**2. Material and methods**

#### **2.1. Plant material**

72 Current Insights in Pollen Allergens

isoforms (Deeks et al. 2005).

(Kovar et al. 2001).

profilin in the pollen tube is expected.

generated among germplasm, as a result of polymorphism. The high variability might result in both differential profilin properties and differences in the regulation of the interaction with natural partners, suggesting that isovariant dynamics may expand the responses of the actin cytoskeleton or buffer it to against stress (Jimenez-Lopez 2008, Jimenez-Lopez et al. 2012).

Profilin is a major regulator of actin dynamics and is crucial for cellular growth, morphogenesis and cytokinesis (Jockusch et al. 2007). In addition to binding to actin, profilins bind to other partners like stretch of poly-proline (PLP) and proline-rich proteins, and phospholipids. The proline-binding ability could be a major function, being different among profilin isoforms, affecting actin-based structures (Kovar et al. 2001). The importance of the binding of profilin to PLP is supported by the finding of the differential preference for profilin isoforms of formin (essential actin-binding and nucleator protein) (Neidt et al. 2009), together with the evidence that *Arabidopsis* formins have preference for different profilin

Another binding ligand of profilin is phospholipids. The binding of profilin to phospholipids links to its potential role in vesicle trafficking (Janssen and Schleicher 2001). Profilins have been revealed as key mediators of the membrane–cytoskeleton communication, acting at critical points of signaling pathways initiated by events in the plasma membrane and transmitted by transduction cascades to promote cytoskeletal rearrangements (Baluska et al. 2002). This functionality arises from their capacity of interaction with phosphatidylinositides (PIP2), as well as with poly-L-proline-rich proteins

Several locations have been attributed to profilin. They have been localized in different plant cells and tissues, including the microspores, pollen grains and pollen tubes (Grote et al. 1993, 1995; Hess et al. 1995; Fischer et al. 1996; Vidali & Hepler 1997; Kandasamy 2002). Plant profilin was reported to be localized in the cytoplasm of pollen tube uniformly (Hess et al. 1995; Vidali & Hepler 1997). However, no clear picture has yet been established about the precise location of profilin in the pollen tubes. In consideration of the existence of calcium gradient in pollen tube and the regulation of profilin's sequestering activity by calcium (Kovar et al. 2000), the existence of a gradient of total sequestering activity of

Upon pollen hydration and pollen germination, profilin was detected close to the site of pollen tube emergence, forming a ring-like structure around the apertural region. Profilin was also detected in the pollen exine of the germinating pollen grains and in the germination medium. Profilin was also localized in the cytoplasm of the pollen tube,

Depending on the fixation and extraction protocol used, nuclear localization has been also observed (Buss et al. 1992). Profilin has also been found in generative and vegetative nuclei of *Ledebouria socialis* pollen (Hess and Valenta 1997), the nuclei of *Phaseolus vulgaris* cells (Vidali et al. 1995) and *Arabidopsis thaliana* and maize root hairs (Braun et al. 1999, Baluska et al. 2001). Microinjection of a fluorescently labeled birch profilin in *Micrasterias denticulata* 

particularly at both the proximal and apical ends (Morales et al. 2008).

also shows an accumulation of profilin in the nucleus (Holzinger et al. 1997).

Olive (*Olea europaea* L.) pollen was individually collected during May and June from olive trees of 24 cultivars, grown in the olive germplasm collection of the Estación Experimental del Zaidín, CSIC, Granada, Spain. Pollen samples were collected in large paper bags by vigorously shaking the inflorescences sequentially sieved through 150 and 50 μm mesh filters to eliminate debris and maintained at -80°C.

Mature seeds from Acebuche (wild olive) and Picual cultivars were obtained from the same collection of well-characterized olive trees growing in the "Estación Experimental del Zaidín" (Granada, Spain), 210 days after anthesis (DAA).

#### **2.2. In vitro pollen germination**

Olive mature pollen from the Picual cultivar was *in vitro* germinated. Pre-hydration was performed by incubation in a humid chamber at 30°C for 30 min. The grains (0.02 g/plate) were then transferred to Petri dishes containing 10mL of the germination medium [10% (w/v) sucrose, 0.03% (w/v) Ca(NO3)2, 0.01% (w/v) KNO3, 0.02% (w/v) MgSO4 and 0.03% (w/v) boric acid], as described by M'rani-Alaoui (2000). The Petri dishes were maintained at 25°C in the dark, and the pollen samples were taken at 5 min, 1, 4, 7, and 18h after the onset of the culture, pollen tube growth was monitoring by light microscopy. Finally, the pollen was pelleted by centrifugation (1000 × g for 20 s).

#### **2.3. Protein extraction**

Mature pollen or germinated samples were re-suspended in an extraction buffer (PBS), pH 7.4 [140 mM NaCl, 2.7 mM KCl, 8.15 mM Na2HPO4 and 1.8 mM KH2PO4] added with 10

μg/μL protease inhibitor cocktail (Sigma, Madrid, Spain) to a proportion of 5 mL solution per 0.5 gram of fresh tissue and then incubated at 4◦C for 4 h with vigorous shaking. After centrifugation at 13000 × g for 30 min at 4◦C, the supernatants were removed, dispensed into aliquots and stored at –20°C. The process was repeated two times and proteins from both protein extractions were precipitated together overnight at -20°C with a trichloroacetic acid solution.

Differential Immune-Reactivity and Subcellular Distribution Reveal

the Multifunctional Character of Profilin in Pollen as Major Effect of Sequences Polymorphism 75

*2.6.1. Immunolocalization of profilins in pollen by Transmission Electron Microscopy* 

although incubation with the primary antibody was omitted.

distinctive protein profiles with different intensity.

not clearly distinguishable (Figure 1).

The germinated pollen grains were fixed for 2 h at 4◦C in 0.1% (v/v) glutaraldehyde (GA) and 4% (w/v) parformaldehyde (PF) in buffer 1. The samples were dehydrated in an ethanol series and embedded in Unicryl resin (BBInternational, Cardiff, UK) following a progressive lowering of temperature (PLT) schedule, as described by Alché *et al*. (1999). Ultra-thin sections (80 nm) were cut with an ultramicrotome (Reichert- Jung, Vienna, Austria) and transferred on to formvar-coated 300-mesh nickel grids. Blocking of non-specific binding sites was carried out by incubation of sections for 2 h in a solution containing 5% (w/v) BSA in washing buffer. The blocking was followed by washing in buffer 1 for 10 min and incubation at RT for 2.5 h with the antibodies described above, and their proper dilutions in the blocking solution. After washing several times with buffer 1, the grids were treated for 2 h with a goat anti-rabbit IgG secondary antibody coupled to 15-nm gold particles (BBInternational) diluted 1/100 in the blocking solution. Finally, they were washed in buffer 1 (5 × 5 min), rinsed in double-distilled water (3 × 5 min) and then stained for 15 min with a solution of 5% (w/v) uranyl acetate in the dark. The observations were carried out with a JEM-1011 (JEOL, Tokyo, Japan) TEM. The treatment of control sections was the same,

Statistical analysis was performed by using the SPSS v.18 statistical software package. A general comparison among multiple sample groups was performed throughout one-way analysis of variance (One-way ANOVA) on the basis of the Fisher-Snedecor distribution test (=0.05, significance value) (Mehta and Patel 1983). Normality and variances homogeneity of the data collection were checked by the Shapiro-Wilk test (= 0.05, significance value) (Shapiro and Wilk 1965) and the Levene test (= 0.05 significance value) (Levene 1960), respectively. Post hoc range probes and pair of species comparisons were carried out with the parametric test of Games-Howell (=0.05, significance value) (Games and Howell 1976).

**3.1. Expression and differential reactivity characterization of profilin from olive** 

SDS-PAGE analysis of protein extracts of mature pollen from 24 olive cultivars showed

30μg of total protein was loaded in each line. The main protein bands of know pollen allergens correspond to the mayor olive pollen allergen Ole e 1, with a molecular weigh between 18-20 kDa. Defined protein bands corresponding to profilin around 13-15kDa are

**2.6. Microscopy analysis** 

**2.7. Statistical analysis** 

**3. Results** 

**cultivars** 

0.5 g of seeds material (cotyledons and endosperms) was directly homogenized to a very fine powder in a liquid nitrogen-precooled mortar with a pestle. 0.1g of this powder was resuspended in 5ml of 1M Tris-HCl buffer, pH 7 plus 0.7% sodium dodecyl sulfate (SDS) and 1% 2-mercaptoethanol (denaturing, reducing conditions). After centrifugation at 10000*g*  for 15 min (4°C), the supernatants were filtered through 0.2mm filter and stored at –20◦C.

The protein concentration was measured following the Bradford (1976) method, using the Bio-Rad reagent and bovine serum albumin (BSA) (Bio-Rad, Barcelona, Spain) as standard. In total, 30 μg of protein per lane was loaded into a 12% sodium dodecyl sulfate (SDS) polyacrylamide gel (PAGE), as described by Laemmli (1970). The proteins were separated using a Mini-Protean system (Bio-Rad). After completion of SDS-PAGE, the gels were fixed and Coomassie blue stained [25% methanol, 10%acetic acid and 0.1% Coomassie blue R250]. Digitized images were obtained using the Power Look III scanner and the MagicScan software (UMAX Systems GmbH, Germany).

#### **2.4. Protein transference and immunoblotting**

After completion of proteins separation by SDS-PAGE, they were transferred onto polyvinylidenedifluoride (PVDF) membranes at 25 V for 30 min in a semi-dry transfer cell (Bio-Rad) with transfer buffer containing 25 mM Tris-HCl pH 8.3, 192 mM, and 20% glycine. For immunodetection, blots were incubated for 4 h at 25◦C with a blocking solution containing 0.1% Tween 20 and 10% dried skimmed milk in Tris-buffered saline (TBS). The membranes were then probed with:


An alkaline phosphatase-Conjugated anti-rabbit IgG (Promega Co) diluted 1:10000 served as the secondary antibody, and the detection reaction was developed using NBT-BCIP colorimetric system.

#### **2.5. Quantitations**

The intensity of each band was calculated from scanned images of gels by using the quantitation tools of the Quantity One v4.6.2 software (Bio-Rad Laboratories, USA).

#### **2.6. Microscopy analysis**

74 Current Insights in Pollen Allergens

software (UMAX Systems GmbH, Germany).

membranes were then probed with:

1996), in a dilution 1:500.

colorimetric system.

**2.5. Quantitations** 

**2.4. Protein transference and immunoblotting** 

anti-Ole e 2 (Morales et al. 2008) in a dilution 1:20000.

solution.

μg/μL protease inhibitor cocktail (Sigma, Madrid, Spain) to a proportion of 5 mL solution per 0.5 gram of fresh tissue and then incubated at 4◦C for 4 h with vigorous shaking. After centrifugation at 13000 × g for 30 min at 4◦C, the supernatants were removed, dispensed into aliquots and stored at –20°C. The process was repeated two times and proteins from both protein extractions were precipitated together overnight at -20°C with a trichloroacetic acid

0.5 g of seeds material (cotyledons and endosperms) was directly homogenized to a very fine powder in a liquid nitrogen-precooled mortar with a pestle. 0.1g of this powder was resuspended in 5ml of 1M Tris-HCl buffer, pH 7 plus 0.7% sodium dodecyl sulfate (SDS) and 1% 2-mercaptoethanol (denaturing, reducing conditions). After centrifugation at 10000*g*  for 15 min (4°C), the supernatants were filtered through 0.2mm filter and stored at –20◦C.

The protein concentration was measured following the Bradford (1976) method, using the Bio-Rad reagent and bovine serum albumin (BSA) (Bio-Rad, Barcelona, Spain) as standard. In total, 30 μg of protein per lane was loaded into a 12% sodium dodecyl sulfate (SDS) polyacrylamide gel (PAGE), as described by Laemmli (1970). The proteins were separated using a Mini-Protean system (Bio-Rad). After completion of SDS-PAGE, the gels were fixed and Coomassie blue stained [25% methanol, 10%acetic acid and 0.1% Coomassie blue R250]. Digitized images were obtained using the Power Look III scanner and the MagicScan

After completion of proteins separation by SDS-PAGE, they were transferred onto polyvinylidenedifluoride (PVDF) membranes at 25 V for 30 min in a semi-dry transfer cell (Bio-Rad) with transfer buffer containing 25 mM Tris-HCl pH 8.3, 192 mM, and 20% glycine. For immunodetection, blots were incubated for 4 h at 25◦C with a blocking solution containing 0.1% Tween 20 and 10% dried skimmed milk in Tris-buffered saline (TBS). The

1. Whole olive and maize profilins polyclonal antibodies anti-ZmPRA (dilution 1:500) and

2. Polyclonal antibodies specific to particular isoforms of maize profilin like ZmPRO5 (Kovar et al. 2000), ZmPRO4 (Gibbon *et al.* 1998) and ZmPRO3 (Karakesisoglou *et al.*

An alkaline phosphatase-Conjugated anti-rabbit IgG (Promega Co) diluted 1:10000 served as the secondary antibody, and the detection reaction was developed using NBT-BCIP

The intensity of each band was calculated from scanned images of gels by using the

quantitation tools of the Quantity One v4.6.2 software (Bio-Rad Laboratories, USA).

#### *2.6.1. Immunolocalization of profilins in pollen by Transmission Electron Microscopy*

The germinated pollen grains were fixed for 2 h at 4◦C in 0.1% (v/v) glutaraldehyde (GA) and 4% (w/v) parformaldehyde (PF) in buffer 1. The samples were dehydrated in an ethanol series and embedded in Unicryl resin (BBInternational, Cardiff, UK) following a progressive lowering of temperature (PLT) schedule, as described by Alché *et al*. (1999). Ultra-thin sections (80 nm) were cut with an ultramicrotome (Reichert- Jung, Vienna, Austria) and transferred on to formvar-coated 300-mesh nickel grids. Blocking of non-specific binding sites was carried out by incubation of sections for 2 h in a solution containing 5% (w/v) BSA in washing buffer. The blocking was followed by washing in buffer 1 for 10 min and incubation at RT for 2.5 h with the antibodies described above, and their proper dilutions in the blocking solution. After washing several times with buffer 1, the grids were treated for 2 h with a goat anti-rabbit IgG secondary antibody coupled to 15-nm gold particles (BBInternational) diluted 1/100 in the blocking solution. Finally, they were washed in buffer 1 (5 × 5 min), rinsed in double-distilled water (3 × 5 min) and then stained for 15 min with a solution of 5% (w/v) uranyl acetate in the dark. The observations were carried out with a JEM-1011 (JEOL, Tokyo, Japan) TEM. The treatment of control sections was the same, although incubation with the primary antibody was omitted.

#### **2.7. Statistical analysis**

Statistical analysis was performed by using the SPSS v.18 statistical software package. A general comparison among multiple sample groups was performed throughout one-way analysis of variance (One-way ANOVA) on the basis of the Fisher-Snedecor distribution test (=0.05, significance value) (Mehta and Patel 1983). Normality and variances homogeneity of the data collection were checked by the Shapiro-Wilk test (= 0.05, significance value) (Shapiro and Wilk 1965) and the Levene test (= 0.05 significance value) (Levene 1960), respectively. Post hoc range probes and pair of species comparisons were carried out with the parametric test of Games-Howell (=0.05, significance value) (Games and Howell 1976).

#### **3. Results**

#### **3.1. Expression and differential reactivity characterization of profilin from olive cultivars**

SDS-PAGE analysis of protein extracts of mature pollen from 24 olive cultivars showed distinctive protein profiles with different intensity.

30μg of total protein was loaded in each line. The main protein bands of know pollen allergens correspond to the mayor olive pollen allergen Ole e 1, with a molecular weigh between 18-20 kDa. Defined protein bands corresponding to profilin around 13-15kDa are not clearly distinguishable (Figure 1).

Differential Immune-Reactivity and Subcellular Distribution Reveal

the Multifunctional Character of Profilin in Pollen as Major Effect of Sequences Polymorphism 77

**Figure 2.** Immuno-reactivity analysis of profilin from 24 olive pollen cultivars. Reactivity of protein crude extracts from 24 olive cultivars was assayed against different maize profilin antisera, A) anti-PRA, B) anti-ZmPRO3, C) anti-ZmPRO5, D) anti-ZmPRO4, E) as well as against olive profilin antiserum anti-Ole e 2. Up to 3 reactive bands about 13, 13.7 and 14.2 kDa, corresponding to different isoforms of profilins were observed. The intensity of the reactive bands was quantitated by a densitometric analysis:

highlighted the differential reactivity (very high or very low) of defined cultivars to particular antisera, whereas blue asterisks highlighted the differential reactivity among cultivar to defined antisera.

yellow plot corresponding to 13kDa bands, pink (13.7 kDa) and blue (14.2 kDa). Red asterisks

**Figure 1.** Protein profile of crude protein extract from 24 olive cultivars.

Cross-reactivity analysis between protein extracts from 24 olive cultivars with different antisera made against profilins from olive and maize pollen showed large differences both qualitative (intensity of bands) and quantitative (number of reactive bands) concerning profilins of MW around 13, 13.7 y 14.2kDa (Figure 2).

Statistical analysis of densitometric quantitations was performed. The variance analysis for the different antisera against different protein extracts from 24 olive cultivars (Figure 2 A to E) showed significant differences (F-ratio=14.06, p<0.05). The reactivity values were inside a normal distribution (Shapiro-Wilk=0.85, p>0.05), while the Levene test indicated non homogeneity of variances (Levene=10.16, p<0.05).

Multiple comparisons of the five antisera by Games-Howell test determined the existence of statistically significant differences (p<0.05) between pairs of serum analyzed (anti-PRA to anti-ZmPRO4, anti-PRA to anti-ZmPRO3, anti-ZmPRO4 to anti-ZmPRO5, and anti-ZmPRO3 to anti-ZmPRO5, with Games-Howell test results of 73705.4, 70384.6, 124308.1 and 53923.4, respectively.

The analysis of the reactivity of the different cultivars against each individual antiserum showed statistical significant differences between Leccino and Picual, Lechín de Sevilla and Picudo, Lucio and Frantoio, Blanqueta and Farga, as well as between Picudo and Cornicabra corresponding to the immunoblots individually probed with anti-PRA, anti-ZmPRO3, anti-ZmPRO5, anti-ZmPRO4 y anti-Ole e 2, respectively (Figure 2A-E). Oppositely, it is possible to observe clear differences between antisera in their reactivity to defined cultivars, such as Hojiblanca, Arbequina, Cornicabra, Bella de España and Empeltre for the immunoblotting corresponding to anti-ZmPRO3; Lechin de Sevilla, Verdial de Huevar, Loaime, Lucio and Leccino for anti-ZmPRO4; and Sourani and Villalonga for anti-Ole e 2 (Figure 2A-E).

**Figure 1.** Protein profile of crude protein extract from 24 olive cultivars.

profilins of MW around 13, 13.7 y 14.2kDa (Figure 2).

homogeneity of variances (Levene=10.16, p<0.05).

53923.4, respectively.

(Figure 2A-E).

Cross-reactivity analysis between protein extracts from 24 olive cultivars with different antisera made against profilins from olive and maize pollen showed large differences both qualitative (intensity of bands) and quantitative (number of reactive bands) concerning

Statistical analysis of densitometric quantitations was performed. The variance analysis for the different antisera against different protein extracts from 24 olive cultivars (Figure 2 A to E) showed significant differences (F-ratio=14.06, p<0.05). The reactivity values were inside a normal distribution (Shapiro-Wilk=0.85, p>0.05), while the Levene test indicated non

Multiple comparisons of the five antisera by Games-Howell test determined the existence of statistically significant differences (p<0.05) between pairs of serum analyzed (anti-PRA to anti-ZmPRO4, anti-PRA to anti-ZmPRO3, anti-ZmPRO4 to anti-ZmPRO5, and anti-ZmPRO3 to anti-ZmPRO5, with Games-Howell test results of 73705.4, 70384.6, 124308.1 and

The analysis of the reactivity of the different cultivars against each individual antiserum showed statistical significant differences between Leccino and Picual, Lechín de Sevilla and Picudo, Lucio and Frantoio, Blanqueta and Farga, as well as between Picudo and Cornicabra corresponding to the immunoblots individually probed with anti-PRA, anti-ZmPRO3, anti-ZmPRO5, anti-ZmPRO4 y anti-Ole e 2, respectively (Figure 2A-E). Oppositely, it is possible to observe clear differences between antisera in their reactivity to defined cultivars, such as Hojiblanca, Arbequina, Cornicabra, Bella de España and Empeltre for the immunoblotting corresponding to anti-ZmPRO3; Lechin de Sevilla, Verdial de Huevar, Loaime, Lucio and Leccino for anti-ZmPRO4; and Sourani and Villalonga for anti-Ole e 2

**Figure 2.** Immuno-reactivity analysis of profilin from 24 olive pollen cultivars. Reactivity of protein crude extracts from 24 olive cultivars was assayed against different maize profilin antisera, A) anti-PRA, B) anti-ZmPRO3, C) anti-ZmPRO5, D) anti-ZmPRO4, E) as well as against olive profilin antiserum anti-Ole e 2. Up to 3 reactive bands about 13, 13.7 and 14.2 kDa, corresponding to different isoforms of profilins were observed. The intensity of the reactive bands was quantitated by a densitometric analysis: yellow plot corresponding to 13kDa bands, pink (13.7 kDa) and blue (14.2 kDa). Red asterisks highlighted the differential reactivity (very high or very low) of defined cultivars to particular antisera, whereas blue asterisks highlighted the differential reactivity among cultivar to defined antisera.
