Using Ustilago maydis as a Trojan Horse for In Situ Delivery of Maize Proteins

The biotrophic pathogen Ustilago maydis is the causative agent of the corn smut disease¹. It infects all aerial parts of maize resulting in large tumors that contain melanized, black spores. On the global level, U. maydis is estimated to cause an annual loss of around 2% of corn yield, while tumors are appreciated as a gastronomical delicacy in Mexico. Plant infection is initiated by an appressorium that secretes cell-wall lysing enzymes to penetrate the first layer of maize epidermal cells. From a primary infection site, U. maydis grows intracellularly and intercellularly, invading one to two cell layers every day^1,2. Successful infection results in plant hypertrophy that turns into visible tumors upon five days post infection^1,3,4. During all infection stages, fungal hyphae invaginate the plant cytoplasm membrane without any direct contact to the host cytoplasm^1,2. The tight apoplasmic space between the infecting hyphae and the plant plasma membrane is considered to be the host/pathogen interactive site, called the biotrophic interaction zone. In order to overcome the plant innate immune system, U. maydis secretes an array of effector proteins into the biotrophic interaction zone¹. Some effectors are taken up by plant cells, while others remain in the biotrophic interaction zone^5,6,7,8. One apoplastic effector is UmPit2, which interacts with apoplastic maize proteases to prevent the release of the signaling peptide ZmZIP1 from ZmPROZIP by apoplastic protease activity^9,10.

Over the last decades, U. maydis became not only a model for fungal genetics in plant-pathogen interaction, but also a valuable tool in biotechnology due to a well-understood life cycle, easy genetic accessibility and heterologous expression of secreted proteins^11,12,13. Signals for both conventional and unconventional protein secretion have been determined allowing the control of posttranslational modifications¹⁴. Recently, U. maydis was employed as a Trojan horse tool to study small, secreted maize proteins in situ¹⁵. The Trojan horse approach was successfully used to analyze the function of the small, secreted protein ZmMAC1 that is involved in anther development. ZmMAC1 induces the periclinal division of pluripotent cells and cell fate specification of the newly formed cells¹⁵. By the same method, the biological function of the maize damage-associated peptide ZmZIP1 was revealed. U. maydis secreting the maize ZmZIP1 resulted in impaired tumor formation¹⁰. Thus, the Trojan horse approach represents a valuable alternative route to protein in situ studies with high spatiotemporal resolution that does neither require generation of stable maize transformation lines nor tissue infiltration with heterologously expressed and purified proteins. In particular, the Trojan horse strategy enables the secretion of any heterologous protein into the maize apoplast and direct comparison of infected versus non-infected plant cells within the same tissue.

This protocol illustrates the major steps for generating an U. maydis Trojan horse strain to study a protein of interest. It further includes precise information on infection procedures of three different maize tissue types (adult leaves, tassels and ears) with U. maydis, which is a prerequisite for studying the spatiotemporal infection progression and protein function in these target tissues. No further specifications are given on maize gene amplification and microscopic imaging techniques, since these steps are target-specific and instrument-dependent. Thus, this protocol is addressed to experienced users of standard molecular biology techniques.

1. Construction of an U. maydis Trojan Horse

NOTE: See Figure 1.

1. Amplify a gene of interest from maize cDNA using gene-specific primers and a proofreading DNA polymerase. Clone the primary PCR product and transform the construct into E. coli following the plasmid vendor's instructions. Verify the correct gene of interest sequence by Sanger sequencing prior to use for the next cloning steps.
   NOTE: PCR specifications need to be optimized due to primer sequence specificities and optimal DNA polymerase reaction conditions.
2. Design primers to amplify the maize gene of interest without the sequence encoding a signal peptide (SP).
3. Extend the 5´ end of the reverse primer with the RSIATA motif and an NcoI cutting site (Table 1).
4. Amplify the maize gene of interest with a proofreading DNA polymerase using the PCR construct generated in 1.1 as the PCR template.
5. Double-digest the PCR product and the U. maydis transformation plasmid, p123-P_Umpit2-Sp_Umpit2-Zmmac1-mCherry-Ha¹⁵, with XbaI and NcoI. Purify the digested PCR product and plasmid.
6. Ligate the digested PCR product into the p123-P_Umpit2-Sp_Umpit2-Zmmac1-mCherry-Ha template using T4 DNA ligase following the manufacturer´s instructions. Transform the ligation product into E. coli and verify the correct gene of interest sequence by Sanger sequencing.
7. Linearize the p123-P_Umpit2-Sp_Umpit2-Zmgene of interest-mCherry-Ha with the restriction enzyme SspI and transform DNA into the solopathogenic U. maydis strain SG200¹⁶. Isolate U. maydis transformants by carboxin selection and confirm isolated transformants by Southern blot analysis¹⁶.
   NOTE: For each protein of interest, at least three independent U. maydis transformants should be isolated and analyzed to estimate any phenotype effects of random background mutations.

2. Culture Media

1. Prepare YEPSlight liquid medium¹⁶: 1.0% (w/v) yeast extract, 0.4% (w/v) Bacto-Peptone, and 0.4% (w/v) sucrose. Dissolve all components in ddH₂O and autoclave at 121 °C for 15 min; autoclaving at a higher temperature, for a longer period of time or repeatedly would reduce the quality of the medium.
2. Prepare potato-dextrose-agar (PD-agar)²⁰: 3.9% (w/v) potato dextrose agar, and 1.0% (w/v) 1 M Tris-HCl pH 8.0 (f.c. 0.01 M). Mix all components directly in the bottle for autoclaving and add ddH₂O plus a magnetic stir bar. Autoclave at 121 °C for 15 min; autoclaving at a higher temperature, for a longer period of time or repeatedly would reduce the quality of the medium.
3. Prepare PD-Charcoal agar²⁰: 3.9% (w/v) potato dextrose agar, 1% (w/v) charcoal, and 1.0% (v/v) 1 M Tris-HCl pH 8.0 (f.c. 0.01 M). Mix all components directly in the bottle for autoclaving and add ddH₂O plus a magnetic stir bar. Autoclave at 121 °C for 15 min; autoclaving at a higher temperature, for a longer period of time or repeatedly would reduce the quality of the medium.

3. Plant Infection

1. Perform analysis of maize cell division in response to Trojan horse delivered protein (e.g., microscopy-based cell counting) beforehand. U. maydis-induced maize cell proliferation and subsequent tumor formation starts around 4-5 days post infection. Quantitative disease assessments are tissue-dependent and should be performed from 6 to 14 days post infection.
   NOTE: Distinct maize cultivars show different levels of susceptibility to U. maydis infection. Maize cvs. W23, A188, Gaspe flint, Early Golden Bantam or Va35 show susceptibility towards this pathogen and are thus suitable cultivars for Trojan horse studies.
2. Preparation of the U. maydis inoculum
   1. Include the progenitor strain SG200 as a negative control in all Trojan horse experiments in order to estimate side effects on the infection by the transgenic strain. Here, use an U. maydis strain expressing a non-secreted version of the protein of interest as a negative control. However, for reasons of practicability (e.g., larger screenings), the progenitor strain may be the easier choice of control.
   2. Before starting the experiment, estimate what amounts of infection culture are needed. Keep in mind that infection of each plant requires 1 - 1.5 mL of U. maydis suspension dependent on the maize tissue type.
      NOTE: Approximately 1 mL of an overnight culture is sufficient for dilution to an OD₆₀₀ of 0.2 in 20 mL of YEPS_light medium, and 25 mL of an U. maydis culture with an OD₆₀₀ of 0.8-1.0 are sufficient for infection of 13-16 plants.
   3. Scratch U. maydis from a PD agar plate using a sterile Pasteur pipette, inoculate in 5 mL of YEPS_light medium and let the culture grow at 28 °C with constant shaking at 200 rpm for 16 h.
   4. Prior to infection, examine the U. maydis inoculation culture by standard light microscopy for proper growth and bacterial contamination at 400X magnification.
      NOTE: In a suitable culture, only the cigar-shaped fungus is visible (Figure 2).
   5. Mix 900 µL of fresh YEPS_light with 100 µL of the overnight culture and measure the OD₆₀₀ using YEPS_light medium as a blank in the spectrophotometer analysis.
   6. Dilute the overnight culture with fresh YEPS_light medium to an OD₆₀₀ of 0.2 and let the culture grow at 28 °C with constant shaking at 200 rpm until reaching the mid-log growth phase indicated by an OD_{600 nm} of 0.8-1.0.
      NOTE: U. maydis cells duplicate every 2 h under these conditions, thus the desired OD₆₀₀ is reached after 4-5 h of cultivation.
   7. Harvest the cells at OD₆₀₀ of 0.8-1.0 by spinning at 3,000 x g for 10 min and discard the supernatant.
   8. Wash the cell pellet one time with ddH₂O. For this purpose, add one culture volume of ddH₂O, spin with 3000 x g for 10 min and discard the supernatant.
   9. Resuspend the cell pellet carefully in ddH₂O using a 20-mL glass pipette, thereby adjusting the final OD₆₀₀ to 3.0 (for a Trojan horse assay) or 1.0 (for disease rating).
3. Verification of the Trojan horse: in planta secretion of the maize fusion protein
   1. Infect maize seedlings with a Trojan horse strain¹⁷.
   2. Perform microscopic imaging of infected seedlings at 2 - 3 days post infection using a confocal laser-scanning microscope. To this end, excise a rectangular piece of the leaf 1 cm below the point of injection, place the sample onto a microscope slide and add a drop of ddH₂O. To visualize mCherry fusion protein, excite specimen at λ= 561 nm and record emission at λ= 580 - 630 nm.
4. Infection of adult maize leaves
   1. Cultivate maize plants to the stage of adult leaves (when at least leaf 7 grows within the stalk).
      NOTE: This stage is reached after four weeks upon sowing under greenhouse conditions of 14 h, 28 °C day/10 h, 22 °C night rhythm using cv. W23. Duration may vary with the maize cultivar and greenhouse conditions.
   2. Transfer the U. maydis culture (see 3.2.9) into a 3-mL syringe with a 20G x 1 hypodermic needle.
   3. Press the stalk carefully to localize the meristem tissue in the stalk. The base of the meristem can be distinguished by a transition from harder stalk to softer tissue.
   4. Mark the meristem on the stalk using a pen.
   5. Inject 1.5 mL of U. maydis culture 1 cm above the shoot meristem or inflorescence meristem.
   6. Rate disease symptoms at 6 and 12 days post infection.
5. Infection of tassels
   1. Grow maize plants until reaching the tassel stage.
      NOTE: A detailed timeline on anther and tassel development in maize cv. W23 was previously described^18,19. Tassels containing pre-meiotic anthers are highly susceptible to U. maydis infection; in the maize cv. W23 tassels are the size of 4 - 7 cm.
   2. Press the stalk carefully to localize the tassel in the stem.
   3. Mark the tip and the base of the tassel on the stem using a pen.
   4. Transfer the U. maydis culture (see 3.2.9) into a 3-mL syringe with a 20G x 1 hypodermic needle.
   5. Inject 1.5 mL of the inoculum around the tassel. To ensure equal distribution of the inoculum, slowly place 0.5 mL each at the tip, the middle part, and the base of the tassel marked with the pen.
   6. Rate disease symptoms at 10 days post infection.
6. Infection of ears
   NOTE: The ear tissue development differs in distinct maize cultivars and greenhouse conditions and must be carefully observed prior to inoculation. Ears starting to outgrow silks are highly susceptible to U. maydis infection.
   1. Transfer the U. maydis culture (see 3.2.9) into a 3 mL syringe with a 20 G x 1 hypodermic needle.
   2. Inject the inoculation needle into the space between the husk leaves as deeply as possible without injuring the ear.
   3. Release 1.5 mL of the inoculum around the ear.
   4. Remove the syringe plus needle and carefully massage the cob to distribute the U. maydis solution equally.
   5. Rate disease symptoms at 14 days post infection.
7. Confirm viability of U. maydis inoculum
   1. Drop 10 µL of the inoculum on a PD charcoal agar plate and incubate at room temperature for 2 days.
      NOTE: If the respective U. maydis culture is able to form filaments, fluffy, white mycelium becomes visible (Figure 5)

Constructs for U. maydis Trojan horse experiments are cloned into the plasmid p123-P_Umpit2-Sp_Umpit2-gene of interest-mCherry-Ha. The maize gene of interest is fused to a mCherry fluorescence reporter and an epitope HA-tag. The expression of the fusion protein is under control of the U. maydis Umpit2 promoter which is specifically activated during infection²¹. To direct secretion of the protein of interest peptide into the biotrophic interaction zone, the coding region is fused to the signal peptide sequence of U. maydis Umpit2²¹ (Figure 1). Upon U. maydis transformation, the transgene is inserted into the SG200 ip-locus by homologous recombination, and targeted genome insertion can be verified by Southern blot analysis. Secretion of the fusion protein is confirmed by confocal laser scanning microscopic imaging in seedling leaves infected with a Trojan horse strain (Figure 3). As an example, seedlings infected with either the U. maydis Trojan horse strain SG200Zmmac1 secreting a ZmMAC1-mCHERRY or a non-Trojan horse control strain expressing noSP-Zmmac1-mCHERRY that lacks the Umpit2-SP, and subsequently does not secret the ZmMAC1-mCHERRY-HA protein, are shown in Figure 3¹⁵. Hyphae secreting ZmMAC1 are surrounded by fluorescent mCherry signal (Figure 3A). In contrast, non-secreting hyphae only show fluorescence signal in the fungal cytoplasm (Figure 3B).

In particular, tassel infection with U. maydis relies on proper tissue inoculation, as described in step 3.5. Improper tassel localization or unequal distribution of the inoculum can result in non-infected tassel (Figure 4A) or only partial infection of the tassel (Figure 4B). To ensure even distribution, the inoculum needs to be slowly released from the inoculation needle to acquire entire tissue infection (Figure 4C). Viability of the inoculum can be verified by placing a droplet on PD-Charcoal agar. Infectious strains form filaments appearing as write fluff on the plate as shown for the solopathogenic Trojan Horse progenitor strain SG200 (Figure 5A) while the U. maydis strain FB1 requires mating before infectious filament formation (Figure 5B).

Table 1: Sequences of primer additions to add restriction sides and the RSIATA motif coding sequence to the maize gene of interest.

Figure 1: Schematic overview of the U. maydis Trojan horse plasmid cloning strategy. Zmmac1 (light grey) is released by double digest from p123-P_Umpit2-Sp_Umpit2-Zmmac1-mCherry-Ha. In parallel, the gene of interest (yellow) is amplified by PCR. For cloning purpose, forward and reverse primers are designed which include XbaI and NcoI cloning sites and a RSIATA linker (purple). The PCR product is digested with XbaI and NcoI. After ligation, the Trojan horse plasmid contains the following elements: Driven by the Umpit2 promoter, an Umpit2-SP (blue) is fused N-terminally to a maize gene of interest ORF (yellow). At the C-terminus, a RSIATA linker, a mCHERRY reporter gene (red) and an HA epitope tag (green) are fused followed by a stop codon. Prior to U. maydis transformation, the Trojan horse plasmid is digested with SspI to allow homologous integration into the U. maydisip-locus (grey).  Please click here to view a larger version of this figure.

Figure 2: Light-microscopic examination of U. maydis inoculation culture. Several cigar-shaped U. maydis sporidia are visible, some of which undergo budding (indicated by asterisks). No further cells are present which would indicate a contamination of the culture. Scale bar = 20 µm. Please click here to view a larger version of this figure.

Figure 3: In planta confocal laser scanning microscopic imaging of an Ustilago Trojan horse strain. Imaging of the mCherry-fused maize protein ZmMAC1 after maize seedling infection using the Trojan horse strain SG200Zmmac1 (A) or a non-Trojan horse strain SG200Zmmac1-noSP lacking the Umpit2-SP (B). Secreted ZmMAC1-mCHERRY-HA fusion protein is located on the surface of U. maydis hyphae (A)¹⁵, indicated by the arrow heads. In SG200Zmmac1-noSP only cytoplasmic localization of ZmMAC1 is visible (B), indicated by the asterisk ¹⁵. Scale bars = 5 µm. Please click here to view a larger version of this figure.

Figure 4: Tassel infection with U. maydis. Unsuccessful infection of tassel with U. maydis (A), partial tassel infection (B) and complete tassel infection (C) 12 days after infection are shown. Please click here to view a larger version of this figure.

Figure 5: Inoculum viability assay on PD-Charcoal agar. The solopathogenic strain SG200¹ was used for Trojan horse generation. SG200 is self-stimulating and forms infectious filaments on a PD-Charcoal agar plate (A). The haploid U. maydis strain FB1 requires mating with a compatible strain prior to filamentous growth (B)²². Scale bars = 2 mm.  Please click here to view a larger version of this figure.

Modern crop research demands protocols for molecular analysis on genetic and protein levels. Genetic accessibility via transformation is not available or inefficient and time-consuming for most crop species such as maize. Moreover, reliable genetic tools such as promoter reporter systems are scarce, which makes it difficult to study in situ protein function with high spatiotemporal resolution at distinct tissue sites. Apoplastic proteins can be studied by infiltration of heterologously expressed and purified proteins into tissues. However, despite advances in heterologous protein expression, targeted infiltration into crop tissues remains difficult and often inefficient for protein functional analysis. The Trojan horse strategy is an alternative approach that does not require transformation or protein infiltration. By employing the secretion apparatus of the maize pathogen U. maydis, delivery of theoretically any protein of interest into the plant apoplast of infected tissue can be achieved. Regarding the size of a protein of interest, the limits of this technique have yet to be explored. In former assays, the 290 amino acids U. maydis effector Cmu1 fused to mCherry was successfully secreted⁷. However, the applicability of the Trojan horse method to bigger proteins remains to be tested. If additions of posttranslational modifications to the protein of interest are undesirable, unconventional secretion may be used as an alternative route to classical secretion via SP¹⁴.

U. maydis is employed as a standard tool in protein biotechnology because of reliable protein folding, posttranslational modification, and secretion efficiencies^11,12,13. Nevertheless, protein secretion for each new Trojan horse strain needs to be carefully analyzed as described in the Step 3. It is recommended to perform an initial testing of every newly generated Trojan horse strain by infecting seedlings that are easy to infect and to inspect by microscopic imaging.

SG200 is a solopathogenic strain that does not require mating prior to the Trojan horse experiments and is thus easier to handle. Although the infection efficiency of compatible strains like FB1 and FB2 is higher²³, the efficiency of SG200 is sufficient for Trojan horse experiments. Some U. maydis effector proteins are taken up by host cells, while others remain in the apoplast. The differentiation between both groups seems to be a controlled and specific process; however, the underlying mechanisms remain elusive⁸ and cannot be taken into account when designing an experiment. Therefore, proteins of interest that have to be integrated into the cell wall or that have to act intracellularly are no suitable candidates for the Trojan horse approach.

Since U. maydis is omnipotent in infecting diverse aerial maize tissues, the Trojan horse method can be applied for multiple proteins, in distinct tissues and at different plant developmental stages, such as seedling leaves, adult leaves, tassels, and ears. To name just a few useful applications, the Trojan horse allows testing local overdosage, offside protein effects, or functional characterization of distinct protein domains.

Maintaining an uncontaminated U. maydis culture is crucial for all described experiments since co-infection with a large amount of bacteria triggers the plant's immune response and alters its reaction towards the protein of interest, thus rendering any results inconclusive. Trojan horse studies in adult leaves, tassels and ears have to be performed with maximally 1.5 mL infection culture as higher volumes may result in tissue damage. Tassel and ear infections can be trained using food color-stained water instead of Ustilago inoculum.