129 I. E. YATES 1 *, K. L. HIETT 1 , D.R.KAPCZYNSKI 1 , W. SMART 2 , A. E. GLENN 1 , D. M. HINTON 1 , C. W. BACON 1 , R.MEINERSMANN 1 , S. LIU 3 AND A.J.JAWORSKI 2  Richard B. Russell Agricultural Research Center, USDA Agricultural Research Service, P.O. Box 5677, Athens, Georgia 30604-5677, U.S.A.  Department of Botany, University of Georgia, Athens 30501, U.S.A.  Clemson University, Clemson, South Carolina 29631, U.S.A. Visual markers detectable by histochemical staining have been developed for analysing the time course and tissue specificity of maize infections by Fusarium moniliforme. Three F. moniliforme strains, RRC 374, MRC 826 and RRC PAT, were transformed with a plasmid, pHPG, containing the gusA reporter gene which codes for β-glucuronidase (GUS) and the hph gene for hygromycin resistance as the selectable marker. Introduction of plasmid DNA into germinating conidia yielded 1210 - transformants per conidium ; expression of both gusA and hph was however, transient. Stable transformants were obtained using protoplasts as the recipient, but transformation frequency was reduced. Southern blot and PCR analyses confirmed incorporation of pHPG into the genome of all three F. moniliforme strains with gusA properly inserted in MRC 826 and RRC PAT, but apparently disrupted in RRC 374. The growth pattern for transformed F. moniliforme isolates and the parental wild types followed a sigmoid curve on minimal and enriched media. Hygromycin totally inhibited growth for wild type isolates, but not of transformants. Transformed isolates maintained the ability to infect the maize plant. Thus, this study is the first report of F. moniliforme transformed with a visibly detectable reporter gene to use for analysing this endophyte-host interaction of world-wide importance to animal and human health. Fusarium moniliforme J. Sheld. [teleomorph  Giberella fujikuroi (Sawada) Ito in Ito & K. Kimura] is a problem in maize because the fungus can produce mycotoxins which are detrimental to animal and human health (Sydenham et al., 1990 ; Norred, 1993 ; Bacon & Nelson, 1994). Disease symptoms are not always visible in F. moniliforme infected maize (Leslie et al., 1990 ; Bacon & Hinton, 1996 a). The fungus can grow as an endophyte without causing visible disease symptoms, i.e., symptomless or asymptomatic infection. Plants with visible disease are not usually sold as food or feed products, but plants with symptomless infections may enter the food chain for animal andor human consumption (Bacon & Hinton, 1996 a). The absence of precise markers to track the fungus in planta dictates using indirect methods to study interactions between fungal endophytes and their host plants (Bills, 1996). Microscopy provides evidence for the presence of a fungus within the plant, but not fungal identity, origin, metabolic status, nor biomass. Furthermore, analysis of all plant tissues in an experimental series is a laborious, time-consuming task. Fungal isolation from plant sections cultured on growth medium, coupled with vegetative compatibility tests, is more conclusive for identification (Correll, Klittich & Leslie, 1987 ; Kedera, Leslie & Claflin, 1994). This methodology, however, shares some limitations with microscopy, such as indefinite assessment of the fungal origin, metabolic status, and biomass. Specific identification of races by serological techniques has been reported for some fungi, including F. oxysporum (Wong, White & Wright, 1988 ; Dickman & Partridge, 1989). Even a combination of microscopy, isolation, compatibility assess- ment and serology provides only indirect evidence on the F. moniliforme–maize relationship. Thus, methodology is lacking for obtaining direct evidence on the tissue specificity and metabolic activity of F. moniliforme during maize infections. Bacteria, fungi and plants have been transformed with a molecular marker using the Escherichia coli gene encoding β- glucuronidase (GUS), which is detectable by histochemical and fluorometric enzymatic assays (Jefferson, 1987 ; Gallagher, 1992). Low endogenous GUS activity in most plants and fungi, and rapid assays for detecting GUS activity, have accounted for the wide application of this method. GUS expression has been used to monitor root colonization by Bipolaris sorokiniana in barley (Liljeroth, Jansson & Scha  fer, 1993) and F. oxysporum in flax (Couteaudier et al., 1993; Eparvier & Alabouvette, 1994). Tagging F. moniliforme with an easily detectable marker gene would provide a system to directly monitor maize infections. The objectives of our study were to transform F. moniliforme with the gusA gene, and to compare transformed colonies with the parent wild type for growth and GUS expression in culture and in planta. MATERIALS AND METHODS Strains and media Recipients for vector DNA were three fumonisin-producing F. moniliforme strains, RRC 374, MRC 826 and RRC PAT, of Mycol. Res. 103 (2) : 129–136 (1999) Printed in the United Kingdom GUS transformation of the maize fungal endophyte Fusarium moniliforme