crw1 - A Novel Maize Mutant Highly Susceptible to Foliar Damage by the Western Corn Rootworm Beetle

Western corn rootworm (WCR), Diabrotica virgifera virgifera LeConte (Coleoptera: Chrysomelidae), is the most destructive insect pest of corn (Zea mays L.) in the United States. The adult WCR beetles derive their nourishment from multiple sources including corn pollen and silks as well as the pollen of alternate hosts. Conversely, the corn foliage is largely neglected as a food source by WCR beetles, leading to a perception of a passive interaction between the two. We report here a novel recessive mutation of corn that was identified and named after its foliar susceptibility to corn rootworm beetles (crw1). The crw1 mutant under field conditions was exceptionally susceptible to foliar damage by WCR beetles in an age-specific manner. It exhibits pleiotropic defects on cell wall biochemistry, morphology of leaf epidermal cells and lower structural integrity via differential accumulation of cell wall bound phenolic acids. These findings indicate that crw1 is perturbed in a pathway that was not previously ascribed to WCR susceptibility, as well as implying the presence of an active mechanism(s) deterring WCR beetles from devouring corn foliage. The discovery and characterization of this mutant provides a unique opportunity for genetic analysis of interactions between maize and adult WCR beetles and identify new strategies to control the spread and invasion of this destructive pest.


Introduction
Corn rootworms are among the most serious insect pests of corn in the United States (U.S.) and have recently been introduced to the European Union (E.U.) as well [1]. The corn rootworm or Diabrotica complex consists of four economically important species: the western corn rootworm Diabrotica virgifera virgifera LeConte (WCR), northern corn rootworm Diabrotica barberi Smith (NCR), Mexican corn rootworm Diabrotica virgifera zeae Krysan and Smith (MCR), and southern corn rootworm Diabrotica undecimpunctata howardi Barber (SCR). Very few insect pests have presented more challenges to the U.S. agriculture than the corn rootworm complex [2]. Historically, the extent of insecticidal use for Diabrotica corn rootworms far exceeds that for any other U.S. agricultural pest [3]. The most devastating of the corn rootworm complex is WCR, which overwinters as eggs that hatch into larvae the following spring [1]. After three larval instars, the WCR larvae pupate briefly prior to the emergence of adult beetles in early summer [4]. The larval stage which feeds on roots of corn and a few other grass species is significantly more damaging than the adult beetle [5][6][7]. Extensive larval feeding can interfere with the ability of plants to absorb water and nutrients causing direct losses in grain fill in addition to plant lodging, complicating harvest activities [8]. Damage caused by larval feeding might also predispose roots to infection by root and stalk rot fungi [9]. Overall, WCR damage results in yield losses and control costs that long exceeded $1 billion per year [10], earning it the nickname of 'billion dollar bug' (www.GMO-Safety.eu.). Crop rotation and use of insecticides (soil and foliar) were the primary management strategies in North America throughout the last half of the 20 th century [11,12]. However, the emergence of WCR variants with either extended diapause, a lack of fidelity to corn for oviposition, the evolution of counter-resistance to insecticides, and larger plantings of continuous corn have each contributed to significant range expansion of this insect pest [1][2]13]. The reduced effectiveness of crop rotation has also significantly increased growers' reliance on soil insecticides for first-year corn [14]. The commercial availability of rootworm-protected transgenic corn hybrids since 2003 has provided an effective alternative [15]. However, none of the transgenic events currently registered for WCR control expose larvae to toxin levels considered to be a high dose, leading to concerns about resistance development [16]. These concerns have been further elevated by recent laboratory and field demonstrations of rapid response to selection in the absence of mating with unexposed beetles [17,18]. An alternate, or complementary, WCR management strategy could involve the use of native plant resistance. Despite extensive screening efforts, corn germplasm with robust resistance to WCR has not yet been identified [1,19].
While most WCR management strategies focus upon reducing larval feeding upon roots, a sustainable alternative could involve managing the population level and behavior of adult WCR, which is a beetle. The WCR beetles are polyphagous i.e. derive nourishment from multiple sources including corn pollen and silks along with pollen of alternate hosts, but very rarely from corn leaves [20,21]. The WCR beetles display very low preference for late vegetative or reproductive stage corn foliage [22,23], suggesting a passive interaction between WCR beetles and corn leaves [20]. Although corn germplasm with robust resistance to WCR larvae is lacking [1], observed variation in susceptibility to foliar feeding by the WCR beetle [24], indicates that corn foliage may have effective mechanisms for resistance.
In the present study we report the identification, phenotypic characterization, and genetic mapping of corn root worm1 (crw1), a corn locus defined by its developmentally-dependent foliar susceptibility to WCR beetles. We demonstrate that the adult leaves of crw1 mutants retain the cell wall biochemistry properties associated with juvenile epidermal cells, including toluidine blue-O (TBO) staining pattern, epidermal lobe formation and reduced accumulation of both hydroxycinnamic acids and lignin. Additionally, the fracture dynamics studies are indicative of reduced structural integrity of the adult mutant leaves. Furthermore, molecular mapping studies using simple sequence repeat (SSR) markers delineated the position of Crw1 to chromosome 6. To our knowledge similar mutant phenotypes have not been previously reported in corn or any other grass species. The crw1 mutant unveils a previously undescribed pathway(s) or mechanism(s) for native resistance to foliar feeding by this most devastating insect pest of maize.

Plant Materials and Growth Conditions
The maize (Zea mays L.) crw1 mutant (crw1) was identified by Guri Johal in three separate F 2 populations which shared a common pedigree. These were originally generated by Tom Brutnell's laboratory to scatter Ac elements across the maize genome, and planted at University of Illinois in the summer of 2005. All field experiments for genetic analysis and phenotypic characterization of the crw1 mutant were performed at Purdue Agronomy Center for Research and Education (ACRE) farm in West Lafayette, Indiana. Eighteen seeds per row were planted in 6 m rows space 0.76 m apart. For stagger planting experiments the wild type and crw1 plants were planted in alternating rows at bi-weekly intervals starting from the end of May. This allowed us to have the wild type and crw1 plants at various growth stages coinciding with the emergence of adult western corn rootworm (WCR) beetles (Diabrotica virgifera virgifera LeConte). All experiments were conducted using natural infestations since the ACRE farm has traditionally exhibited heavy WCR pressure.
The crw1-4 mutant was identified by Steve Moose from M 2 families derived from ethyl methanesulfonate mutagenesis of an inbred line that was homozygous for the ramosa1-RS mutation [25], on the basis of purple staining with TBO of 5-mm diameter circular punches obtained from adult leaves. The TBO staining was performed following a published protocol [26]. The crw1-4 mutant allele was introgressed into the B73 and Mo17 inbred lines by six backcrosses followed by two generations of self-pollination to recover homozygous near isogenic lines. The B73 and Mo17 crw1-4 near-isogenic lines were crossed to each other to generate hybrid seeds. The B73 6 Mo17: crw1-4 hybrid, along with a similarly-constructed near-isogenic hybrid for the glossy15-H mutation and the wild-type B73 6 Mo17 hybrid, were each planted in adjacent field plots (5.3 m long and 0.76 m row spacing, 30 plants per row) at the Crop Sciences Research and Education Center at the University of Illinois during the 2008 growing season. The extent of foliar feeding damage was quantified using a rating system of 1 for no foliar feeding to 10 for full defoliation of the most distal 25 cm of the leaf subtending the ear node. Root damage ratings were collected from excavated roots that were soaked, washed and rated for damage using the 0-3 scale [27]. Grain yields were estimated by harvesting all ears from plots, weighing the shelled grain, and adjusting for moisture content measured by near-infrared spectroscopy on a Dickey-John Instalab 600 instrument.
Double mutant analyses of crw1 were conducted by crossing the B73: crw1-4 near-isogenic line to either a B73: gl15-H nearisogenic line, or the brown midrib 1-4 mutant stocks (119F, 408E, 515D, 918B) available from the Maize Genetics Cooperation Stock Center (www.maizegdb.org). F 1 plants were then selfpollinated and at least 100 F 2 progeny planted in the field were phenotyped for TBO staining in adult leaves, leaf waxes and macrohairs, midrib color, and foliar damage from WCR beetles.

Genetic Mapping of crw1
Following three backcrosses of the crw1-4 mutation to the maize inbred line B73 and self-pollination, a population of 441 plants was scored for the crw1 mutant phenotype by TBO staining of leaf 10. The population produced 323 wild-type plants that stained aqua with TBO and 118 crw1-4 mutant plants that stained purple. Genomic DNA was isolated from three separate pools of 12 wildtype and 12 crw1-4 mutant plants and each pool was screened by PCR amplification with a set of 383 simple sequence repeat (SSR) markers selected to be evenly distributed across the maize genetic map (available from Sigma-Aldrich and www.maizegdb.org). Eight markers located to chromosome six showed amplification patterns indicative of segregation distortion and hence potential linkage with crw1. Five of these markers (bnlg161, gpc2, umc1753, bnlg1867, umc1229) were then used to genotype 112 crw1-4 mutant individuals. Genotyping data was used to estimate recombination frequencies between crw1 and each of these markers.

Toluidine Blue-O (TBO) Staining of Maize Epidermal Peels
Leaf discs from juvenile and adult leaf discs (defined here as leaves number 1-6 and 13 and above respectively, counting the first leaf to be initiated as leaf number 1) from nine plants each of crw1 and wild-type were collected. These samples were fixed in 4% formaldehyde in PBS and 0.2% saponin (Sigma-Aldrich, St. Louis) pH 7, washed and incubated with 0.1% pectinolyase overnight [28]. The epidermis was then peeled from the rest of the leaf using a fine point forceps and incubated in 0.05% TBO pH 4.0 until evenly stained. The stained samples were washed and imaged using bright field optics (Olympus Vanox, Olympus Corporation, New Hyde Park, New York).

Cryo-scanning Electron Microscopy
Two leaf strips, K cm wide by 1 cm long, were cut 5 cm from the apical and basal portions of juvenile (leaf number 3), transition (leaf number 7) and adult (leaf number 13) leaves from 9 plants each of crw1 and wild-type. Sample pieces were mounted on a flat holder covered with double-coated carbon tape and cryo-adhesive. They were further secured using narrow carbon tape strips on each side of samples. The sample holder was plunged into liquid nitrogen slush. A vacuum was pulled and the samples were transferred to the Gatan Alto 2500 pre-chamber (cooled to ,170uC). Samples were sublimated at 290uC for 7 min followed by sputter coating for 90 sec. with platinum. The sample holder was then transferred to the microscope cryo-stage (,2130uC) for imaging. Samples were imaged using an FEI NOVA nano SEM field emission scanning electron microscope (FESEM, FEI Company, Hillsboro, Oregon) operated under high vacuum conditions using an Everhart-Thornley secondary electron detector and Through-the-Lens (TLD) high resolution detector. Parameters were 5 kV, Spot 3, 5 mm working distance, and magnifications of 150, 500, and 20006.

Uni-axial Tensile Stage Test
Uni-axial tensile test was performed using a GATAN (Warrendale, PA, USA) low force tensile stage with a 20 Newton load cell. Adult leaf (leaf number 13) samples from 9 plants each of crw1 and wild-type were prepared by cutting 7 mm wide by 4 cm long leaf strips from a constant position along the leaf lamina. A 1 mm incision was made exactly at the midpoint of each leaf strip tested, using a razor blade mounted on a pre-fabricated metallic aid under a stereomicroscope. The samples were affixed between the clamps across the crossheads of the tensile stage. The tensile tests were conducted at a constant extension rate of 1 mm/min. The stage was set to a minimal extension and force at zero before each run. Two replicates from the same position constituted one sample and two samples, one each from the apex and base were taken from every leaf tested. The cryo system attached to the tensile stage enabled us to monitor the incipient crack propagation and SEM imaging of the fracture surface at the point of final failure.

Analysis of Cell Wall Bound Hydroxycinnamic Acids
For cell wall preparation, whole plant leaf tissue from 3 biological replicates (3 plants/replicate) each of V3 and V8 stage crw1 and wild-type were harvested and flash frozen in liquid nitrogen after removing the midribs. The leaf samples were grounded to a fine powder in liquid nitrogen and extracted with neutral phosphate buffer, 70% ethanol and acetone [29,30]. Cell wall esterified phenolics were released by saponification with 1 M sodium hydroxide (NaOH) for 16 h at room temperature and the hydrolysis products were extracted in ethyl acetate and separated by reverse phase HPLC using detection at 280 and 320 nm.

Lignin Analysis
Whole plant leaf tissue from 9 biological replicates (3 plants/ replicate) each of V8 stage crw1 and wild-type adult plants were harvested and flash frozen in liquid nitrogen after removing the midribs. Cell walls were prepared from these leaf samples as mentioned in the previous section. Two milligrams of the dried cell wall material was used for acetyl bromide-soluble lignin (ABSL) analysis following the published protocol [31], with slight modifications. Briefly, the cell wall material was dissolved in 100 ml freshly made acetyl bromide/acetic acid solution (20:80, v/ v) in a 2 ml volumetric flask. The solution was incubated at 50uC for 2 hr with vortexing every 15 min. The mixture was brought to room temperature by cooling on ice and lignin was extracted using a mixture of 400 ml 2 M NaOH and 70 ml 0.5 M hydroxylamine hydrochloride. The volume of the mixture was made up to 2 ml with glacial acetic acid and mixed thoroughly. The percentage of ABSL (% of ABSL) was determined by pipetting 200 ml of the solution into a UV-specific 96 well plate and read in an ELISA reader at 280 nm. Three readings were performed for every sample extracted and the proportion of ABSL lignin was determined by using 18.21 as an extinction coefficient.

Discovery and Genetics of crw1
Approximately 2000 F 2 families derived from Ac transposonmediated mutagenesis of the maize inbred W22 [32] were planted at the University of Illinois in the summer of 2005 and visually examined for a diverse set of phenotypes at the post-flowering stage. Three independent F 2 populations were identified that segregated for plants which displayed signs of insect herbivory. Only a subset of the individual plants within each of the affected families showed insect damage, and these were adjacent to plants with no foliar feeding, suggesting a segregating genetic basis for the phenotype. The absence of the insect causing the damage at this late stage of crop development precluded us from identifying the species responsible.
Additional seeds of the affected F 2 populations were acquired and replanted at Purdue University's Agronomy Center for Research and Education (ACRE) farm in West Lafayette, Indiana in 2006. Plants were checked regularly for germination, growth, development, and symptoms associated with biotic or abiotic stress. None of the plants in any of these F 2 families exhibited any kind of abnormality during the first few weeks of growth. However, when plants reached V5-V6 stage, a few plants from each of the three populations came under attack by the WCR beetles. The number of beetles increased over time on these selected plants which led to their substantial defoliation. The rest of the plants in each F 2 family remained largely untouched by the beetles. These results clearly indicated that genetic mutations led to exceptional foliar herbivory by the WCR beetles. To address whether these mutants were allelic, they were crossed with at least three wild-type plants within each F 2 family and the resulting progenies were tested for susceptibility to WCR beetles. Roughly half of the plants in 15 of the 24 progenies thus produced displayed foliar susceptibility to WCR beetles. The appearance of WCR susceptible mutants in all possible mutations in a single locus. This locus has been designated crw1, for susceptibility to corn root-worm1. These results also suggested that crw1 was probably a recessive mutation. To test this hypothesis, one of the crw1 alleles (crw1-1) was crossed to B73, a maize inbred often used for genetic and genomics studies. As expected, the resulting F 1 hybrid was resistant to WCR, but the crw1 mutant phenotype reappeared in the F 2 population at a frequency consistent with inheritance as single locus recessive allele (17 mutants among 74 individuals, X 2 , p.0.05, 1 d.f.).

Developmental Manifestation of crw1
An interesting feature of the crw1 phenotype that we noticed was that significant WCR beetle damage did not start until the mutant reached the age of about 5-6 weeks. This could have been coincidental, given that our genetic nurseries are about 5-6 weeks old in late June or early July when the WCR beetles emerge. To examine whether the delayed manifestation of crw1 susceptibility was merely due to the absence of WCR beetles during earlier stages of plant growth, crw1 mutants (crw1) and their isogenic wildtypes were planted at weekly intervals from the middle of May to the end of June. This planting scheme allowed the availability of crw1 of all ages at the time of beetle emergence. It was found that regardless of the availability of mutant juvenile seedlings, the WCR beetles always preferred the foliage of crw1 plants starting from V5-V6 stage (Fig. 1A, left panel). Interestingly, once the susceptibility to the WCR beetle became established in crw1, it lasted throughout the life of the mutant plants (Fig. 1B, left panel). In contrast, the wild-type isogenics lacking the crw1 mutation remained resistant to WCR beetle damage throughout their lives ( Fig. 1A and B, right panels).
Unusual susceptibility of crw1 to WCR beetles enabled us to examine the feeding behavior of the WCR beetle. The beetle prefers to eat crw1 leaves from below, the abaxial side (Fig. 1C). The damage often initiates near the tip of leaves first and proceeds largely downwards in a basipetal fashion (Fig. 1C and D). The WCR beetles scrape or chisel away all the leaf tissue from the upper epidermis, which is left intact. This results in transparent areas on the damaged leaves that have a ''window pane'' appearance (Fig. 1E). Being weak and flimsy post damage, the upper epidermis however withers away over time, giving the damaged plants a defoliated appearance with bare midribs (Fig. 1B, left panel). The entire foliage of crw1 can be eaten away if the population level of the beetle is very high. However, if the beetle pressure is low, the damage to crw1 is accordingly also low. In the absence of WCR beetles, or when plants are grown in a greenhouse, the mutant plants are almost indistinguishable from their wild type siblings in growth and vigor. Thus, the phenotype of crw1 is conditional, being contingent on the presence of WCR beetles.

Crw1 is also Required for Biochemical and Morphological Features of Adult Leaf Epidermal Cell Identity
The unique phenotype of late-onset susceptibility to foliar feeding by WCR beetles was also observed to be associated with another maize mutant named epidermal cell wall1 (ecw1). The ecw1 phenotype was identified in a screen of an ethyl methanesulfonate (EMS) mutagenized population for mutants that affected the Figure 1. Phenotype of field grown crw1. A, Initiation of WCR beetle feeding of crw1 around week 5-6 after planting (left, area between arrowheads). B, Once initiated, the WCR beetle feeding results in complete loss of foliage during the adult stage (left, area between arrowheads). The wild type (W22 inbred) on the other hand does not show WCR beetle damage at the same stages (A and B right). C and D, stripping of epidermal tissue on underside of leaves in a basipetal fashion in crw1 by WCR beetles. E, characteristic ''window pane'' pattern resulting due to WCR feeding of crw1 leaves. doi:10.1371/journal.pone.0071296.g001 changes in cell wall biochemistry which distinguish juvenile and adult leaf epidermal cell identity, as visualized by differential staining reaction with TBO (Fig. 2). The first 5-6 juvenile leaves of wild-type plants stain purple and possess a weakly-invaginated epidermal lobes. In contrast, subsequent adult leaves stain aqua with TBO and exhibit extensive lobing that interlocks the adjacent intercostal cells. The juvenile leaves of the ecw1 mutant stain purple and are weakly-invaginated as in wild-type; however, these features also continue to persist in all adult leaves of the ecw1 mutant.
Considering the fact that maize mutants exceptionally susceptible to the WCR beetles were never reported before, we considered it likely that ecw1 may be another mutant allele of crw1. To test this hypothesis, we checked the TBO staining and epidermal lobe pattern of crw1. We found that the TBO staining and epidermal lobing pattern of crw1 was identical to ecw1 ( Fig. 2  and 3). We then crossed crw1 and ecw1 (introgressed into the B73 inbred background) to test for genetic complementation. Although the resulting F 1 appeared as robust as the normal B73/W22 hybrid, its adult leaves stained purple with TBO as observed for ecw1 and crw1. In addition, it was highly susceptible to the WCR beetles, clearly demonstrating that crw1 and ecw1 are allelic and result from mutations in the same gene. As a result, ecw1 is now designated as crw1-4.
Importantly, the crw1 mutants were not impaired in the replacement of juvenile waxes or production of macrohairs ( Fig. 4), two other leaf identity traits associated with vegetative phase change [26]. To further assess its relative position within the vegetative phase change pathway, the crw1-4 mutation was crossed to the glossy15 (gl15) mutation that coordinately regulates the entire suite of juvenile versus adult leaf epidermal traits [26]. The F 1 plants produced fully normal juvenile and adult leaves as in wildtype plants, indicating crw1 was not allelic to gl15. Self-pollination of the F 1 plants produced an F 2 population of 182 plants with four phenotypic classes in proportions consistent with two independently segregating recessive alleles (9:3:3:1, X 2 , p.0.05, 3 d.f.): 112 fully wild type, 30 typical glossy15 phenotype where all adult epidermal cell traits appear earlier at leaf 3, 32 crw1 type with prolonged purple TBO staining but otherwise normal vegetative phase change, and 8 crw1; gl15 double mutants that exhibited accelerated onset of glossy leaf wax phenotype and macrohairs at leaf 3, but purple TBO staining and weakly-invaginated walls in all leaves. These observations suggested that crw1 is required for the biochemical features that define adult leaf epidermal cells, and indicated that the cell shape differences between juvenile and adult leaf epidermal cells are linked to these same biochemical changes.
Near-isogenic lines were created for the crw1-4 by backcrossing to both the B73 and Mo17 inbreds. The susceptibility to foliar feeding and TBO staining phenotypes were very similar in both genetic backgrounds. Because maize is grown commercially as hybrids and tolerance to root damage by WCR larvae differs between inbreds and hybrids [33], we produced the B736Mo17 hybrids from the near-isogenic lines for crw1-4 and evaluated them for foliar feeding, root damage ratings, and grain yield (Fig. 5, Table 1). A near-isogenic hybrid for the gl15 mutation was also included for comparison. The leaves of the crw1-4 hybrid showed much greater damage from foliar WCR beetle feeding compared to either the normal or gl15 mutant hybrids (Fig. 5). There were no significant differences in root damage rating among the three hybrids and the crw1-4 hybrid also produced approximately 30% less grain (Table 1) compared to the normal or gl15 mutant hybrids. These observations suggest that crw1 primarily affects foliar feeding by WCR adult beetles, and its pleiotropic effects on leaf development can reduce grain yield under moderate levels of WCR infestation.

Reduced Levels of Hydroxycinnamic Acids and Lignin in crw1 Leaf Walls
What could be the reason for the adult leaves of crw1 to show a staining reaction with TBO like those of juvenile leaves? The polychromatic staining of TBO depends on the density of charged functional groups in cell wall components, with purple staining indicative of a higher density of reactive groups [34]. A likely source of differential staining reactions with TBO in the maize leaf epidermis is the altered accumulation of cell wall bound phenolic compounds [35], which are key determinants of TBO staining in other species [34,36], and which are important constituents of adult grass cell walls. Two key phenolics in this regard are pcoumaric acid (pCA) and ferulic acid (FA), which serve to crosslink cell wall polysaccharides and lignins [37]. It had been found previously that leaves with adult identity, including the precociously adult leaves of the glossy15 mutation, exhibit higher concentrations of pCA and FA [35]. To address if the levels of pCA and FA were compromised in crw1, we isolated these hydroxycinnmates from the juvenile (V3 stage) and adult (V8) walls of both crw1 and wild type leaves by saponification of the isolated cell walls and analyzed them by HPLC. Not too surprisingly, minor variation could be detected in the total contents of pCA and FA in the juvenile leaves of crw1 and wild type leaves. However, substantial reductions in the amounts of these cell wall bound hydroxycinnmates especially, pCA was found in the adult leaves of crw1 (33% reduction at p,0.001; unpaired t test) compared to wild -type plants (Fig. 6A).
Given the intimate association between the accumulation of pCA and lignification in grasses [37]; we tested whether there were differences in the levels of lignin in the adult walls of crw1 leaves compared to their wild-type counterparts. To address this likelihood, lignin was extracted from isolated cell walls as acetyl bromide soluble (ABS) fraction and analyzed by UV spectroscopy. Consistent with the reduction in the levels of pCA in crw1 adult leaf walls, the levels of ABS lignin are lower (p,0.001; unpaired t test) in the crw1 leaves at maturity compared to wild-type leaves (Fig. 6B). The brown midrib (bm) mutations of maize also exhibit reductions in lignin and changes in pCA, FA and their esterification [38], yet do not suffer from enhanced susceptibility to foliar feeding by WCR beetles or exhibit altered TBO staining. Similarly, although the crw1 mutant shows reductions in lignin, they do not show the brown midrib phenotype, indicating crw1 affects cell wall phenolics via a distinct genetic pathway. This hypothesis was confirmed by the double mutant analysis of crw1-4 in combination with each of the bm1, bm2, bm3 and bm4 mutations. Within each population of F 2 progeny produced from the cross of crw1 with a brown midrib mutation, the segregation ratios for single and double mutant phenotypes were consistent with two independently-acting recessive mutations. Approximately 1/16 of the individuals produced completely normal juvenile leaves and adult leaves that possessed both reddish-brown midribs and that stained uniformly purple with TBO.

Reduced Tensile Strength of crw1 Leaves
The adult crw1 leaves were found to be compromised in three strength-imparting components of cell walls, namely the epidermal cell invaginations, cellulose tethering hydroxycinnamates, and lignification. Accordingly, crw1 leaves appear to be less rigid than wild-type. We compared tensile strength of crw1 and wild-type leaves with a tensile stage SEM. The results obtained indicated that there are significant differences in both crack propagation and  fracture dynamics between crw1 and wild type leaves. While adult leaves of wild type plants from any position along the leaf lamina form even fracture surfaces, equivalent crw1 leaves form uneven fractures when pulled to the point of final failure ( Fig. 7A and B). Moreover, crw1 leaves tend to form additional fractures along the axis of the incipient crack propagation (Fig. 7C). Because the formation of uneven and multiple fracture lines are often associated with low structural integrity in plant tissues [39], these results indicate that crw1 leaves have a lower tensile strength compared to wild-type leaves.

Crw1 Maps to Chromosome 6
A population of 441 plants segregating for the crw1-4 allele in the B73 inbred background were scored for the crw1 mutant phenotype and used to obtain a genetic map position for the crw1 locus. As expected for a single recessive locus, approximately 25% (118) of these individuals exhibited the crw1 mutant phenotype. Pools of wild type and crw1-4 mutant plants were used in a bulk segregant analysis approach and genotyped with 384 SSR markers selected to uniformly span the maize genetic map. Five markers (bnlg161, gpc2, umc1753, bnlg1867, umc1229) located to chromosome bin 6.00 on the IBM Neighbors2 map were found to exhibit differential amplification among the wild-type and crw1-4 mutant pools, indicative of segregation distortion and genetic linkage. Further genotyping of 112 crw1-4 mutant individuals with these five markers refined the genetic map in this region and located Crw1 in a 17 centiMorgan interval between bnlg161b (8 recombinants) and gpc2 (12 recombinants). In silico mapping of these flanking markers on the B73 reference genome sequence (www. maizesequence.org) indicated this chromosomal region spanned nearly 6 Mbp.

Discussion
We describe here the discovery, mapping and preliminary characterization of crw1, a corn mutant that was identified on the basis of its exceptional foliar susceptibility to WCR beetles. Its recessive loss-of-function nature suggests that a key characteristic which is usually required for deterring WCR beetles is lacking in this mutant. It is also possible that this mutation leads to perturbations in biochemical pathways resulting in the accumulation of pathway intermediates that might enhance the attraction of corn foliage to WCR beetles. One curiosity is that this mutant was never reported before, perhaps because the most dramatic  visual aspect of the crw1 phenotype is conditional upon sufficient WCR beetle feeding pressure.
Although the adult hosts of WCR beetles are varied [19][20][21], extensive feeding (to the point of total defoliation) upon corn foliage is very unusual. The WCR beetles primarily feed on corn pollen and silks but target the relatively nutrient-deficient leaves only as a last resort [19]. WCR beetles are known to survive on pollen of alternate hosts (mostly weeds) during first few weeks of their adult life, until corn pollen and silks becomes available [40]. They show very low preference to late vegetative or reproductive stage corn foliage [22][23]. The continued feeding upon crw1 foliage by WCR beetles starting from early adult vegetative stage (V5-V6 stage) through reproductive maturity ( Fig. 1A and B, left panel) is a stark deviation from the norm. This unusual feeding choice demonstrates that an active deterrence mechanism might be present in corn foliage which restricts the feeding preference of WCR beetles to silks and pollen. We hypothesize that the absence of this active deterrence mechanism in crw1 leads to its exceptional foliar susceptibility.
The onset of foliar susceptibility in crw1 coincided with vegetative phase change, and the purple TBO staining and moderate lobing of crw1 leaf epidermal cells ( Fig. 2 and 3) suggested that crw1 may function in the transition from juvenile to adult leaf identity. Interestingly, there have been previous reports of enhanced susceptibility to insect herbivory when the juvenile phase and purple TBO staining of leaf epidermal cells is prolonged by the maize Corngrass1 mutation [41][42][43][44] that overproduces microRNA156 (45). However, except perhaps for Corngrass1 (Cg1), none of the other known maize mutants that alter vegetative phase change (Teopod1 (Tp1), Teopod2 (Tp2), GA-deficient or insensitive loci, early phase change1 (epc1), or glossy15 (gl15), exhibit enhanced susceptibility to foliar feeding by WCR (Table 1, S. Moose, Unpub.). Conversely, crw1 only affects TBO staining and cell shape but not other traits associated with vegetative phase change (Fig. 4). Double mutant analysis indicated that Crw1 function is required for the accelerated onset of aqua TBO staining in leaves 3-7 of glossy15 mutants. These observations indicate that Crw1 promotes the biochemical and morphological features of intercostal cells within the epidermis of normal adult leaves, and may be repressed by glossy15 in normal juvenile leaves. Another possibility for the late onset of foliar susceptibility in crw1 might be the DIMBOA pathway. The DIMBOA pathway confers early stage resistance to insects in corn however, the potency of this resistance declines with age [45]. However, double mutant analysis of crw1 with bx1, a mutant impaired in DIMBOA pathway, indicated no link between DIMBOA with the foliar susceptibility conditioned by crw1 (BP Venkata, Unpub.). These observations are consistent with the recent findings that WCR larvae are completely immune to the products of the DIMBOA pathway [46].
The altered TBO staining pattern of the adult crw1 intercostal cells might be due to reduced levels of cell wall bound phenolic acid monomers (pCA and FA), which appear to be critical determinants of the staining pattern [34][35][36]. Consistent with this hypothesis, the levels of pCA are substantially lower in adult crw1 leaves compared to the wild type (Fig. 6A). Further, since pCA is esterified to the c position of S lignin subunit in grasses, its accumulation is usually considered to be a marker for lignin deposition [37,47,48]. Not surprisingly, crw1 shows a reduction in foliar lignin content during adult stages (Fig. 6B). In general, higher levels of lignin and cross linking of cell wall components by cell wall bound phenolic compounds enhance both leaf and stalk stiffness in grasses [37]. In addition, there have been previous reports correlating increased phenolic acid and lignin levels with resistance to insect pests in corn [37,49]. Despite this, none of the corn bm mutants (bm1, bm2, bm3 and bm4) with altered lignin content and/composition [38] exhibit unusual foliar susceptibility to WCR beetles. Interestingly, the percentage reduction of lignin in crw1 is comparable to the overall reduction in the lignin levels of bm2 and bm3 [50,51], negating the exclusive role of reduced lignin content in its foliar susceptibility to WCR beetles. Further, none of the bm mutants of maize shared the altered TBO staining phenotype with crw1, nor do crw1 mutations exhibit a brownmidrib phenotype. Therefore, our finding that the crw1 mutation acts in an independent genetic pathway from the bm mutants indicates that the impacts of crw1 on maize cell wall phenolics may be mediated by changes in either the degree of cross-linking among hydroxycinnamic acids or between these phenolics and the glucoronoarabinoxylans that comprise the major carbohydrate polymer in maize leaf cell walls. Consistent with the above, the 17 centiMorgan interval on chromosome 6 to which Crw1 maps is devoid of any known genes or regulators for phenylpropanoid biosynthesis.
In grass species like corn the intercostal cell lobes function to enhance intercellular adhesion through physical interlocking [52]. The moderate epidermal lobing pattern in crw1 in conjunction with reduced levels of cell wall components including lignin might contribute towards its lower leaf toughness during the adult stage. The results from uniaxial tensile stage SEM support this hypothesis. The dynamics of incipient crack extension in the adult crw1 leaves is suggestive of lower resistance around the plane of crack extension that resulted in an uneven fracture surface at the point of critical failure (Fig. 7B). This is in agreement with previous reports that associate this characteristic with lower structural integrity [39]. The development of additional fracture lines along the axis of the incipient crack propagation in the adult crw1 leaves (Fig. 7C) further substantiates the above argument. There is a strong body of evidence that leaf mechanical properties, including strength and toughness, have a detrimental effect on insect herbivory [37,53,54]. The mechanical properties leading to leaf toughness might deter insect herbivory either by mechanical constraints or by dilution of the nutrients [55,56]. Therefore, the lower structural integrity of crw1 leaves during the adult stage might be conducive for preferential feeding by WCR beetles. However, we also note that other maize genotypes that display ultrastructural compromise, such as the brittle stalk (bk2) mutation [57] or populations with low rind penetrometer resistance do not show enhanced susceptibility to foliar damage by WCR beetles. Similarly, although the intercostal cell lobing pattern observed for crw1 is identical to the Extra cell layer (Xcl1) [58] mutant, neither Xcl1 nor the brick1 (brk1) or brick2 (brk2) mutants that completely lack intercostal cell lobes [59] display enhanced foliar susceptibility to WCR beetles.

Conclusion
Taken together, the data presented here indicates that the crw1 mutation causes pleiotropic defects on leaf epidermal cell wall biochemistry and shape, yet defines a genetic pathway that contributes to resistance to foliar feeding by the WCR beetle. It does not appear that any single perturbation of leaf epidermal cell walls is specifically important for the foliar susceptibility to WCR beetle, suggesting that this feature of the crw1 mutant phenotype is a combinatorial effect of reduced lobes, lowered lignification and altered cell wall composition as indicated by changes in TBO staining. In such a scenario, it should be possible to reconstitute the WCR foliar susceptibility in a triple mutant of gl15 (a phase transition mutant), bm2 (lignification mutant) and brk1 (a mutant devoid of epidermal cell lobes). The generation of such triple mutant is currently underway. It is also possible that other mechanisms in addition to cell wall biochemistry may contribute to the enhanced susceptibility of crw1 leaves to WCR feeding. The discovery of crw1 opens up a new avenue to identify native resistance to WCR that might significantly mitigate the economic scourge caused by this insect pest.