatures required for pelleting feed (+80°C), and by loss of activity during storage. These problems might be overcome if phytase were added to the repertoire of digestive enzymes produced endogenously by swine and poultry. Such endogenous phytase could increase the bioavail- ability of plant phytate and in turn lead to reduced phosphorus output from animal production. To test the feasibility of this hypothesis, we produced transgenic mice that secrete phytase in their saliva. The transgenes used in these studies contain the E. coli appA gene 10,11 , which was recently shown to be effective in poultry 12 , regulated either by the inducible proline-rich protein (PRP) R15 promoter from the rat 13 or the constitutive parotid secretory protein (PSP) promoter from the mouse 14 . The salivary appA phytase pro- duced in these transgenic mice leads to a significant reduction of fecal phosphorus levels. Results Production of transgenic mice. The appA gene from E. coli was inserted downstream of the salivary-specific promoters R15-PRP and PSP, to obtain the inducible R15/APPA (Fig. 1A) and the consti- tutive PSP/APPA constructs (Fig. 1B). Because phytase can dephosphorylate inositol phosphates, some of which are involved in essential cellular functions 15 ,our first model was designed with the inducible R15 promoter, in order to evaluate the possible deleterious effects of phytase expression on the animal. With the R15/APPA transgene, eight transgenic founder (G 0 ) mice (four males and four females) were obtained, of which three did not pass the transgene to progeny and were probably mosaics. In the remaining five lines, phytase expression was induced by isopro- terenol injection 13 . Isoproterenol injections increased the size of the salivary glands, an observation reported previously 16 . Phytase expression was not detected in either uninduced mice or in induced nontransgenic mice. A wide range of phytase expression was observed in the various R15/APPA transgenic lines (Table 1). We did not detect any deleterious effect of phytase expression on mice. Induced transgenic animals were fertile, producing both male and female offspring. As a second model, transgenic mice were generated using the con- stitutive PSP/APPA transgene (Table 1). Two transgenic founders (male and female) were produced. Despite the detection of the trans- gene by PCR, no phytase production was detected in saliva of the male founder or in his G 1 offspring. The female founder produced a single transgenic G 1 male out of 35 offspring. Phytase activity (30 U/ml) was detected in the saliva of both the G 0 female founder and the G 1 male offspring. The number of newborns in the G 2 gener- ation, their gender, and their viability were the same in the transgenic and nontransgenic animals, further documenting the transgene’s lack of toxicity. Expression of the phytase transgene in salivary glands. Northern blot analysis demonstrated very strong expression of phytase mes- senger RNA (mRNA) in parotid glands and a fivefold lower expres- Department of Microbiology 1 , Department of Molecular Biology and Genetics 2 , Department of Pathobiology 3 , University of Guelph, Guelph, Ontario N1G 2W1, Canada. *Corresponding authors (cforsber@uoguelph.ca or jphillip@uoguelph.ca). 429 © 2001 Nature Publishing Group http://biotech.nature.com © 2001 Nature Publishing Group http://biotech.nature.com