The natural adhesive of the goose barnacle Lepas anatifera: The functional morphology and chemistry of the adhesive gland and an investigation of the adhesive proteins

Abstract

The adhesive and adhesive glands of the stalked barnacle Lepas anatifera were investigated using a wide range of techniques to characterise them and compare them to distantly related acorn barnacles, as well as other adhesive models such as mussels, tubeworms and echinoderms. Live samples of L. anatifera were collected from Irish coastal waters and maintained in a laboratory aquarium. Histochemical and immunological analyses showed that the adhesive production system in barnacles was unique compared to other adhesive animal models and showed that, within the body, the adhesive is slightly acidic, contains some carbohydrate and does not contain L-3.4-dihydroxyphenylalanine or phosphorylated serines. Ultrastructural analyses revealed a series of single celled adhesive glands, drained by ever-larger canals. The first, the intracellular canal, is immediately adjacent to the gland cell cytoplasm and responsible for transferring the secretion granules across the cell membrane through apocrine and merocrine processes. Once they had been released from the gland cell cytoplasm, the adhesive components were completely unbound by granular membranes as they made their way through the adhesive drainage system. These observations raise questions about how premature hardening of the adhesive is prevented.

Because of their relatively large size, it was possible to elementally map the gland cells for the first time using SEM-EDS. These analyses showed that the gland cells had a far less varied elemental composition than the adhesive. A wider range of elements than has been recorded previously was observed in the adhesive, although some of these were probably environmental contaminants. Elements were sometimes variable, for example, phosphorus was sometimes present in the gland but mostly absent from the adhesive. Sulfur was absent from the gland but always present in the adhesive. FTIR and Raman spectroscopy indicated that the sulfur content of the adhesive was not present in disulfide form. Spectroscopic analyses confirmed the absence of phosphorylated proteins and protein-metal interactions. CaCO3 involvement in barnacle adhesion was also ruled out but Raman spectroscopy showed a signal consistent with tyrosine in the adhesive. There were marked differences in Raman spectra between stalked and acorn species.

The hardened adhesive was solubilised most successfully by increasing the amount of urea/thiourea buffer prior to 1D SDS-PAGE. Approximately 12 bands were observed ranging in size from less than 20 kD to 200 kD. Repeating 1D SDS-PAGE showed consistent bands at 30, 70, 80, 90, 110 kD. Mass spectrometry of these protein bands showed no significant similarity to any known proteins, including published barnacle adhesive proteins. De novo sequencing of adhesive protein bands produced seventy-eight peptide fragments of up to 16 amino acids in length. The longest de novo sequences (more than 11 amino acids) were subjected to multiple BLAST analyses and compared to translated cDNA open reading frame sequences from a transcriptome of Amphibalanus amphitrite and an unpublished database for Tetraclita, but these returned no matches. Several of the de novo sequences were repeated across several of the bands that were analysed, indicating possible multiple variants of the same protein. There was some limited similarity in the protein size bands and peptide sequences of L. anatifera and related species Dosima fascicularis, but less similarity to size ranges of acorn barnacle proteins. All attempts to amplify adhesive protein genes using PCR with primers based on de novo peptide sequences were unsuccessful.

Immunohistochemical assays using polyclonal antibodies raised against acorn barnacle adhesive proteins resulted in positive reactions in L. anatifera adhesive gland tissue for two out of the three proteins tested, suggesting that some homology exists between stalked and acorn barnacle adhesive. However, extensive attempts to isolate adhesive protein genes from L. anatifera cDNA using primers based on acorn barnacle adhesive protein gene alignments, as well as acorn and stalked barnacle cDNA sequences for adhesive genes, were unsuccessful. Homology between acorn barnacles and Pollicipes pollicipes sequences ranged from 26% to 36% depending on the gene. Overall, there were strong distinctions between the adhesives of distantly related (different taxonomic order) barnacle species. The overall picture which emerges is one where the functional homology of barnacle proteins does not strictly depend on the characteristics of size, pI or primary sequence similarity.