An actin-binding protein, LlLIM1, mediates calcium and hydrogen regulation of actin dynamics in pollen tubes

Abstract

Actin microfilaments are crucial for polar cell tip growth, and their configurations and dynamics are regulated by the actions of various actin-binding proteins (ABPs). We explored the function of a lily (Lilium longiflorum) pollen-enriched LIM domain-containing protein, LlLIM1, in regulating the actin dynamics in elongating pollen tube. Cytological and biochemical assays verified LlLIM1 functioning as an ABP, promoting filamentous actin (F-actin) bundle assembly and protecting F-actin against latrunculin B-mediated depolymerization. Overexpressed LlLIM1 significantly disturbed pollen tube growth and morphology, with multiple tubes protruding from one pollen grain and coaggregation of FM4-64-labeled vesicles and Golgi apparatuses at the subapex of the tube tip. Moderate expression of LlLIM1 induced an oscillatory formation of asterisk-shaped F-actin aggregates that oscillated with growth period but in different phases at the subapical region. These results suggest that the formation of LlLIM1-mediated overstabilized F-actin bundles interfered with endomembrane trafficking to result in growth retardation. Cosedimentation assays revealed that the binding affinity of LlLIM1 to F-actin was simultaneously regulated by both pH and Ca(2+): LlLIM1 showed a preference for F-actin binding under low pH and low Ca(2+) concentration. The potential functions of LlLIM1 as an ABP sensitive to pH and calcium in integrating endomembrane trafficking, oscillatory pH, and calcium circumstances to regulate tip-focused pollen tube growth are discussed.

Figure 1.

Spatial and temporal expression and protein structure of LlLIM1. A, LlLIM1 expression in different organs and different developmental stages of pollen were determined by semiquantitative RT-PCR. Three micrograms of lily total RNA extracted from root, leaves, stem, pistil, pollen grains, pollen tubes cultured in medium for 12 and 24 h, and anthers collected from flower buds of different lengths (10–20 mm, premeiosis; 20–30 mm, meiosis; 30–45 mm, microspore stage 1; 45–60 mm, microspore stage 2; 60–70 mm, mitosis; 70–90 mm, pollen maturation stage 1; 90–130 mm, pollen maturation stage 2; 130–150 mm, pollen maturation stage 3) were used for reverse transcription to obtain corresponding cDNAs for PCR, with the use of a LlLIM1 coding-region-specific primer set. rRNA amplified by the specific primer set was used as a loading control. B, Alignment analysis of the deduced amino acid sequences of LlLIM1, HaWLIM1 (sunflower; 85.1% similarity and 78.7% identity), AtWLIM1 (Arabidopsis; 84.7% similarity and 79.5% identity), NtWLIM1 (tobacco; 81.3% similarity and 76.2% identity), and OsWLIM1 (rice; 83.1% similarity and 77.9% identity). Black shading, identical and conserved amino acids; gray shading, similar amino acids; dark gray underline, LIM domain; asterisks, identical Cys or His residues for zinc finger motifs.

Figure 2.

LlLIM1 functions as an ABP in elongating lily pollen tubes and promote F-actin bundling, as revealed by cosedimentation assays. Confocal images show lily pollen tubes coexpressed with GFP and RFP:mTalin fusion genes (A), LlLIM1:GFP and RFP:mTalin fusion genes (B), and GFP:LlLIM1 and RFP:mTalin fusion genes (C). The images in D were from the lily pollen tube shown in C, except that the transformed pollen tubes were treated with 200 nm LatB for 1 h before recording the images. In all cases, hydrated pollen grains were cobombarded with 2.5 μg of each indicated plasmid, followed by germination in culture medium for 6 h before images were taken by fluorescent laser scanning confocal microscopy. Different fluorescent channels of images are indicated at the top of the panels. Medial sections through pollen tubes lying flat on the cover-slide surface are shown and represent at least 15 similar images collected from at least three independent experiments. DIC, Differential interference contrast. High- and low-speed cosedimentation assays were used to examine the capability for binding (G) and bundle assembly (H), respectively, of LlLIM1 recombinant proteins to F-actin. In F, 1 mg mL⁻¹ commercial F-actin, bovine serum albumin (BSA), and the recombinant proteins indicated were added to the binding solution and centrifuged at 100,000g for 45 min. Subsequently, equal amounts of pellets (P) and supernatants (S) were analyzed by SDS-PAGE and Coomassie Blue staining. Only F-actin, not the remaining soluble proteins, was precipitated after high-speed centrifugation. Various amounts (0–48 μm) of LlLIM1 (G and H) and 48 μm BSA (F) were incubated with 4 μm F-actin for 1 h, then centrifuged at 100,000g (G) or 12,500g (H) for 45 min and analyzed by SDS-PAGE. Equal amounts of pellets (P) and supernatants (S) were analyzed by SDS-PAGE and Coomassie Blue staining.

Figure 3.

Overexpression of LlLIM1 impaired pollen germination, tube growth, and pollen tube morphology. A and B, Lily pollen grains (P) bombarded with 7.5 μg of GFP-expressing plasmids, LlLIM1:GFP, GFP:LlLIM1, or cobombarded with 2.5 μg of GFP-expressing plasmid and 5 μg of LlLIM1-expressing plasmid were cultured in media for 12 h and observed by epifluorescence microscopy with a GFP filter. C, Projective images of elongating pollen tubes showed multiple pollen tubes emerging from a single pollen grain. Z-serial sections used to assemble the projective images were recorded and edited by fluorescent laser scanning confocal microscopy and an application program. D, Analysis of pollen tube lengths of pollen bombarded with 7.5 μg of the individual plasmids indicated or various combinations of LlLIM1- and GFP-expressing plasmids, but 2.5 μg of GFP-expressing plasmids was used as the minimal amount of transformation marker. A representative example from three independent data sets is shown: control pollen tubes overexpressing GFP showed normal morphology and average length (2.00 ± 0.10 mm), but the mean length of pollen tubes was reduced with increasing amounts of LlLIM1, such as bombardment with 7.5 μg of LlLIM1:GFP (0.31 ± 0.03 mm) or GFP:LlLIM1 plasmids (0.27 ± 0.02 mm) or cobombardment with 5 μg of LlLIM1 and 2.5 μg of GFP plasmids (0.3 ± 0.05 mm). Error bars indicate 95% confidence intervals (n = 60–80).

Figure 4.

LlLIM1 overexpression caused the mislocalization of signal molecules involved in the PLC signaling pathway in the clear zone of elongating pollen tubes. The images show the typical (A) and missed (B and C) subcellular localization of the fluorescent DAG marker, CYS1:GFP, in CFP-co-overexpressed (A) or LlLIM1-co-overexpressed (B and C) pollen tubes stained with FM4-64 to show the distribution of secreted/endocytic vesicles. B and C represent pollen tubes showing slow and stopped growth, respectively. In all experiments, pollen grains were cobombarded with 2.5 μg of expressing plasmid of the indicated florescent markers and 5 μg of expressing plasmid of CFP or LlLIM1, cultured in germination medium for 6 h, and then confocal images were recorded with the proper fluorescent channels. Images were obtained from the central section of pollen tubes lying flat on the cover-slide surface and represent typical examples from at least 10 similar images collected from at least three independent experiments.

Figure 5.

LlLIM1 overexpression-aggregated Golgi apparatus at the subapical zone of the elongating pollen tube. The subcellular localization of the fluorescent Golgi apparatus was revealed by the marker NAG:CFP in GFP-co-overexpressed (A; control) or LlLIM1:GFP-co-overexpressed (B and C) pollen tubes, which represent pollen tubes showing slow or stopped growth, respectively. In all experiments, pollen grains were cobombarded with 2.5 μg of expressing plasmid of the indicated fluorescent markers and 5 μg of expressing plasmid of GFP or LlLIM1:GFP, cultured in germination medium for 6 h, and then confocal images were recorded with the proper fluorescent channels indicated at the top of the panels. Images were obtained from the central section of pollen tubes lying flat on the cover-slide surface and represent typical examples from at least 10 similar images collected from at least three independent experiments.

Figure 6.

Periodic appearance of asterisk-shaped actin aggregates closely associated with oscillatory growth in pollen tubes moderately expressing LlLIM1:GFP. A, Serial images from one typical pollen tube bombarded with 3 μg of LlLIM1:GFP-expressing plasmids showed the oscillatory appearance of asterisk-shaped actin aggregates and growth after culture in germination medium for 6 h. Consecutive confocal time-lapse images were recorded at 9-s intervals with the GFP channels; white lines represent the initial length (at 0 s) of the pollen tube. Arrows indicate the subapical localization of asterisk-shaped actin aggregates. In all cases, the central section of pollen tubes lying flat on the cover-slide surface are shown and represent at least three similar serial images collected from at least two independent experiments. B, Diagram of the curve showing the continuing growth rate of the pollen tube obtained from A; asterisks indicate the time points around fast growth peaks when asterisk-shaped actin aggregates appeared. [See online article for color version of this figure.]

Figure 7.

In vitro high-speed cosedimentation assays revealed that [H⁺] and [Ca²⁺] simultaneously regulated the F-actin affinity of full-length and the N-terminal half of LlLIM1. High-speed (100,000g) F-actin in vitro cosedimentation assay samples were collected from reaction buffers containing 4 μm F-actin and 2 μm LlLIM1 (A–C), 6 μm LlLIM1N (D–F), or 28 μm LlLIM1C (G–I) under different pH conditions (pH 6.25–7, MES buffer; pH 7–8, Tris buffer) without (A, D, and G) or with (B, E, and H) 4 mm EGTA. C, F, and I were samples from reaction buffers containing 4 μm F-actin and 2 μm LlLIM1, 6 μm LlLIM1N, or 28 μm LlLIM1C under pH 6.25 conditions and in the presence of different amounts of EGTA (0–8 mm) used for cosedimentation assays with the estimated [Ca²⁺] indicated. In all experiments, after 1 h of incubation, the samples were centrifuged at 100,000g for 45 min, and the resulting pellet (P) and supernatant (S) fractions were analyzed by SDS-PAGE and Coomassie Blue staining. e[Ca²⁺], Free Ca²⁺.