The mechanism responsible for fluorescein isothiocyanate (FITC)-albumin internalization by columnar cells in culture obtained from the midgut of Bombyx mori larvae was examined by confocal laser scanning microscopy. Protein uptake changed over time, and it appeared to be energy dependent, since it was strongly reduced by both low temperatures and metabolic inhibitors. Labeled albumin uptake as a function of increasing protein concentration showed a saturation kinetics with a Michaelis constant value of 2.0 ± 0.6 μM. These data are compatible with the occurrence of receptor-mediated endocytosis. RT-PCR analysis and colocalization experiments with an anti-megalin primary antibody indicated that the receptor involved was a putative homolog of megalin, the multiligand endocytic receptor belonging to the low-density lipoprotein receptor family, responsible for the uptake of various molecules, albumin included, in many epithelial cells of mammals. This insect receptor, like the mammalian counterpart, required Ca2+ for albumin internalization and was inhibited by gentamicin. FITC-albumin internalization was clathrin mediated, since two inhibitors of this process caused a significant reduction of the uptake, and clathrin and albumin colocalized in the intermicrovillar areas of the apical plasma membrane. The integrity of actin and microtubule organization was essential for the correct functioning of the endocytic machinery.
- lepidopteran larval midgut
- columnar cells in culture
- albumin endocytosis
- clathrin-mediated endocytosis
the active absorption of proteins across the intestinal wall in mammals has been studied extensively (53). Transcytosis, the route followed by proteins to cross epithelial barriers, implies their internalization by endocytosis at one pole of the cell plasma membrane and, after vesicle-mediated internal transport, their release by exocytosis in the extracellular milieu of the opposite plasma membrane domain. The intracellular pathway followed by the vesicles containing the protein depends on the plasma membrane domain where endocytosis had occurred and implies their fusion with specific cell compartments (2).
Protein absorption by the insect midgut currently attracts increasing research efforts, in part fostered by the need of developing efficient strategies for oral delivery of bioinsecticides targeting hemocoelic receptors. To date, even though gut absorption of undegraded proteins has been unequivocally demonstrated in a number of insect species in vivo (1, 3, 18, 21, 22, 23, 29, 34, 37, 39, 46), it is still unclear which cellular pathway is involved in this process. We recently demonstrated in vitro that the isolated midgut of Bombyx mori larvae is able to perform transepithelial translocation of fluorescein isothiocyanate (FITC)-albumin (9) and horseradish peroxidase (10) by transcytosis. In vitro approaches provide a powerful tool to study in a simplified experimental context the pathways involved in the absorption process, which, however, have always to be interpreted with caution when placed in the more complex physiological scenario occurring in vivo. Even though with the isolated midgut it is possible to measure the transepithelial unidirectional fluxes and their modification under controlled experimental conditions, its use appears inadequate to study on a finer-grained scale the absorption process at the cellular level. The internalization of proteins and the sequence of the intracellular steps undertaken by the loaded vesicles can be better analyzed in single columnar cells of the larval midgut in culture. Mature columnar and goblet cells, the two main cell types present in the Lepidoptera midgut epithelium, can be obtained from the proliferation and differentiation of stem cells detached from the insect midgut epithelium and can survive in culture up to 6 mo (6, 30, 40, 47, 48). Thus, to define at cellular level the steps involved in the transcytosis of macromolecules in the insect midgut, we recently established a culture of midgut cells from B. mori larvae (12), which is used here to investigate albumin endocytosis.
The uptake by endocytosis of macromolecules starts with the formation of vesicles derived from the invagination and pinching off of plasma membrane portions. This phenomenon is a multifaceted mixture of multiple mechanisms that fall into two broad categories (16): phagocytosis (i.e., the uptake of large particles) and pinocytosis (i.e., the uptake of fluid and solutes). Although phagocytosis occurs in specialized cells, pinocytosis is common to all cells and occurs by four different modalities: macropinocytosis, clathrin-mediated endocytosis, caveolae-mediated endocytosis, and clathrin-and caveolae-independent endocytosis (16). The best-studied cell uptake mechanism is clathrin-mediated endocytosis, which ensures an efficient internalization of diluted solutes by high-affinity receptors concentrated in pits coated on their cytosolic side by clathrin and other associated proteins. These coated pits invaginate and pinch off to form the endocytic vesicles that are delivered to early endosomes, from which ligands and receptors will be addressed to their proper destination. A large number of regulatory proteins are involved in this process (16), and the cytoskeleton is implicated both in the formation of the endocytic vesicles and in their release from the plasma membrane (44, 51).
In the framework of a larger research effort, aiming at characterizing the transepithelial transport of proteins in the insect midgut, here we report experimental data on the endocytic mechanism responsible for albumin uptake in vitro by cultured columnar midgut cells from B. mori larvae. FITC-albumin was used as a model protein, and we examined in detail the general features of its uptake by columnar cells, the mechanism responsible for its internalization, the possible receptor(s) involved, and the role of the cytoskeleton.
MATERIALS AND METHODS
Isolated midgut cells in culture.
A detailed description of the protocol followed to obtain isolated midgut cells in culture from B. mori larvae is reported in Cermenati et al. (12), where all the modifications from the original procedure (48) are specified. Briefly, larvae of B. mori just before the fourth molt were anesthetized with CO2 and surface sterilized. Silkworms were cut between the second and the third pair of thoracic legs and behind the third pair of abdominal appendages, to exclude the foregut and the hindgut. Next, the midgut was deprived of the peritrophic membrane along with the enclosed intestinal contents. The central part of the larva was transferred to a petri dish containing a sterile Insect Physiological Solution (IPS) (47 mM KCl, 20.5 mM MgCl2, 20 mM MgSO4, 4.3 mM K2HPO4, 1.1 mM KH2PO4, 1 mM CaCl2, 88 mM sucrose, pH 7) modified by addition of 0.2% (vol/vol) gentamicin (50 mg/ml, Sigma), 0.01% (vol/vol) 1× antibiotic-antimycotic solution (Sigma), 0.003‰ (vol/vol) sodium hypochlorite. The cuticle of the ventral side was longitudinally cut, and the midgut was isolated. Midguts dissected from 8–10 animals were cut along the longitudinal axis and rinsed two times (10 min each) in sterile modified IPS, then two times (10 min each) in the same solution without sodium hypochlorite. Midguts were pooled in a strainer (100 μm pore size), placed in a petri dish containing a few milliliters of the latter solution, and left under mild agitation for 1 h. In these conditions, the loosely attached stem cells migrated away from the tissue. The tissue in the sieve was discarded, and the free cells in the filtrate, mostly stem cells (12), were collected and pelletted by gentle centrifugation at 400 g for 5 min. Cells were then resuspended in a growth medium (12) supplemented with 6 × 10−8 M 20-hydroxyecdisone (Sigma) and 100 ng/ml α-arylphorin [purified according to Blackburn et al. (6) in the Insect Bio-Control Laboratory, United States Department of Agriculture, Beltsville, MD] kindly donated by R. S. Hakim (Howard University, Washington DC). All solutions were sterilized by filtration (Nalgene, 0.2 μm pore size) before use. Cell suspension (3 ml) was distributed in the wells (35 mm in diameter) of six-well plates. Cells, which grow and differentiate in suspension, were incubated at 25°C, and 1 ml of medium from each well was replaced with 1 ml of fresh medium once a week. The cells to be used for the experiments were harvested from 3-wk-old cultures in which differentiated cells considerably exceeded stem cells: of total viable cells, as determined in three different experiments, 8.9 ± 0.6% were stem cells, 34.6 ± 3% were differentiating cells, 27.6 ± 3.4% were goblet cells, and 28.9 ± 2.7% were columnar cells (12). These last ones were characterized in their ability to internalize albumin since in vivo the completely differentiated columnar cells are responsible for digestion and absorption of molecules. For the confocal analysis (see Fluorescence acquisition and analysis), only those cells that could be considered fully mature columnar cells on the basis of their morphological features (12) were chosen.
Internalization of FITC-albumin by columnar cells.
The cultured cells were pelletted by gentle centrifugation at 400 g for 5 min, and 2.5 ± 0.6 × 104 cells were resuspended in 300 μl of the modified IPS, devoid of sodium hypochloride, for each experimental set up. The incubations, performed at 25°C unless otherwise specified, started when FITC-albumin, in the concentration indicated in the legends for Figs. 1–11, was added to the cells. At the end of the incubation, cells were rinsed three times in IPS and fixed for 10 min with 4% paraformaldehyde. After three rinses in PBS (in mM: 137 NaCl, 2.7 KCl, 4.3 Na2HPO4, and 1.4 KH2PO4), the samples were mounted in DABCO (Sigma)-Mowiol (Calbiochem), covered with a cover slip, and then examined with a confocal microscope as described in Fluorescence acquisition and analysis.
When drugs were used, the cells were preincubated for 30 min in the absence (control) or in the presence of the drug. The effect on FITC-albumin internalization of the following compounds was tested: 100 μM 2,4-dinitrophenol (DNP), 10 mM sodium azide, 100 μM chlorpromazine, 20 μM phenylarsine oxide, 27 μM nocodazole, and 20 μM cytochalasin D. If the drugs were dissolved in dimethyl sulfoxide (DMSO), control cells were incubated with a corresponding amount of solvent. In all these experiments, the cells were then incubated in the presence of 1.4 μM FITC-albumin for 20 min. Afterward, the cells were fixed and processed for confocal microscopy observations as reported above. To test if the drugs affected cell vitality, midgut cells were incubated for 30 min in the absence (control) and in the presence of either the drug concentrations used for the experiments or the corresponding amount of DMSO. The Trypan blue test showed that none of the experimental conditions induced dye diffusion into the cells (data not shown).
In the experiments in which the ability of 10 mM gentamicin to inhibit 1.4 μM FITC-albumin internalization was tested, cells were incubated for 20 min at 25°C in the absence (control) or in the presence of the polybasic drug.
To evaluate the effect of Ca2+ on 1.4 μM FITC-albumin internalization, cells were incubated for 20 min at 25°C in the absence (control) or in the presence of the Ca2+ chelator EDTA (20 mM). For these experiments, the incubation medium had the following composition (in mM): 47 KCl, 4.3 K2HPO4, 1.1 KH2PO4, 179 sucrose, pH 7, and 20 EDTA. At the end of the incubation, cells were rinsed three times in the same solution and fixed as reported above. Control cells were incubated in the modified IPS, devoid of sodium hypochloride, rinsed, and fixed as reported above.
Midgut cells were pelletted and resuspended in IPS as described in Internalization of FITC-albumin by columnar cells. After 20 min of incubation with 1.4 μM FITC-albumin at 25°C, cells were rinsed three times in IPS and fixed for 10 min with 4% paraformaldehyde. After three rinses in PBS, the samples were permeabilized for 4 min with 0.1% Triton X-100 in PBS and washed three times in PBS. Cells were then incubated for 15 min in PBS containing 1% BSA and for 1 h in the same buffer added with anti-clathrin heavy-chain mouse IgG (Affinity BioReagents) diluted 1:400 or with the anti-megalin rabbit IgG A55 (kindly donated by Dr. M. Marinò, Dipartimento di Endocrinologia, Università di Pisa, Pisa, Italy) diluted to 400 μg/ml. The cells were then washed three times in PBS containing 1% BSA and incubated for 1 h in the same buffer added with either Alexa Fluor 594-conjugated goat anti-mouse IgG antibody (Molecular Probes) diluted 1:1,000 or with Alexa Fluor 594-conjugated donkey anti-rabbit IgG antibody (Molecular Probes) diluted 1:1,000. After three rinses in PBS, the samples were mounted in DABCO (Sigma)-Mowiol (Calbiochem), covered with a cover slip, and examined with a confocal microscope (see Fluorescence acquisition and analysis). Controls were carried out in the same manner, except for omitting the incubation with the primary antibody (data not shown).
Immunodetection of microtubule organization.
Cultured midgut cells, pelletted and resuspended in IPS as described in Internalization of FITC-albumin by columnar cells, were incubated with or without 27 μM nocodazole. At the end of the incubation, cells were rinsed three times in IPS, fixed and permeabilized for 5 min with ice-cold methanol. After three rinses in PBS, cells were incubated for 15 min in PBS containing 1% BSA and for 1 h with the anti-α-tubulin mouse IgG (Sigma), diluted 1:500 in the same buffer. The cells were then washed three times in PBS containing 1% BSA and incubated for 1 h with Alexa Fluor 488-conjugated donkey anti-mouse IgG antibody (Molecular Probes), diluted 1:1,000 in PBS containing 1% BSA. After three rinses in PBS, the samples were mounted in DABCO (Sigma)-Mowiol (Calbiochem), covered with a cover slip, and examined with a confocal microscope (see Fluorescence acquisition and analysis). Controls were carried out in the same manner, except for omitting the incubation with the primary antibody (data not shown).
Detection of actin filaments.
Cultured midgut cells were pelletted, resuspended in IPS as described above, and incubated in the absence or in the presence of 20 μM cytochalasin D. At the end of the incubation, cells were rinsed three times in IPS and fixed for 10 min with 4% paraformaldehyde. After three rinses in PBS, the samples were permeabilized for 4 min with 0.1% Triton X-100 in PBS and washed three times in PBS. The cells were then incubated for 20 min with 4.3 μg/ml TRITC-phalloidin (Sigma). After three rinses in PBS, the samples were mounted in DABCO (Sigma)-Mowiol (Calbiochem), covered with a cover slip, and examined with a confocal microscope (see Fluorescence acquisition and analysis).
Fluorescence acquisition and analysis.
Fluorescence was acquired by using a confocal microscope CLSM TCS SP2 AOBS (Leica Microsystems Heidelberg) equipped with an argon ion laser (458, 476, 488, 496, or 514 nm excitation), two HeNe lasers (543, 594, and 633 nm excitation), and tunable emission wavelength collection. A ×63 Leica oil immersion plan apo (numeric aperture 1,4) objective and a ×2 zoom were used for all observations. FITC and Alexa Fluor 488 were excited with the 488 nm laser line, and the emitted fluorescence was collected between 500 and 560 nm; Alexa Fluor 594 was excited with the 594 nm laser line, and the emitted fluorescence was collected between 605 and 700 nm; TRITC was excited with the 543 nm laser line, and the emitted fluorescence was collected between 555 and 620 nm. To compare different experimental conditions (i.e., cells incubated with FITC-albumin for different time intervals, with different protein concentrations, or in the presence of various drugs), fluorescence acquisitions were always performed with the same hardware settings (laser intensity, sampling, acquisition rate, pinhole, and photomultiplier settings). To evaluate FITC-albumin internalization avoiding the contribution of unspecific binding to the cell membrane, a single optical section in a middle cell focal plane (where the nucleus was clearly evident) was acquired. Regions of interest, precisely defining the cell cytoplasm, were drawn, and the calculated mean gray values were used. Ten or more cells from at least two independent preparations were analyzed for each experimental condition. The data, expressed as arbitrary units of fluorescence intensity (8-bit acquisition), are reported as means ± SE. For each set of experiments, Student's t-test was used for statistical analysis. In all cases, P values ranged between <0.001 and <0.02.
Western blot analysis.
B. mori midguts isolated from fifth instar larvae were homogenized using a glass/Teflon Potter with nine strokes at 2,000 revolutions/min (IKA-Labortechnik RE 16 apparatus; Janke & Kukel) in 100 mM mannitol, 10 mM HEPES-Tris, pH 7.2, 1 mM dithiothreitol, 1 mM phenylmethanesulfonyl fluoride, 0.1 mM leupeptin, 2 μg/ml aprotinin, and 1 μM pepstatin A (all purchased from Sigma) (10 ml/g tissue). Protein concentration was determined according to Bradford (7) with BSA as standard.
Aliquots of midgut homogenates (50 μg of protein) were solubilized in sample buffer and resolved by 7.5% SDS-PAGE, as described by Laemmli (38). Proteins were transferred to nitrocellulose membranes at 350 mA for 90 min. Membranes were left overnight at 4°C in 150 mM NaCl, 50 mM Tris·HCl at pH 7.4, 5% wt/vol nonfat dry milk, and 0.1% vol/vol Tween 20 and then washed three times for 15 min in 150 mM NaCl, 50 mM Tris·HCl at pH 7.4, and 0.1% (vol/vol) Tween 20. Incubations with the primary antibody were performed for 1 h at room temperature using anti-clathrin heavy-chain mouse IgG (Affinity BioReagents) diluted 1:500 in 150 mM NaCl, 50 mM Tris·HCl at pH 7.4, 3% wt/vol nonfat dry milk, and 0.1% vol/vol Tween 20. Membranes were then washed three times (15 min for each wash), and the primary antibody was detected by the enhanced chemiluminescence method (Amersham Biosciences), with peroxidase-conjugated goat anti-mouse IgG (Santa Cruz Biotechnology) at 1:1,000 dilution as a secondary antibody.
Analysis of mRNA expression of a putative megalin homolog.
Total RNA (0.8 μg) isolated from B. mori larval midgut with Tri Reagent (Sigma), following the manufacturer's instructions, was used for RT to generate the cDNA with the ImProm-II Reverse Transcription System (Promega). The generated RT-cDNA was used for PCR amplification.
The primers for B. mori putative megalin homolog were sense 5′-TGGACTGGGTGGGCGACAAG-3′ and antisense 5′-TTCATGTGAGTGCCGTCCATGT-3′ (expected size for PCR product 208 bp) (see Identification of the receptor involved in FITC-albumin internalization in results). The conditions for PCR were: denaturation, 96°C for 30 s; annealing, 61°C for 30 s; extension, 72°C for 1 min (60 cycles). Amplification of B. mori cytoplasmic actin A3 (40 cycles) was included as internal control; the primers, designed according to the sequence available in the database (GenBank accession number U49854), were sense 5′-ATGTGCGACGAAGAAGTTGC-3′ and antisense 5′-CTCACCTGTTGGCCTTGG-3′ (expected size for PCR product 331 bp).
The PCR products with or without reverse transcription were separated by electrophoresis on 1.5% agarose gel and visualized under ultraviolet light with ethidium bromide.
Albumin uptake by columnar cells.
FITC-albumin uptake was measured at a fixed concentration of 1.4 μM. Figure 1A, shows the time course of the protein uptake by columnar cells evaluated by confocal laser microscopy. After 5 min of incubation, a weak fluorescence was present inside the cell, which progressively increased up to 30 min and showed no detectable changes for the remaining experimental time points considered. This was further corroborated by the time course analysis of fluorescence intensity in single optical sections (Fig. 1B), which clearly indicated that protein uptake reached a steady state in 30 min.
FITC-albumin transport was then characterized in cells incubated for 20 min. Albumin uptake was significantly inhibited when the incubation temperature was lowered from 25°C to 4°C (Fig. 2, A and B, respectively). From the fluorescence intensity recorded in single optical sections, a 60% inhibition of the uptake at low temperature was calculated (Fig. 2C). Moreover, in cells pretreated for 30 min with the two metabolic inhibitors DNP and sodium azide, a significant reduction of FITC-albumin transport was measured, with a 50 and 80% decrease of the measured fluorescence intensity, respectively (Fig. 3).
Figure 4 shows FITC-albumin uptake as a function of increasing extracellular protein concentration. The experimental values fitted a Michaelis-Menten equation, with a Michaelis constant (Km) of 2.0 ± 0.6 μM.
These data confirm that albumin uptake by columnar cells is mediated by an active process and indicate that a receptor is involved in the internalization process.
The endocytic mechanism involved in albumin uptake.
We examined if clathrin, a protein involved in coated pit formation, was expressed in B. mori midgut. Immunoblotting experiments, performed with an anti-clathrin heavy-chain primary antibody on midgut tissue homogenates, revealed a band of ∼180 kDa, corresponding to the molecular weight of clathrin heavy chain (Fig. 5A).
We verified the actual expression of clathrin also in columnar cells cultured in vitro by staining the cells, previously incubated for 20 min with FITC-albumin, with the antibody used in the immunoblotting experiment. As shown in Fig. 5B, clathrin (red) was effectively present in the intermicrovillar areas of the apical membrane and partially colocalized (white) with albumin (green), a clear indication of the possible involvement of clathrin in albumin internalization. Therefore, we tested the effect on albumin uptake of two inhibitors of clathrin-mediated endocytosis, chlorpromazine and phenylarsine oxide. Pretreatment of the cells with the two drugs significantly reduced FITC-albumin transport (Fig. 6), with a 75 and 50% inhibition of the uptake, respectively.
These data support a clathrin-mediated mechanism for albumin internalization.
Identification of the receptor involved in FITC-albumin internalization.
Various mammalian epithelial tissues express the multiligand endocytic receptor megalin, a member of the low-density lipoprotein (LDL) receptor family that recognizes molecules of different structure and function, albumin included (15, 43). A gene has been recently identified as a megalin homolog in Drosophila melanogaster (Flybase Gene ID CG12139, http://www.ensembl.org). Thus we investigated if a putative megalin homolog was expressed also in B. mori larval midgut. By BLAST analysis, we identified in a EST database of B. mori (http://www.ab.a.u-tokyo.ac.jp/silkbase/), a sequence (clone name NV021862) with a 74% identity with the putative megalin gene of Drosophila. The expression of the putative megalin homolog mRNA in B. mori larval midgut was examined by RT-PCR analysis, using specific primers that were designed across sequence regions highly conserved. PCR products of the expected size (208 bp) were observed in samples of the midgut after reverse transcription (Fig. 7). Positive (actin) and negative (no RT samples of RNA) controls gave the expected amplimer or no amplification products, respectively.
To prove that this receptor was involved in albumin uptake, we incubated columnar cells with FITC-albumin and then stained the cells with an anti-megalin primary antibody. The antibody recognized a protein in the apical region of the cell (red) that colocalized (white) with albumin (green) in the intermicrovillar areas of the plasma membrane (Fig. 8). Because binding of ligands to megalin is Ca2+ dependent (14) and megalin can bind and internalize several polybasic drugs, including gentamicin, in renal proximal tubule (42), we measured the internalization of FITC-albumin in the presence of the Ca2+ chelator EDTA or of the polybasic drug gentamicin. In these conditions, we observed a 62 and 40% inhibition of the protein uptake, respectively (Fig. 9, A and B). These experiments, together with the result obtained with RT-PCR, strongly support the conclusion that the receptor involved in albumin internalization is a putative megalin homolog.
Role of the cytoskeleton in albumin endocytosis.
The cytoskeleton is involved in a wide variety of cellular activities, including endocytosis and vesicle transport inside the cytoplasm. To assess if albumin endocytosis was regulated by microtubules and/or by actin filaments, we preincubated columnar cells with either nocodazole or cytochalasin D and then measured FITC-albumin uptake. As shown in Fig. 10, albumin uptake was significantly reduced by both drugs, with a 70 and 40% reduction of the measured fluorescence intensity with nocodazole and cytochalasin D, respectively (Fig. 10D).
To prove that the reduction of the uptake with the two drugs was due to their specific action on the cell cytoskeleton, columnar cells were incubated in the presence or in the absence (control) of nocodazole or cytochalasin D and then stained with an anti-α-tubulin antibody to disclose microtubules or with TRITC-phalloidin to detect microfilaments. Figure 11A shows the organization of microtubules in a control columnar cell and Fig. 11B its dramatic disorganization, especially at the apical pole, in nocodazole-treated cells. Therefore, the effect of the drug on albumin endocytosis was effectively associated with the disruption of microtubule organization. Figure 11C reports a typical columnar cell showing the arrangement of actin filaments: phalloidin distinctly stained microfilaments within the brush-border microvilli. Treatments with cytochalasin D had no visible effects on actin scaffolding (Fig. 11D), in spite of the fact that this drug clearly altered albumin endocytosis (Fig. 10, C and D).
In this study, we examined in vitro the mechanism involved in FITC-albumin uptake in isolated columnar cells obtained from the midgut of B. mori larvae. These cells in culture proved to be a powerful tool to disclose the mechanism responsible for the internalization of FITC-albumin, a model protein transported by transcytosis in the larval midgut of B. mori (9).
Albumin uptake linearly increased over time and reached a steady state after 30 min (Fig. 1) as a result of the balance between protein internalization and its subsequent fate, which can include, in addition to the transepithelial route (9), the recycling to the plasma membrane domain where the vesicle had originated and/or the targeted delivery to lysosomes for intracellular digestion (2, 20). The punctuate distribution of FITC-albumin inside the cytoplasm (Fig. 1A) is consistent with a vesicular compartmentalization, which is expected for a protein that follows these intracellular pathways.
Albumin uptake showed a marked temperature dependence (Fig. 2) and was inhibited by the two metabolic inhibitors DNP and sodium azide (Fig. 3). This indicates that the protein uptake is, as expected, energy dependent. Albumin internalization was a saturable process (Fig. 4) with a Km value of 2.0 ± 0.6 μM, suggesting that the protein could be taken up by receptor-mediated endocytosis and not by simple fluid-phase endocytosis. In this latter case, as highlighted by Gekle (28), the molecule, like the classical markers inulin or dextran (13, 45, 52), is not enriched at the plasma membrane surface, and its concentration in the endocytic invagination is the same as that present in the extracellular fluid. Thus molecule uptake is linearly correlated with its extracellular concentrations. Conversely, when a receptor is involved in the process, the ligand concentration in the endocytic invagination exceeds severalfold that of the extracellular fluid, and the uptake exhibits a saturation kinetics (24). Receptor-mediated endocytosis is clearly a more efficient mechanism of transport. The Km value for albumin, calculated in the kinetics experiments performed with B. mori columnar cells in culture, is in the same range of the apparent affinity constants for albumin endocytosis in mammalian absorptive cells (24, 25, 27).
The expression of clathrin in B. mori midgut epithelium (Fig. 5A) as well as in columnar cells cultured in vitro (Fig. 5B), and the colocalization of FITC-albumin with clathrin in the intermicrovillar areas of the apical plasma membrane of these cells (Fig. 5B) strongly indicate that clathrin is involved in albumin internalization. This conclusion is reinforced by the significant inhibition of albumin uptake observed in columnar cells preincubated with two inhibitors of clathrin-mediated endocytosis (Fig. 6). Chlorpromazine prevents assembly of the coated pit at the cell surface (54), whereas phenylarsine oxide presumably blocks endocytosis by cross-linking the clathrin coat (33).
No information is currently available regarding the molecular mechanisms of protein endocytosis in the insect midgut. In mammalian absorptive epithelia megalin, a multiligand receptor plays a fundamental role in clathrin-mediated endocytosis of different ligand. Megalin is a 600-kDa transmembrane protein belonging to the LDL receptor family, acting in mammalian polarized epithelia as a receptor for hormones, vitamin-binding proteins, enzymes, enzyme inhibitors, and proteins like albumin and lactoferrin (15, 43). It is abundant in renal proximal tubule, where it is responsible for the tubular reabsorption of filtered proteins (28). In the same epithelium, it also acts as a membrane anchor for cubilin, a second multiligand receptor for albumin (28) that has no transmembrane domain and little structural homology with other known endocytic receptors (15, 43). In mammals, both megalin and cubilin are expressed in the intestinal brush border, where they are involved in the gastrointestinal uptake and transport of vitamin B12 and folate (4, 5, 57).
Because our data show that in B. mori columnar cells a receptor is involved in albumin internalization via clathrin-mediated endocytosis, we tested the suggestive hypothesis that a megalin-like receptor was responsible for albumin recognition in the insect midgut. This was stimulated and made possible by the presence of a megalin homolog in the genome of both D. melanogaster and B. mori. RT-PCR analysis (Fig. 7) as well as colocalization (Fig. 8) and inhibition (Fig. 9) experiments indicate that a putative megalin homolog involved in albumin endocytosis is indeed expressed in B. mori columnar cells. As far as we know, this is the first time that a functional description of a putative megalin homolog is reported for an insect. Megalin expression is not restricted to mammals: it plays a role in the development of central nervous system in zebrafish (41), and the megalin homolog LRP-1 (49) is essential for growth and development in the nematode Caenorhabditis elegans (32, 59). These reports, the recent identification of the putative megalin homolog in D. melanogaster, and the data reported here indicate that this protein is highly conserved in evolutionary distant organisms.
It is, of course, logical to question what could be the role in vivo of the megalin-like receptor in the midgut of lepidopteran larvae, especially if we consider that albumin, a protein so readily recognized and internalized, is absent in the silkworms diet. We can speculate that the receptor may be responsible for vitamin B absorption like in mammalian intestine and that it may be involved in the uptake of those proteins present in the diet that escape degradation by digestive enzymes. Actually, in different insect species, dietary (37) or exogenous proteins added to the diet can avoid digestion and enter the hemolymph undegraded (1, 3, 18, 21, 22, 23, 29, 34, 39, 46). The mechanism involved in the absorption of at least two proteins in B. mori larvae was trancytosis (9, 10). Therefore, the megalin-like receptor could mediate their transepithelial transport but could also provide substrates for columnar cell metabolism, since proteins, once internalized, can be directed to the intracellular degradative pathway, to supply further free amino acids to midgut cells.
Microtubules are dynamic protein filaments that provide a mechanical basis for cell polarity and for transport of organelles and vesicles within the cell. These elements of the cytoskeleton are critical players in endocytosis, since drugs that interfere with their organization reduce receptor-mediated endocytosis in mammals (2, 19, 26, 31), although a detailed analysis of their role is still lacking. As discussed in detail in Cermenati et al. (12), in B. mori cultured columnar cells (Fig. 11A) microtubules are arranged in bundles oriented longitudinally from the apical to the basal pole of the cell as in other absorptive epithelia (55), whereas, at odds with the latter, numerous microtubules run parallel to the apical cell surface just below the microvilli, forming a dense network. The cells treated with nocodazole displayed a remarkable disorganization of microtubules (Fig. 11B) and a strong inhibition of FITC-albumin internalization (Figs. 10, B and D), proving that also in the insect midgut these cytoskeletal elements are critical for the endocytic process.
Actin is an ubiquitous eukaryotic protein that forms dynamic polar microfilaments of the cell cytoskeleton. Unequivocal data on the involvement of these filaments in endocytosis come from studies on budding yeast, in which actin nucleation at the endocytic site, its role in driving membrane invagination and vesicle scission, as well as the function of actin-regulatory proteins have been thoroughly clarified [recently reviewed by Smythe and Ayscough (51) and Kaksonen et al. (36)]. In mammals, different orthologs of yeast proteins that cap, bundle, and stabilize actin are present (51) and, like in yeast, membrane invagination and vesicle scission are affected by the depolymerization of actin filaments (58). The localization and timing of actin polymerization in yeast and mammal (35, 36, 50) are strikingly similar, and it is apparent that the endocytic internalization processes in these so different organisms are variations of a same ancestral theme. In B. mori columnar cells, cytochalasin D induced a significant reduction of FITC-albumin endocytosis (Fig. 10, C and D), suggesting that actin structures involved in the formation of the endocytic vesicles in these insect cells follow the same pattern described for yeast and mammalian cells. However, a clear-cut disorganization of actin filaments was not apparent (Fig. 11D), so we assume that cytochalasin D concentration was adequate to alter actin filaments involved in the endocytic process but insufficient to cause a significant alteration of actin cytoskeleton in the microvilli and in the submicrovillar area, where microfilaments in insects are present in considerable density (17).
Perspectives and Significance
Gut absorption of intact proteins in insects had been unequivocally demonstrated in vivo (1, 3, 18, 21, 22, 23, 29, 34, 37, 39, 46), but the cellular pathway involved in the process was unknown, and only speculations on the mechanism were advanced. We showed recently that albumin, a protein absent from the insect diet and chosen as a model molecule, was readily absorbed by the isolated B. mori larval midgut by transcytosis (9), and now we have characterized, for the first time in an insect, the mechanism of its internalization by midgut cells. Unexpectedly, the receptor and the endocytic mechanism involved resemble those described in the kidney epithelium for the reabsorption of plasma proteins, albumin included. To understand how conserved this physiological process is, further studies are undoubtedly necessary. We intend then to define the possible coexpression of a cubilin homolog, which, if present, would further reinforce the emerging notion that absorption mechanisms of intestinal epithelia in insects and mammals share important functional similarities at the cellular and molecular level, as we recently demonstrated also for sugar absorption in a parasitic wasp (8). We want to detect how similar is the specificity of the lepidopteran receptor(s) for the large number of megalin and, possibly, cubilin ligands of the mammalian counterpart(s) (15). Finally, we want to disclose the effective functional role of the receptor for the insect physiology. We are confident that the future directions of our research will lead to a more complete knowledge of the physiological role of these multiligand endocytic receptors from an evolutionary point of view.
Moreover, our study contributes to a current innovative research effort aiming at defining new strategies for insect control, including agronomical pests and insects vectors for a multitude of human pathogens. Since the 1940s, with the introduction of dichloro-diphenyl-trichloroethane (DDT), insect pest control has been performed essentially with chemical insecticides (11). The well-documented environmental and health impact of these chemicals, their poor species specificity, and the rapid development of insect resistance have stimulated the search for new insecticidal compounds, novel molecular targets, and alternative control methods. In the last decades, a variety of biocontrol methods employing peptides and proteins derived from microorganisms, animals like predator or parasitoids artropods, and plants has been examined [recently reviewed by Whetstone and Hammock (56)]. In most cases, these proteins have hemocoelic targets and must pass the gut barrier undegraded to exert their activity. For their successful delivery, it is essential to develop basic information on the molecular mechanisms mediating the absorption of macromolecules by the insect midgut, which is a physiological process still poorly understood.
This work was supported by the Italian Ministry of University and Research (COFIN 2006, project no. 20060794417).
We thank Dr. M. Beltrame (University of Milan) for support in the RT-PCR experiments and Dr. M. Marinò (University of Pisa) for the supply of the anti-megalin antibody.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society