The fish skin is a living, mucosal tissue where the mucus is essential to protect the animal [1
]. Fish skin mucus is a rich source of molecules with immune relevant functions [2
]. Lectins are proteins that recognize sugar moieties through their lectin domain(s) and hence bind to carbohydrates, glycolipids and glycoproteins. Lectins can also bind directly to proteins [3
] or adenine [4
] in a non-carbohydrate dependent manner.
The diversity in binding partners results in multiple functions of lectins. In mucosal surfaces they can contribute to immune defense by agglutinating pathogens and inhibit uptake [5
]. However, lectins can also contribute to pathogenesis probably by stimulating pathogen uptake [6
]. Extracellular functions of lectins include opsonization, cellular uptake through endocytosis and phagocytosis, start of the lectin pathway of the complement system and cell-cell interaction. Intracellularly lectins have diverse roles including roles in protein folding [7
], in organelle movement, in uptake (into endocytic, autophagic and lysosomal organelles) and subsequent destruction of virus and bacteria found in cytosol [8
], as well as roles in mRNA splicing and stability [9
]. Lectins were first purified and characterized from plants; later lectins have been reported from microorganisms to humans, including aquatic animals such as teleosts [10
We have previously isolated and characterized β-galactoside binding galectin 1-1 and galectin 1-2 [13
] from Atlantic cod Gadus morhua
skin mucus. In the present study we aim to isolate galectin from Atlantic salmon skin mucus by affinity purification by lactose-binding. Three groups of galectins exist, the prototype galectins where the whole protein is essentially a globular carbohydrate binding domain (such as in galectin-1), chimera type galectins with a N-terminal tail in addition to the carbohydrate binding domain (galectin-3) and tandem repeat galectins where there are two carbohydrate binding domains (such as galectin-4). Previously characterized skin and/or skin mucus galectins AJL-1 from Japaneese eel (Anguilla japonica
], Congerin I and II from conger eel (Conger myriaster
], and galectin 1-1 and 1-2 from Atlantic cod (Gadus morhua
] were all prototype galectins, hence the isolation of prototype galectin(s) from Atlantic salmon skin mucus was expected to be the outcome of this study.
Further we aim to identify and characterize the isolated protein(s) and we hypothesize that galectin present in skin mucus can bind to and affect bacteria relevant for skin disease. The bacteria used Moritella viscosa
(formerly named Vibrio viscosa)
cause skin ulcers, winter ulcers, in Atlantic salmon at low water temperatures [16
Mucosal surfaces of animals including humans have living cells that are vulnerable for invasion of pathogens. The molecular defense system in fish skin include the mucus that both provides a physical barrier and is a reservoir for defense molecules [1
]. One group of defense molecules is the lectins, glycan binding proteins with a large range of specificities and immune related functions. Galectins can for example both stimulate and inhibit inflammation and galectin-1, -2, -3, -4, -7, -8, -9, and -12 can all induce T-cell apoptosis [18
]. Galectins can also work as opsonins to stimulate phagocytosis, they are produced and secreted by many cells including immune cell (reviewed in [19
The galectins that were previously characterized in skin and/or mucus are prototype galectins consisting of only the globular carbohydrate recognition domain [13
], the size of our isolated protein (Figure 1
) fits with the expected molecular weight of carbohydrate recognition domains of approximately 15 kD. However, the Atlantic salmon skin mucus galectin was identified as galectin-3 by mass spectrometry. Galectin-3 is a chimeric galectin with both a N-terminal domain and a C-terminal carbohydrate recognition domain, and an expected molecular weight of ~30 kD. The observed molecular weight of the isolated protein was however around 15 kD, as expected for prototype galectins. This discrepancy in observed and expected molecular weight was explained when further analysis showed that we isolated a truncated form, galectin-3C, with only the C-terminal carbohydrate recognition domain. In accordance with our observation, in humans galectin-3 (~30 kD) can also be found in a truncated form due to cleavage by metalloproteases [20
], resulting in a 22 kD protein. Galectin-3 is a multifunctional protein, the protein’s location and activity is thought to be modulated by post-translational modifications and different interaction partners. In mammals including humans, results suggest that extracellular full length galectin-3 promotes transformation and metastasis of tumor cells, whilst galectin-3C can inhibit metastasis [21
], showing that the post-translational cleavage of galectin-3 affects its function.
Galectin-3 does not have a signal peptide for translation on the endoplasmic reticulum, for later export out of cells through the endoplasmic reticulum-Golgi secretion pathway. It is thought to be exported from the cytosol by non-classical mechanism(s) [22
] likely through exosomes [17
]. The N-terminal of the protein is needed for secretion [17
] and a sequence needed for import into multivesicular endosomes and export in exosomes, P(S/T)AP [17
], is present once as PTAP in the N-terminal of salmon galectin-3. For humans it has been shown that the N-terminal domain is sensitive to metalloproteases [20
]. Metalloproteases are present in the skin mucus of several fish species [25
] including Atlantic salmon [26
], and could hence be responsible for removal of the N-terminal resulting in truncated galectin-3 in skin mucus.
We found the highest expression of galectin-3 mRNA in skin and gills. High skin mRNA levels could suggest that that galectin-3C is made in its full-length form in skin cells, is exported with the N-terminal intact and then cleaved by metalloproteases in skin mucus. However, further studies are needed to confirm this hypothesis.
Galectin-3 can polymerize using its N-terminal usually to make pentamers [27
]. Dimers can also be formed with interactions between two C-terminals [28
]. The dimerization of the C-terminal part can be inhibited by the addition of glucan ligand [28
]. Galectin-3C isolated in this study gives hemagglutination and bacterial agglutination in a lactose inhibitable manner.
is a bacterium that causes ulcers in the skin of Atlantic salmon [16
]. The binding of galectin-3C to Moritella viscosa
could be part of defense against the bacterium. Bacterial aggregates will have difficulties in penetrating host cells and can be sloughed off the fish with skin mucus during swimming. If the level of molecules that cause bacterial agglutination, including galectin-3C can be modulated for example with functional feeds, this holds a potential to reduce infection by pathogens that enter through mucosal surfaces. Agglutination of bacteria by immunoglobulin hinders bacterial colonization and also transfer of bacteria between animals [29
]. Some galectins such as galectin-4 and 8 can directly kill bacteria [30
], the killing activity is specific and restrained to only some bacteria [30
]. Mucus galectin-3C did not inhibit growth of Moritella viscosa
as assessed by measurement of turbidity nor did it produce clear zones in a zone of inhibition test [31
] (results not shown). There was an increased optical density in the presence of galectin-3C at later time points (Figure 6
), this could be explained by the formation of aggregates (Figure 7
). This is in accordance with findings in a study of chitosan introduced aggregates of Escherichia coli
]. Low concentrations of chitosan give abrupt increase in optical density in the Escherichia coli
growth curve, but measurement of green fluorescence (from the bacterial strain’s green fluorescent protein (GFP)) amount indicates that there are fewer rather than more bacteria present than in the control samples. Scanning electron microscopy shows bacterial aggregates in samples with abrupt optical density increases, but not in control samples or at chitosan concentrations where no sudden increases in optical density was observed [32
]. In the presence of lactose the bacteria grew slower, in accordance with the known inhibition of bacterial growth of carbohydrates due to hyperosmotic extracellular fluid and in accordance with a previous study showing no oxidative production of acid for the bacteria in the presence of lactose [33
]. Lactose inhibited galactose-3C induced hemagglutination. Limited bacterial agglutination was observed in samples incubated with both galectin-3C and lactose. In both the presence and absence of lactose, galectin-3C aggregate formation formed at later time points. This could indicate that the aggregates are forming on the galectins as the bacteria divide, and not by simple aggregation of already present bacteria. Aggregate formation by bacterial division on immunoglobulin A has been shown to be responsible for aggregate formation in the mucosal surface in the gut [34
To further study the interaction between galectin-3C and Moritella viscosa
, the bacteria incubated with and without galectin-3C was studied by two-dimensional gel electrophoresis. Increased spot densities for multi drug transporter, high-molecular weight cobalt-containing nitrile hydratase subunit alpha, transcription termination factor Rho, and ribosomal proteins, were found. The multidrug transporter is used to export harmful substances from bacteria, and also is involved in cell-cell contact and biofilm formation [35
]. Nitrile hydratases hydrates harmful nitriles to amides, hence protect the bacteria. It can be upregulated in stressed bacteria as shown in bacteria treated with the pesticide linuron [36
]. Transcription termination factor Rho is as the name says involved in termination of transcription, it is targeted by the antibiotics bicyclomycin. Interestingly, knock down of Rho has been linked to increased Staphylococcus aureus
], indicating that bacteria under less favrable conditions increase their virulence after Rho silicing.
The bacterial ribosome consists of two parts, the 50S and the 30S subunits. Since ribosomes are key for cell growth and survival, targeting them can kill cells, there are antibiotics targeting both the 30S [38
] and 50S [39
] subunits. Galectin-3C induced spot differences in 50S ribosomal protein L7/L12 (two spots), 30S ribosomal protein S13, and 30S ribosomal protein S2. Typically changes in ribosomal subunits for example have been observed when bacteria are in non-favorable conditions and proteomics changes in ribosomal subunits have been found with antibiotics that are inhibitors of translation elongation [40
Overall, the changes observed in Moritella viscosa indicate that galectin-3C is not simply binding to the bacteria and aggregating them, but are also inducing changes in signaling pathways to change proteins in the bacterial cytoplasm. Further studies are needed to understand the interaction between galectin-3C in the host immune system and pathogens, and whether galectin-3 and galectin-3C have the same function.
In mammals galectins are focused upon because they have important roles in human disease as well as regulating functions such as inflammation and metabolism [18
]. For experimental purposes they are targeted with inhibitors and their amounts increased by injection. Galectins from other species, including fish, could have potential to be used to inhibit or kill bacteria and to promote wound healing. A galectin from Japanese eel Anguilla japonica
, AJL-1 inhibits biofilm formation by human bacteria important for periodontal disease [41
]. Fish skin has antibacterial and wound healing promoting properties and are appealing for use in human medicine as there is less risk of disease transfer including prions than when mammalian grafts are being used [42
]. Since galectins can kill bacteria [30
], inhibit bacterial biofilm formation [41
] and have wound healing properties [43
], at least some of the wound healing property of fish skin could be due to the presence of galectin. Galectins could be isolated from by-products from industrial processing of fish. However, additional research is needed to establish the effectivity and biosafety of fish galectins on human pathogens and human wound healing.