Emptying and refilling of slime glands in Atlantic (Myxine glutinosa) and Pacific (Eptatretus stoutii) hagfishes

Hagfishes (Myxini) are known for their unique defensive slime (Downing et al., 1981a; Newby, 1946), which is an effective defense against attacks by gill-breathing predators (Lim et al., 2006; Zintzen et al., 2011). The slime is produced by many slime glands that occur in a line down both sides of the hagfish. The glands are connected to the skin surface by a short duct, and are surrounded by striated muscle and a connective tissue capsule (Downing et al., 1981b; Blackstad, 1963; Koch et al., 1991). The slime glands contain two main secretory cells: gland thread cells and gland mucous cells. The gland thread cells each contain a single, elaborately coiled protein-rich thread formation, known as a skein, which is precisely bundled and organized into stacked thread loops within the cell (Winegard et al., 2014; Downing et al., 1984; Spitzer et al., 1984). The gland mucous cells are filled with hundreds of tiny mucus-containing vesicles (Luchtel et al., 1991; Salo et al., 1983). When the musculature around the slime glands contracts, it initiates a rapid holocrine secretion event; the two gland components are forced out through the narrow gland pore and, in the process, their plasma membranes are sheared off (Downing et al., 1981a,b). In seawater, the thread and mucus components interact to yield an ultra-dilute mucous gel (Downing et al., 1981b; Spitzer et al., 1987; Fudge et al., 2005; Ewoldt et al., 2011).

The mass of slime exudate from all the glands of a Pacific hagfish (Eptatretus stoutii) has been estimated to be 3–4% of body mass (Fudge et al., 2005), and therefore represents a significant energetic commitment. In spite of the large amounts of pre-exudate stored in the slime glands, hagfish appear to be economical with their use of the slime in two important ways. First, the slime is remarkably dilute, with mucus concentrations three orders of magnitude lower than those of other mucous secretions, such as gastric mucus (Fudge et al., 2005). Second, hagfishes conserve slime by only releasing exudate from glands in the vicinity of an attack, and not as a whole-body response (Lim et al., 2006; Zintzen et al., 2011). The large number of slime glands possessed by hagfishes (∼79 pairs in Pacific hagfish and ∼97 pairs in Atlantic hagfish) and their ability to release exudate locally likely reduce the chance that a hagfish will be left without slime to defend itself in subsequent attacks (Fernholm, 1998). Vulnerability to predators will also be affected by how many times a single gland can release exudate before it is empty, and how quickly depleted slime glands can recharge, both major foci of this study.

Although the time scale for slime gland recovery after a sliming event has not been documented in any species of hagfish, Lametschwandtner et al. (1986) speculated about the relative rates of recovery in Pacific hagfish and Atlantic hagfish (Myxine glutinosa) based on differences in the vascular anatomy of the slime glands in these two species. In Atlantic hagfish, capillaries form a cage around the slime gland, whereas in Pacific hagfish, capillary loops also descend into the gland interior, which may increase the rate of nutrient delivery to the slime glands and decrease the recovery time in this species (Lametschwandtner et al., 1986).

In this study, we aimed to answer several fundamental questions about the refilling of slime glands in hagfishes. Using a population of captive Atlantic and Pacific hagfishes, we measured the time it takes for slime glands to refill after they have been emptied of slime exudate. We also investigated the possibility that a single slime gland can eject multiple boluses of exudate and, if so, whether the composition of the exudate changes with successive ejections. In addition, we examined the effects of recovery time after a sliming event on the composition of exudate in a gland, the relative proportions of threads and mucus in the exudate, as well as the morphometrics of the thread skeins.

Pacific hagfish [Eptatretus stoutii (Lockington 1878)] were collected from Bamfield Marine Station in Bamfield, BC, Canada, and Atlantic hagfish (Myxine glutinosa Linnaeus 1758) were collected from Passomoquody Bay, NB, Canada. All hagfishes utilized in this study were adults, but the age and sex of each individual was unknown. The two species were housed together in a 2000 l environmentally controlled aquatic recirculating system filled with chilled artificial seawater (34 ppm, 10°C) at the Hagen Aqualab at the University of Guelph, Guelph, ON, Canada. Hagfish were isolated in floating bins within the tank for a minimum of 1 month (30 days) prior to our experiments, to make sure their slime glands were completely full. Hagfish were fed squid to satiety once per month. All housing and feeding conditions were approved by the University of Guelph Animal Care Committee (Animal Utilization Protocol #2519).

Hagfish were anesthetized by placing them in 3 l artificial seawater (Coralife, Energy Savers Unlimited, Inc., Carson, CA, USA) with 3 ml of a clove oil (Sigma-Aldrich, Oakville, ON, Canada) anesthetic solution (1:9 clove oil to 95% ethanol) and left until they ceased to respond to touch. Once fully anesthetized, the hagfish were removed from the artificial seawater, rinsed with deionized water, patted dry using Kim Wipes (Kimberly-Clark Corporation, Irving, TX, USA) and placed on a dissection tray. The slime glands of the hagfish were induced to secrete slime exudate via mild electrostimulation (60 Hz, 18 V) using a Grass SD9 electric stimulator (Grass Instruments, Quincy, MA, USA). The expressed exudate was collected from the skin using a scoopula, and either gently stirred into a 0.9 mol l⁻¹ sodium citrate (Fisher Scientific, Ottawa, ON, Canada), 0.1 mol l⁻¹ Pipes [piperazine-N,N′-bis(2-ethanesulfonic acid); Sigma-Aldrich] stabilization buffer (pH 6.7) or collected onto a pre-weighed paper towel (Fudge et al., 2003). If a hagfish produced slime at any point during its handling, it was excluded from the trials.

Samples from successive stimulations of the same glands were collected and analyzed to test whether the composition of the exudate changes as more and more exudate is expressed from the glands. Pacific hagfish (n=20) with full glands were anesthetized, and ten glands posterior to the gill pouches were stimulated. A single stimulation was defined as touching the stimulator electrodes to each slime gland just long enough for exudate to be released. The exudate produced from the first five glands during each stimulation was pooled and collected into individual microfuge tubes (Fisher Scientific) containing 1.5 ml stabilization buffer to be used for later analysis (one tube per stimulation). Exudate from the final five glands was collected onto pre-weighed paper towels to determine the mass of exudate produced. Exudate from each stimulation was collected in this manner until the glands were exhausted. The time between stimulations ranged from 15 to 30 s, depending on the time needed to collect samples from the previous stimulation. Microfuge tubes containing samples of exudate from each stimulation were gently mixed by inversion, and 20 µl subsamples were removed and analyzed as described above.

To determine how long it takes for hagfish slime glands to refill, a series of exudate collection trials were conducted using 100 individuals of similar mass from each species (Pacific: 51.54±2.53 g; Atlantic: 42.14±1.23 g), with ten individuals used for each of the ten time intervals examined. For each recovery time treatment, all glands on the left side of each specimen were exhausted completely (i.e. they no longer produced exudate) at time zero using repeated electrostimulation, while glands on the right side were unstimulated (‘full’) and served as an internal control. Experimental hagfish were kept separate from others by placing them in a smaller floating container in the tank. Hagfish were isolated for various time intervals (4 days, 8 days, etc.) and removed from the tank at desired days post-sliming to have their slime exudate collected. Each individual had one refilling time point (left side of hagfish) and full glands (right side of hagfish) from which slime was collected. After recording hagfish body mass, we collected slime from each side of the animal separately onto pre-weighed paper towels to determine the mass of slime produced by the refilling side and the full side of the animal.

Exhausted and unstimulated glands were harvested from five Pacific hagfish for histological analysis. Hagfish were anesthetized then euthanized by severing the notochord and dorsal nerve cord using a large pair of surgical scissors. Hagfish were dissected according to a protocol developed by Winegard (2012), which allows the slime glands to be separated from the myotomal muscle in which they are embedded while still attached to the skin at the gland pore. Dissected slime glands were placed in fixative for later histological analysis (see below).

Harvested gland tissue was fixed in 4% paraformaldehyde (Electron Microscopy Sciences, Fort Washington, PA, USA) solution in preparation for paraffin embedding. Paraformaldehyde-containing fixatives were prepared in stabilization buffer (0.9 mol l⁻¹ sodium citrate, 0.1 mol l⁻¹ PIPES; pH 6.7) to reduce swelling of the mucous vesicles within the gland. Fixed tissues were processed for paraffin embedding in the Ontario Veterinary College Veterinary Histology Unit using a routine overnight protocol. Hagfish slime gland diameter was measured with calipers, and the glands were trimmed and sagittally sectioned (5 μm thick sections) to their approximate center using a rotary microtome. Sections were placed on frosted glass microscope slides (Fisher Scientific) for histological staining. Whole glands were stained with hematoxylin and eosin (H&E; Fisher Scientific), and slides were covered with a cover glass (22×50 mm; Fisher Scientific) and sealed with Cytoseal XYL (Richard Allen Scientific, San Diego, CA, USA). The Cytoseal was allowed to dry overnight before visualization.

Slides with exudate and histology samples were analyzed using a Nikon Eclipse 90i Epi-fluorescence microscope (Nikon, Tokyo, Japan). Brightfield color images were taken using a Q Imaging EXi 12-bit color camera (Q Imaging, Surrey, BC, Canada) driven by NIS-Elements AR software (Nikon Instruments Inc., Melville, NY, USA). For each stimulation subsample, three images of the exudate were taken for analysis (three replicates per sample). For histological sections, multiple tile images of gland sections taken at 10× magnification were stitched together using the ‘scan image’ function in NIS-Elements. Exudate sample images were analyzed using ImageJ software (Abramoff et al., 2004). Exudate sample images were analyzed for the number of thread skeins present (number of thread skeins/area of view), thread cell diameter along the short axis (d₁) and long axis (d₂) (µm), and mucous vesicle concentration (number of vesicles/area of view) (Fig. 1). The long- and short-axis measurements of thread skeins were used to calculate thread skein volume [4/3π(d₁/2)²×d₂/2], where each thread skein was treated as an ellipsoid. Histology samples were analyzed using NIS-Elements AR software.

All statistical analyses were run using SPSS v.23.0 (SPSS Inc., Armonk, NY, USA) with α=0.05. Non-positive values (0 or negative values) were excluded from the analyses. Outliers identified by SPSS were also excluded from analyses and figures. For trials involving successive stimulations of full glands, non-linear regression analyses were performed on the data for the percent of exudate released with each stimulation, as well as on the data for the mucous vesicle to thread skein ratio released with each stimulation. A one-way ANOVA test with post hoc least significant difference (LSD) was conducted on the thread skein morphometric (short axis d₁, long axis d₂, volume) data to test for differences in thread skein size between stimulations. Because of the nested design of the study, replicate number (three replicates per subsample) and individual number (n=20 for successive stimulation trials) were used as break variables for statistical analysis of the composition of exudate samples following successive stimulation. A two-way ANOVA with post hoc Tukey's HSD and LSD tests was conducted on gland refilling data for both species to test for the effects of days post-sliming and species on the proportion of refilling. A two-way ANOVA with post hoc Tukey's HSD and LSD tests was also conducted on the absolute slime gland refilling timeline data (total slime from full glands/number of slime glands; total slime from full glands/mass of the hagfish) between the two species. A one-way ANOVA with post hoc LSD testing was used to analyze whether the morphometrics of skeins in the collected exudate differed significantly as a function of days post-sliming. Non-linear regression analysis was performed on the mucous vesicle to thread skein ratio data as a function of days post-sliming.

The percent of total exudate released significantly decreased with successive stimulation of the full glands, with more than half of the gland contents being released during the first and second stimulation (42.1±3.2%, 23.7±1.5%, respectively) (Fig. 2A). A logarithmic curve was fitted to the data for the percent of total exudate released per stimulation (R²=0.731; F=359.419; P<0.001). Non-linear regression analysis revealed that stimulation number had a significant effect on the percent of total exudate released (d.f.=2, 132; F=429.378; P<0.0001).

Thread skein dimensions were remarkably consistent over successive stimulations of full glands (Fig. 2B). Full slime glands, on average, released larger sized thread skeins (d₁=72.5±0.4 µm, d₂=158.8±0.8 µm, volume=4.6×10³±5.6×10³ µm³) over successive stimulations. Thread skein short-axis length (d₁; d.f.=7, 125; F=0.448; P=0.870), long-axis length (d₂; d.f.=7, 125; F=0.779; P=0.606) and volume (d.f.=7, 125; F=0.318; P=0.945) were found to not significantly differ among stimulations. The mucous vesicle to thread skein ratio in the exudate decreased significantly with successive stimulation of the full Pacific hagfish slime glands (Fig. 2C). The mucous vesicle to thread skein ratio was highest in the initial stimulation of the slime gland, and consistently decreased with each successive stimulation. An inverse curve was fitted to the mucous vesicle to thread skein ratio data (R²=0.552; F=143.982; P<0.001). Non-linear regression analysis revealed that stimulation number had a significant effect on this ratio (d.f.=2, 117; F=167.716; P<0.0001).

H&E-stained histological sections from Pacific hagfish slime glands that were stimulated until exhaustion revealed that these glands were not devoid of slime cells (Fig. 3A). Full glands on average had nearly 3.5 times the cross-sectional area of exhausted glands (full glands, 7.83±1.10 mm²; exhausted glands, 2.15±0.25 mm²) and contained almost triple the number of thread cells (full glands=680±91, exhausted glands=246±31) (Fig. 3B). Compared with full glands, many of the exhausted gland sections seemed to have fewer gland thread cells in the center of the gland, with more gland thread cells found near the periphery of the gland, near the gland capsule. The circumference of exhausted glands was about half that of the corresponding full glands (C_exhausted/C_full=0.53±0.06).

We found that slime gland refilling is a process that takes multiple weeks in both Pacific and Atlantic hagfishes (Fig. 4). Refilling slime glands for both species were found to release masses of exudate equivalent to that of full slime glands by 24–28 days post-sliming, indicating that they had refilled by this time point [Tukey post hoc test; Pacific hagfish: 24, 28 and 32 days post-sliming (P=0.829, P=0.953 and P=1.000, respectively); Atlantic hagfish: 28 and 32 days post-sliming (P=0.995 and P=1.000, respectively)]. A two-way ANOVA revealed that days post-sliming had a significant effect on the proportion of refilling (d.f.=8; F=129.784; P<0.0001), but revealed no significant effect of species (d.f.=1; F=3.464; P=0.065) and no significant interaction effect of species and days post-sliming on the proportion of refilling (d.f.=8; F=1.745; P=0.092).

Exudate mass measurements from full slime glands also allowed us to compare the absolute amount of stored exudate in Pacific and Atlantic hagfishes. By dividing the total exudate mass collected from one side of a hagfish by the number of slime glands (79 for Pacific hagfish, 97 for Atlantic hagfish), we calculated the average amount of exudate obtained from each gland. For Pacific hagfish, the average was 0.0053±0.0034 g and, for Atlantic hagfish, the average was 0.0028±0.0012 g; their difference was statistically significant (Student's t-test: d.f.=198, t=6.811, P<0.001). When the same data were normalized for body size (and not the number of slime glands), the values were 8.3±4.6 and 6.5±2.0 g exudate kg⁻¹ for Pacific and Atlantic hagfish, respectively, which is also a significant difference (t-test: d.f.=198, t=3.557, P<0.001).

Pacific hagfish slime glands in the later stages of refilling (21, 28 days post-sliming) and those that were full required significantly more stimulations to exhaust them compared with slime glands in the earlier stages of refilling (7, 14 days post-sliming) (one-way ANOVA: d.f.=4; F=18.023; P<0.0001). Slime glands in the earlier stages of refilling (14, 21 days post-sliming) also released significantly smaller thread skeins on average compared with those in the later stages of refilling and those that were full (post hoc LSD test: P<0.05) (Fig. 5A). Thread skein short-axis length (d₁) (d.f.=4, 35; F=4.881; P<0.01), long-axis length (d₂) (d.f.=4, 35; F=4.536; P<0.01) and volume (d.f.=4, 34; F=4.723; P<0.01) differed significantly over the course of Pacific hagfish slime gland refilling. The mucous vesicle to thread skein ratio significantly increased with refilling of the Pacific hagfish slime gland, but was significantly reduced at 14 and 21 days post-sliming (post hoc LSD test: P<0.0001) compared with the proportion released from full glands (Fig. 5B). A quadratic curve was fitted to the mucous vesicle to thread skein ratio in refilling exudate data (R²=0.378; F=10.627; P<0.001). Non-linear regression analysis revealed that the number of days post-sliming had a significant effect on this ratio (d.f.=3, 35; F=68.803; P<0.0001).

This study provides the first ever timeline of slime gland refilling in hagfishes. The refilling process takes several weeks after the slime glands have been exhausted, with Pacific hagfish glands refilling marginally faster than Atlantic hagfish glands. It is important to keep in mind that the data shown in Fig. 4 are normalized to the mass of exudate obtained from the previously unstimulated side of the body, and therefore provide no information about the absolute rate of exudate production. Absolute measurements of exudate mass from full glands of both species reveal that Pacific hagfish glands contain almost double the mass of pre-exudate that Atlantic hagfish glands have. Thus, although the glands in the two species refill in approximately the same amount of time, the absolute rate of exudate production is about twice as fast in Pacific hagfish, a result that is consistent with the more elaborate vascular anatomy in this species (Lametschwandtner et al., 1986). This increased absolute rate of exudate production in Pacific hagfish may reflect more intense predation pressures on this species. Furthermore, the lower metabolic rate and cardiac function of the Pacific hagfish means that they expend an even larger fraction of their total energy budget to refill a given slime gland (Munz and Morris, 1965; Hansen and Sidell, 1983; Steffensen et al., 1984; Forster et al., 1991).

Although there is no striking difference in relative refilling rates between the two species, the time required for complete refilling of the glands, more than three weeks in both species, is surprising. Such a long recovery time raises the question of how hagfishes avoid being depleted of slime, which would make them vulnerable to attacks by predators. One possibility is that each slime gland contains enough exudate to participate in several defensive sliming events. Our successive stimulation data demonstrate that it is at least physiologically possible for hagfishes to release only a fraction of their slime gland contents at a time. However, it is also possible that our results are an artefact of the electrical stimulation protocol we used, and that the release of slime in the wild is an all-or-nothing response. In this case, a subsequent attack by a predator in the same area would indeed lead to a diminished sliming defense response.

It is also not clear why the slime glands take so long to refill. One factor may be the low metabolic rates of hagfishes, which are the lowest of any craniate (Forster, 1990). It is also possible that the elaborate thread skeins produced by gland thread cells simply take a long time to manufacture and are a limiting step in the refilling process. Although the mechanism of skein production in gland thread cells is unknown, the leading hypothesis suggests that the 150 mm thread elongates from a single growing end (Fudge and Schorno, 2016). If this is the case, then the rate of thread elongation must be on average about 6 mm per day, or about 60 times the length of a single gland thread cell, assuming a 24 day maturation time. Although it is known that the thread undergoes both lengthening and thickening during its development, the relative rates and timing of these processes are unknown (Winegard et al., 2014). Given that most mucus-producing epithelia, including those in the hagfish epidermis, continuously produce mucus, it is difficult to imagine that mucus production is the limiting factor in the refilling of the slime glands.

It is not clear whether ejection of exudate from the slime gland initiates refilling, or whether production of slime cells is a continuous process, similar to the production and turnover of gametes in seminiferous tubules (Clermont and Perey, 1957; Klein et al., 2010). We propose that slime cell production is initiated after the exudate is ejected, with division, differentiation, maturation and growth of slime cells continuing until the gland is full. However, full slime glands are known to contain small gland thread cells near the gland epithelium, which, at first glance, may imply a continuous production of new cells (Newby, 1946; Downing et al., 1981a, 1984). Another interpretation is that these small gland thread cells are arrested in their development and resume growing and maturing after the exudate is ejected and refilling commences. Keeping numerous small thread cells near the epithelium may reduce the time to refilling compared with a process that relies completely on the production of new cells. Further examination of slime glands at several stages of refilling, including staining for apoptotic and proliferative markers, should allow us to answer these remaining questions about the cellular mechanisms of slime gland refilling.

As seen in histological cross-sections, exhausted Pacific hagfish slime glands were, on average, about one-third the area of full glands, and exhausted glands contained fewer gland thread cells on average. However, exhausted glands were not devoid of gland thread cells and gland mucous cells, which raises the question of how and why some cells are ejected from the glands and others are retained. One way to approach this question is to consider the mechanics of the thin layer of striated muscle (i.e. the musculus decussatus) that surrounds the slime gland capsule. When these muscle fibers contract, slime cells are squeezed out through the narrow gland pore and rupture in the process, releasing their thread and mucous secretory products. Striated muscle can generally contract about 10% of its total length (Rassier et al., 2003; Peterson et al., 2004; Herzog et al., 2008). If we assume the slime gland is a sphere with radius r, circumference C, maximal cross-sectional area A and volume V, then a 10% contraction of the muscle around the gland will reduce the circumference to 0.9C, the radius to 0.9r, the area to 0.81A and the volume to 0.73V. However, histological analysis of empty glands reveals that these values are 0.53C, 0.53r, 0.28A and 0.15V, respectively, compared with those in their corresponding full glands. These numbers change the pertinent question from ‘how do the slime glands retain the cells they do?’ to ‘how do the glands eject as many cells as they do?’, given the limitations of striated muscle shortening. One possible mechanism may be that the muscle fibers in the capsule are connected in series to elastic elements that stretch as the gland is refilled, even after the muscle fibers have stretched to their limit. In this scenario, contraction of the muscle fibers causes some shortening but, more importantly, increases pressure within the gland above the threshold pressure needed for pre-exudate to begin to flow through the narrow gland pore. Once resistance at the gland pore is overcome, exudate is ejected out of the gland primarily by the relaxation of previously stretched series elastic elements. Another possibility is that contraction of the adjacent myotomal muscle assists in squeezing more exudate out of the glands than could be achieved by contraction of the musculus decussatus alone. Further work is needed to test the viability of these hypotheses.

The fact that not all mucous and thread cells are ejected from slime glands stimulated to exhaustion raises the question of which cells get ejected and which remain. Intercellular adhesion forces between cells may be relevant here, with immature cells being strongly attached to their neighbors and retained, and mature cells exhibiting lower adhesion and thus a greater likelihood of getting squeezed out the gland pore. A network of putative nurse cells (gland interstitial cells) may also be involved in the retention of immature cells in the gland (Fudge et al., 2015). Regardless of the mechanism, it seems sensible that the gland can preferentially eject mature cells and retain immature ones, as deviating from this pattern would be a waste of resources and likely would reduce the slime's predator-repelling efficacy.

Compositional analysis of Pacific hagfish slime exudate revealed that thread skein size in full slime glands was conserved from the first stimulation to the last. This is interesting, given that histological sections show small gland thread cells in full glands near the gland epithelium (Fig. 3). The discussions of muscle mechanics and cell adhesion above are the simplest explanations for these observations, i.e. small cells are not released from full glands because the musculus decussatus can only contract so much, and the smaller gland thread cells near the gland epithelium are well adhered. The result is a preferential release of mature gland thread cells closer to the center of the gland. The conservation of thread skein size over multiple stimulations of full glands is also interesting because glands that are in the process of refilling release thread skeins that are substantially smaller than those released from full glands, but release lower mucous vesicle to thread skein ratios at these time points. This is consistent with the findings of Spitzer et al. (1988), who showed that, in recently slimed glands, smaller thread skeins represent a higher percent of the total number of thread skeins present within a dissected slime gland. A proximate explanation of this pattern is that gland mucous cells regenerate and grow faster than gland thread cells in depleted glands, which makes sense given the complexity of thread production in gland thread cells. Thus, contraction of the musculus decussatus results in the ejection of the most mature thread cells, which, in the case of depleted glands, are not fully mature and are smaller than those ejected from full glands. Although ejection of smaller thread skeins likely has consequences for the function of the slime, such effects have not yet been investigated. It has been previously suggested that the threads provide a wide array of properties to the slime, including imparting cohesiveness, preventing mucin wash-out allowing for better clogging and providing anchoring points for the mucins within the slime (Fudge et al., 2005; Lim et al., 2006; Böni et al., 2016). It is possible that ejection of smaller thread skeins results in a reduced clogging ability of the slime; however, ejection of sub-optimal slime is undoubtedly preferable to releasing no slime at all during an attack.

We found a stark decrease in the mucous vesicle to thread skein ratio over successive stimulations of the full Pacific hagfish slime glands to exhaustion. This pattern may arise as a result of gland mucous cells being more abundant in areas of the gland interior that are close to the gland pore and hence squeezed out first. The functional significance of changing the mucous vesicle to thread skein ratio is not entirely clear, although Koch et al. (1991) demonstrated that manipulating this ratio in vitro using sodium citrate-stabilized slime exudate can result in predictable changes in slime cohesion. This ratio also varied as a function of the refilling time. At 14 and 21 days post-sliming, the mucous vesicle to thread skein ratio was significantly lower than that at all other time points in the refilling cycle. This pattern may simply be the result of a spike in the number of thread skeins that are mature enough to be ejected but not yet fully mature. Alternatively, slime glands may have evolved an adaptive mechanism to compensate for releasing smaller thread skeins during early refilling. Releasing smaller cells in larger quantities may make up for the shorter length of the thread skeins being released. Perhaps it is not the mucous vesicle to thread skein ratio that is important but rather the vesicle to thread length ratio. Because the relationship between thread length and skein size is not known, we are currently unable to evaluate whether this ratio is conserved.

This study provides detailed information about hagfish slime exudate and how it varies as a function of several factors, including recovery time after the slime glands are depleted, species and, in experiments where the glands were stimulated in rapid succession to release exudate, stimulation number. Our data demonstrate that slime glands from Atlantic and Pacific hagfishes take 3–4 weeks to refill completely, with Pacific hagfish achieving faster rates of absolute exudate recovery. We found that individual Pacific hagfish slime glands can release multiple boluses of exudate, with the mass of each successive bolus decreasing exponentially. In full Pacific hagfish glands, thread skein morphometrics were conserved from the first bolus of exudate expressed to the last, although the mucous vesicle to thread skein ratio declined as more exudate was expressed. We also found that exudate composition and thread cell morphometrics shift during Pacific hagfish gland refilling, with a general trend toward larger skeins being expressed as gland recovery time increases. Finally, histological analysis of full and exhausted glands from Pacific hagfish revealed that a larger volume of cells is expressed than can be explained by the contraction of striated muscle fibers in the musculus decussatus, suggesting that other mechanisms may be involved in the ejection of holocrine secretion products from the slime glands.

We would like to thank Sarah Boggett for help with slime collection for the successive stimulation trials and Helen Coates in the Ontario Veterinary College Histoprep Lab for her guidance in preparing the histological sections. Thanks also to Drs Kevin Jagnandan, Charlene McCord, and several undergraduate members of the Fudge Lab at Chapman University for providing thoughtful feedback on the manuscript. We would also like to thank Matt Cornish and Mike Davies at the Hagen Aqualab, as well as volunteers Samantha Nieuwold and Viktorya Hlamazda, for the care of the hagfish used for this study.