a putative role for natriuretic peptides in fish …osmoregulation the sequences of teleost (eel and...

Other functions of PKC in vascular smooth muscle

PKC activation and translocation may also play roles in long-term re- sponses such as gene expression and cell proliferation (Fig. 1). For instance, PKC has been shown to affect DNA synthesis by activating serum response elements associated with immediate early gene transcription (11). These effects may be related to the finding mentioned above that PKC is prominently located to a nuclear area in vascular smooth muscle.

Several lines of evidence also suggest that PKC modulates ion con- ductance by phosphorylating membrane proteins such as channels, pumps, and ion-exchange proteins, PKC has been proposed to play a role in extrusion of Ca2+ immediately after its mobilization into the cyto- sol; the Ca’+-transport ATPase is a possible target of this protein kinase. The Na+-H+ exchange protein has been reported to be another target that is activated by phorbol esters or by permeant DAG analogues, and thereby PKC may function to increase cytoplasmic pH (11) (Fig. 1).

Thus PKC appears to perform a variety of functions in vascular smooth muscle. Activated PKC may translocate to various cellular mem- branes. The catalytic domain may remain in the vicinity of the mem- branes and phosphorylate nuclear proteins or membrane pumps; it may relocate to the cell interior; or it may act through third messengers to phosphorylate cytoplasmic substrates that induce or enhance smooth muscle contraction. Differ- ent PKC isoforms may have different locations, substrates, and functions.

The authors acknowledge the secretarial assistance of Jason Kravitz and the artistic and photographic assistance of Al Brass and Rob Littlefield.

The experimental studies were supported by National Heart, Lung, and Blood Institute Grants HL-31704 and HL-42293. R. A. Khalil is a fellow of the Massachusetts Affiliate of the American Heart Association.

References 1. Adam, L. P., J. R. Haeberle, and D. R.

Hathaway. Phosphorylation of caldesmon in arterial smooth muscle. 1. Biol. Chem. 264:7698-7703,1989.

2. Brozovich, F. V., M. P. Walsh, and K. G.

3.

4.

5.

6.

7.

8.

9.

Morgan. Regulation of force in skinned, single cells of ferret aortic smooth muscle. Pflugers Arch. 416: 742-749, 1990. Griendling, K. K., S. E. Rittenhouse, T. A. Brock, L. S. Ekstein, M. A. Gimbrone, Jr., and R. W. Alexander. Sustained diacyl- glycerol formation from inositol phospho- lipids in angiotensin II-stimulated vascular smooth muscle cells. J. Biol. Chem. 261:5901-5906,1986. Hai, C. M., and R. A. Murphy. Ca2+, cross- bridge phosphorylation, and contraction. Annu. Rev. Physiol. 51: 285-298, 1989. Hidaka, H., and M. Hagiwara. Pharmacol- ogy of the isoquinoline sulfonamide protein kinase C inhibitors. Trends Pharma- cok Sci. 8: 162-164, 1987. House, C., and B. E. Kemp. Protein kinase C contains a pseudosubstrate prototope in its regulatory domain. Science Wash. DC 238:1726-1728,1987. Jaken, S. Protein kinase C and tumor pro- moters. Curr. Opin. Cell Biol. 2: 192-197, 1990. Khalil, R. A., and K. G. Morgan. Imaging of protein kinase C distribution and translocation in living vascular smooth muscle cells. Circ. Res. 69: 1626-1631,1991. Kobayashi, E., H. Nakano, M. Morimoto,

10.

11.

12.

13.

14.

15.

and T. Tamaoki. Calphostin C (UCN- 1028C), a novel microbial compound, is a highly potent and specific inhibitor of protein kinase C. Biochem. Biophys. Res. Commun. 159: 548-553, 1989. Marston, S. B. The regulation of smooth muscle contractile proteins. Prog. Bio- phys. Mol. Biol. 41: l-41, 1982. Nishizuka, Y. The family of protein kinase C for signal transduction. J. Am. Med. Assoc. 262:1826-1833,1989. Sobue, K., and J. R. Sellers. Caldesmon, a novel regulatory protein in smooth muscle and nonmuscle actomyosin systems. J. Biol.Chem. 266:12115-12118,1991. Stull, J. T., P. J. Gallagher, B. P. Herring, and K. E. Kamm. Vascular smooth muscle contractile elements: cellular regulation. Hypertension 17: 723-732, 1991. Suematsu, E., M. Resnick, and K. G. Mor- gan Change of Ca2+ requirement for my- osin phosphorylation by prostaglandin Fzn. Am. J. Physiol. 261 (Cell Physiol. 30): c253-c258,1991. Winder, S. J., and M. P. Walsh. Smooth muscle calponin. Inhibition of actomyosin MgATPase and regulation by phosphorylation. J. Biol. Chem. 265: 10148-10155, 1990.

A Putative Role for Natriuretic Peptides in Fish Osmoregulation David H. Evans and Yoshio Takei

Emerging evidence indicates that atria1 natriuretic peptide-like peptide hormones may play a significant role in various aspects of fish osmoregulation. Surprisingly, the bulk of current evidence supports a role in salt rather than volume regulation.

Introduction

In the 10 years since de Bold and his colleagues first described the natriuresis induced in the rat by injection of atria1 extracts, there has been intense interest in what is now known to be a family of natriuretic

21). H. Evans is in the Dept. of Zoology at the University of Florida, Gainesville, FL 2261 I, USA, and is Director of The Mount Desert Island Biology Laboratory, Salsbury Cove, ME 04672, USA. Y. Takei is in the Dept. of Physi- ology, Kitasato University School of Medicine, Kanagawa 228, Japan.

peptide hormones. These contain atria1 natriuretic peptide (ANP), brain natriuretic peptide (BNP) (g), C-type natriuretic peptide (CNP) (11, 12), and the recently described ventricular natriuretic peptide (14; Fig. l), named, with the exception of CNP, after the respective sites of synthesis: atrium, brain, and ventri- cle (6). Interestingly, it is now clear that the atrium is the major site of BNP synthesis, and CNP appears to be the major peptide in the brain, although a substantial amount is apparently present in the dogfish shark

0886-1714/92 $2.00 0 1992 Int. Union Physiol. Sci./Am. Physiol. Sot. Volume 7/February 1992 NIPS 15

heart, suggesting that CNP may be a circulating hormone in fish (12).

Moreover, ANP-like immunoreac- tivity has been localized in a variety of other tissues including the lung, adrenal glands, gonads, gastrointes- tinal tract, thymus, spleen, pancreas, eye, and salivary gland (16). Mem- bers of this hormone family produce relaxation of vascular smooth muscle and natriuresis by direct glomerular and tubular effects on the kidney and indirectly by inhibition of renin, aldosterone, and arginine vasopressin secretion by the kidney, adrenals, and neurohypophysis, re- spectively (e.g., Refs, 4, 6, 9).

For obvious biomedical reasons, the vast majority of the substantial literature on these potentially natu- ral antihypertensive hormones deals with mammals. However, in the past 4 years, a data base has emerged suggesting strongly that ANP, as well as other members of this family, is present in fish and may play a significant role in their osmoregulation (6) .

Mechanisms of fish osmoregulation

Extant fish can be divided into three major systematic groups: the agnatha (hagfish and lampreys), chondrichthyes (sharks, skates, and rays), and osteichthyes (bony fish, mostly teleosts). These are the only primarily aquatic vertebrates and are modern representatives of the earliest vertebrates,

With the apparent exception of the marine hagfish (which have plasma NaCl concentrations almost identical to seawater), both marine and freshwater fish have all evolved from freshwater ancestors and therefore have a plasma NaCl con- centration some one-third that of seawater but some 300 times that of freshwater. Chondrichthyes have similar plasma NaCl concentrations but maintain substantial urea and trimethylamine oxide levels in their body fluids, and therefore their plasma is slightly hypertonic to seawater.

Thus hagfish face no substantial osmoregulatory problems in seawater, marine chondrichthyes face a potential hypervolemia and hypernatremia, marine teleosts face hy- povolemia and hypernatremia, and freshwater teleosts (and the occa-

sional shark or ray that enters freshwater) face hypervolemia and hypo- natremia (e.g., Ref. 5). Thus a variety of osmoregulatory problems are pre- sented to specific fish groups, espe- cially those that are euryhaline and can tolerate a range of salinities.

Contrary to terrestrial mammals, where renal function dominates osmoregulation, fish utilize an array of tissues to maintain plasma tonicity in the face of net fluxes of water and salts (5). Briefly, chondrichthyes balance the osmotic influx of water with relatively high glomerular fil- tration rates (GFR) and urine flows. The urine is approximately isotonic, however, because the loop of Henle is not present, so diffusional gain of salt is balanced by secretion of NaCl by the unique rectal gland. Because sharks in which the rectal gland has been removed still survive in seawater, one must postulate that other extrarenal (probably gill) salt extrusion mechanisms are also present.

Marine teleosts drink seawater to counter the osmotic loss of water, transport the salt (with water follow- ing osmotically) across the intestine, and excrete excess salt (= diffusional + intestinal gain) via the gills. Te- leosts also lack a loop of Henle and therefore do not produce a urine that is hypertonic to the plasma, despite the fact that, like in the shark kidney, secretion of NaCl apparently takes place in the proximal tubule (3) .

It is of interest to note that NaCl transport across the shark rectal gland, teleost intestine, shark and teleost proximal tubule, and teleost gill are all via the Na-K-Z1 cotransport system, which is sensitive to loop diuretics (e.g., furosemide) and also is present in the thick ascending limb of the loop of Henle, as well as a variety of other cells and epithelia (10). The only variation is that the cotransporter is apical in the intestine but basolateral (along with Na- K-activated ATPase) in the shark rectal gland, shark and teleost proximal tubule, and teleost gill, all secretory rather than absorptive epithelia. Freshwater teleosts excrete large volumes of dilute urine to counter the osmotic influx of water and extract needed Na and Cl from the medium via Na-H and Cl-HCO, exchange in the gill epithelium, although there is some evidence that Na and H movements may be linked electrically rather than biochemi- cally. Thus fish utilize renal, rectal, intestinal, and branchial epithelia in osmoregulation (5).

ANP effects on fish osmoregulation

The sequences of teleost (eel and killifish) and shark (dogfish) hormones in the ANP family (Fig. 1) have only been known for 18 mo, so the majority of studies examining a putative role for these hormones in fish osmoregulation has utilized

HUMAN ANP

PIG CNP G-L-S-K G-C-F-G-L-K-L-D-R-I-G-S-M-S-G-L-G-C

CHICKEN CNP

KILLIFISH CNP G-w-N-R-G-C-F-G-L-K-L-D-R-I-G-S-M-S-G-L-G-C

EEL CNP

DOGFISH CNP

EEL VNP

IG-W-N-R-G-C-F-G-L-K-L-D-R-I-G-SI;(S-G-L-G-C

N-S-F-R---Y-COOH

N-S-F-R---Y-COOH

--G---R-R-F-COOH

N-S---R-K-COOH

N-V-L-R-R-Y-COOH

K-V-L-R-R-H-COOH

COOH

COOH

COOH

COOH

COOH

N-S-L--K-N-G-T-K-K-K-I-F-G-N-COG

FIGURE 1. Amino acid sequences of selected members of atria1 natriuretic peptide (ANP) family. BNP, brain natriuretic peptide; Hum, human; CNP, C-type natriuretic peptide; VNP, ventricular natriuretic peptide.

16 NIPS Volume 7/February 1992

either heterologous peptides or antibodies raised against mammalian ANP or BNP. Nevertheless, some interesting and unexpected patterns have emerged (6). Contrary to what might be expected from the natriuretic and diuretic action of ANP (hereafter used to designate the en- tire family) in mammals, the majority of the extant data supports the conclusion that ANP functions in seawater rather than freshwater osmoregulation in fishes but may produce branchial hemodynamic effects that would exacerbate osmoregulation in either medium. What is the basis for these conclusions?

Initial studies (see citations 25 and 72 in Ref. 6) demonstrated that injection of mammalian ANP produced natriuresis in both the freshwater trout and the marine toadfish, although in both cases quite high concentrations (-0.1 PM) were nec- essary. Importantly, the toadfish is aglomerular, directly demonstrating for the first time that ANP-induced natriuresis could be produced with- out changes in GFR. At least in the freshwater trout, the ANP-stimulated natriuresis was significantly larger than the diuresis, somewhat surprising in a fish facing hyponatre- mia, and suggesting that salt extrusion was more sensitive to ANP than water extrusion.

Natriuresis in both of these species may have been secondary to ANP stimulation of proximal NaCl secretion (see above) in a manner similar to that demonstrated for the shark rectal gland and teleost gill (see be- low), but this proposition remains unstudied. More recently, it has been shown that physiologically relevant concentrations of ANP (-130 pg/ml) actual1 y produce a fall in the GFR in the shark SquaJus acanthias, although volume loading, caused by placing this species in 90% seawater, resulted in glomerular diuresis sub- sequent to injection of the same dose of mammalian ANP (2). This suggests the interesting possibility that the renal response may be keyed to the salinity, although the shark also faces a volume load in seawater (see above). Nevertheless, it is clear that the expected correlation between volume load and diuresis is not supported by the extant data in fish, and additional studies are warranted.

The data on other osmoregulatory organs in fish are somewhat clearer, albeit limited. NaCl uptake, subse- quent to ingestion of seawater, in the intestine of the marine flounder is inhibited by mammalian ANP, but salt extrusion by the marine killifish opercular epithelium (which models the gill epithelium) is stimulated by ANP (citation 114 in Ref. 6). Salt secretion by the shark rectal gland is also stimulated by ANP apparently directly (8) as well as indirectly via the release of glandular vasoactive intestinal polypeptide (citation 118 in Ref. 6).

As indicated above, each of these tissues transports salt via the Na- K-Xl cotransporter, although the transport geometry (basolateral vs. apical placement of transporters) of the cells varies somewhat (see above), possibly accounting for the divergent effects. However, stimulation of gill or rectal gland salt secretion is of obvious osmoregulatory utility only in seawater, ANP inhibition of salt uptake in the seawater teleost intestine is somewhat more difficult to rationalize. It would cer- tainly decrease the salt loading of these hypotonic fish, but it is the only way that ingested water (criti- cal for water balance) can be moved from the lumen across the intestinal epithelium. Such an effect therefore makes ionoregulatory, but not osmoregulatory, sense.

The proposition that ANP is important in seawater osmoregulation in teleosts is supported by our dem- onstration that plasma levels [measured via radioimmunoassay (RIA) using antibodies against human ANP] decreased in two species of euryhaline marine teleosts when they were adapted to 20% seawater (citation 34 in Ref. 6). Moreover, ac-

climation of a freshwater fish to higher salinity (-35% salt water) also resulted in a significant increase in plasma immunoreactive ANP (citation 126 in Ref. 6; Table 1). Eels may be an exception to this pattern; when acclimated to seawater, their plasma levels of ANP (measured by an eel-specific RIA) fell substantially (Takei, unpublished observations).

Published and theoretical consid- erations also indicate that ANP may function in fish osmoregulation indirectly via interactions with other hormones known to be involved in salt and water balance. For instance, a very recent study (1) has found that, both in vivo and in vitro, mammalian ANP stimulated cortisol secretion by the trout interrenal gland (homologue of the mammalian adrenal) but only when the fish were acclimated to seawater. Because cortisol is known to be a major osmoregulatory hormone in seawater fish, involved in the stimulation of salt extrusion (citation 30 in Ref. 6), these data suggest that ANP also may have indirect effects on the gill transport cells.

Importantly, ANP apparently inhibits cortisol secretion in mammals (citation 47 in Ref. 6). In mammals, ANP is known also to inhibit prolactin secretion by the pituitary (4), and prolactin has been long accepted as a major effector in the osmoregulation of freshwater teleosts (citation 30 in Ref. 6). However, no data have been published relating ANP to prolactin secretion in fish.

Finally, it is clear that, in mammals, ANP inhibits the production of angiotensin via direct or indirect effects on kidney renin secretion (4). Because angiotensin stimulates drinking in fish (citation 30 in Ref. 6), potential inhibition by ANP is

TABLE 1. Effect of salinity on immunoreactive ANP in plasma of euryhaline fishes

Species High Low

Salinity Salinity Reference

Sculpin Flounder Chub Eel

102 t 8.0 9.6 k 2.1 Citation 34, Ref. 6 32 Ifi: 4.9 2.5 t 0.4 Citation 34, Ref. 6

347 k41 146k27 Citation Ref. 6 126, 343 t59 689 t 184 Takei, unpublished observations

All concentrations are means t SE in pg/ml plasma. Chub (Gila atraria) is a freshwater teleost, sculpin (Myoxocephalus octadecimpinosus) and flounder (Pseudopleuronectes ameri- canus) are euryhaline marine teleosts, and the eel (Anguilla japonica) is a euryhaline cultured freshwater teleost.

Volume 7/February 1992 NIPS 17

consistent with ANP’s inhibition of salt uptake across the intestine (see above), despite the uncertainty about the osmoregulatory utility of this effect in marine teleosts (see above). Studies investigating this potential inhibitory activity of ANP on the renin-angiotensin system in fish need to be undertaken.

ANP effects on hemodynamics

ANP has also been shown to be vasoactive in fish. Infusion of mammalian ANP produced a fall in pressure in the dorsal aorta of both the shark (2) and the eel (13) but increased the dorsal aortic pressure in the trout, apparently via release of other hormones (citation 96 in Ref. 6). Infusion of the newly described eel ANP (13) or eel CNP (15) also produced systemic hypotension in the eel, with a greater efficacy than human ANP, suggesting increased sensitivity to homologous peptides. Correlation between systemic blood pressure and salinity in fish is un- clear, but, intuitively, one might sus- pect that ANP-induced hypotension would be physiologically relevant in freshwater, rather than seawater, fish.

Using vascular rings from the ventral aorta of a teleost, shark, and hagfish, we demonstrated that rat ANP produces relaxation with a half-maximal effective concentra- tion in the nanomolar range, a sensitivity similar to that described in isolated mammalian vascular smooth muscle rings (Fig. 2). Inter- estingly, eel ANP and killifish CNP do not show any greater efficacy in relaxing ventral aortic rings from a teleost, suggesting that there may be as much variability of ANP sequences within the piscine vertebrates as within the vertebrates gen- erally. There are not enough fish sequences published yet to deter- mine whether this is true, but at least eel and killifish CNP differ by only one amino acid residue and shark CNP by four or five from the other two (Fig. 1). Finally, the sensitivity of the toadfish (teleost) aortic rings to rat ANP increased lo-fold when that species was adapted to 5% artificial seawater (citation 34 in Ref. 6), suggesting upregulation of recep- tor numbers correlated with a fall in plasma ANP in lower salinities.

•I TOADFISH:SEA WATER

A TOADFISH:LOW SALINITY

0 DOG FISH SHARK

x [fAGFISH

110

1

+ TOADFISH:EEL ANP OR KILLIFISH CNP

0 B n

A x 8

A +

?!I x +

0

0

ANP CONC., MOLAR

FIGURE 2. Effect of mammalian and fish atria1 natriuretic peptide (ANP) and C-type natriuretic peptide (CNP) on aortic vascular smooth muscle rings from spiny dogfish (Squalus acanthias), hagfish (Myxine glutinosa), and toadfish (Opsanus beta). Data are redrawn from Refs. 7 and 11, Data for effect of eel ANP and killifish CNP are overlapped because they are nearly identical.

It was also shown, using isolated perfused toadfish (teleost) heads, that rat ANP produced a net fall in the vascular resistance of this com- plex vasculature, suggesting that ANP vasodilates the branchial vasculature, which presumably pre- dominates in the perfused head, as well as the ventral aorta (citation 34 in Ref. 6). Because increased perfu- sion of the branchial vasculature would probably also be associated with an increase in the surface area for net osmotic and ionic losses, it is difficult to see the adaptive value of this response, at least with regard to osmoregulation.

Could it be that the ANP family of peptide hormones had some function in the original vertebrates, not associated with defense against osmotic and ionic problems in noniso- tonic solutions such as seawater and freshwater? The fact that rat ANP vasodilates in the ventral aorta of the hagfish (7), which does not have any substantial osmotic or ionic problems in seawater, suggests that this might be the case.

Clearly, further studies on putative roles for ANP in fish physiology are warranted. However, it is clear that these “lower” vertebrates give us an important opportunity to study the evolution of both the structure

and function of the ANP family of peptide hormones.

The restriction of editorial guidelines has limited references cited. A more complete listing of contributors to the knowledge about this topic may be found in Ref. 6.

The writing of this review, as well as our recent research, has been supported by Grant DCB 8916413 from the National Science Foun- dation (to D. H. Evans) and Grant 02640584 from the Ministry of Education, Science, and Culture of Japan (to Y. Takei).

References

1.

2,

3.

4.

5.

6.

7,

8,

9.

10.

11.

Arnold-Reed, D. E., and R. J. Balment. Atria1 natriuretic factor stimulates in- vivo and in-vitro secretion of cortisol in teleosts. J. Endocrinol. 12: R17-R20, 1991.

Benyajati, S., and S. D. Yokota. Renal effects of atria1 natriuretic peptide in a marine teleost. Am. 1. Physiol. 258 (Regu- latory Integrative Camp. Physiol. 27): R1201-R1206,1990.

Beyenbach, K. W., and M. D. Baustian. Comparative physiology of the proximal tubule. In: Structure and Function of the Kidney, edited by R. Kinne. Basal: Karger, 1989, vol. 1, p. 103-142.

Brenner, B. M., B. J. Ballermann, M. E. Gunning, and M. L. Zeidel. Diverse biological actions of atria1 natriuretic peptide. Physiol. Rev. 70: 665-699, 1990. Evans, D. H. Fish. In: Comparative Physi- ology of Osmoregulation in Animals, edited by G. M. 0. Maloiy. Orlando, FL: Academic, 1979, vol. 1, p. 305-390. Evans, D. H. An emerging role for a car- diac peptide hormone in fish osmoregulation. Annu. Rev. Physiol. 52: 43-60, 1990. Evans, D. H. Rat atriopeptin dilates vascular smooth muscle of the ventral aorta from the shark (Squalus acanthias) and the hagfish (Myxine glutinosa). 1. Exp. Biol. 157: 551-554, 1991. Karnaky, K. J., Jr., J. D. Valentich, M. G. Currie, W. F. Oehlenschlager, and M. P. Kennedy. Atriopeptin stimulates chloride secretion in cultured shark rectal gland cells. Am. 1. Physiol. 260 (Cell Physiol. 29): C1125-C1130, 1991.

Needleman, P., E. H. Blaine, J. E. Green- wald, M. L. Michener, C. B. Saper, P. T. Stockman, and H. E. Tolunay. The bio- chemical pharmacology of atria1 peptides. Annu. Rev. Pharmacol. Toxicol. 29: 23- 54, 1989. O’Grady, S. M., M. W. Musch, and M. Field. Characteristics and functions of Na- K-Cl cotransport in epithelial tissues. Am. j. Physiol. 253 (Cell Physiol. 22): Cl77- C192, 1987. Price, D. A., K. E. Doble, T. D. Lee, S. Galli, B. M. Dunn, B. Parten, and D. H. Evans. The sequencing, synthesis and biological activity of an ANP-like peptide

18 NIPS Volume 7/February 1992

isolated from the brain of the killifish, Fundulus heteroclitus. Biol. Bull. Woods Hole 178: 279-285, 1990.

12. Suzuki, R., A. Takahashi, N. Hazon, and Y. Takei. Isolation of high-molecular- weight C-type natriuretic peptide from the heart of a cartilaginous fish (European dogfish, Scy2iorhinus canicula). FEBS Lett. 282: 321-325, 1991.

13. Takei, Y., A. Takahashi, T. X. Watanabe,

14.

K. Nakajima, and S. Sakakibara. Amino acid sequence and relative biological activity of eel atria1 natriuretic peptide. Biochem. Biophys. Res. Commun. 164:

537-543, 1989.

Takei, Y., A. Takahashi, T. L. Watanabe, K. Nakajima, and S. Sakakibara. A novel natriuretic peptide isolated from eel car- diac ventricles. FEBS Lett. 282: 317-320,

1991.

15. Takei, Y., A. Takahashi, T. X. Watanabe, K. Nakajima, S. Sakakibara, T. Takao, and Y. Shimonishi. Amino acid sequence and relative biological activity of a natriuretic peptide isolated from eel brain. Biochem. Biophys. Res. Commun. 170: 883-891,

1990.

16. Vollmar, A. M. Atria1 natriuretic peptide in peripheral organs other than the heart. Klin. Wochenschr. 68: 699-708, 1990.

The Insect Goblet Cell: A Problem in Functional Cytoarchitecture David F. Moffett and Alan Koch

The midgut of some insects actively transports K+ from blood to lumen. The transporting cells extrude K+ into an apical goblet cavity, from which it diffuses into the gut lumen via a small valve. The reasons why such a complicated cytoarchitecture envelops an ion transport process are explored.

In 1961, when epithelial physiology was a relatively young science, Wil- liam Harvey, then a visiting scientist with Karl Zerahn at the Institute of Biological Chemistry of the Univer- sity of Copenhagen, made a sac prep- aration of the midgut of a larva of the silkworm moth Samia cecropia in the manner Zerahn was using to study sodium transport by toad uri- nary bladder. To their surprise, the tissue developed a transepithelial potential of X00 mV.

In a series of papers (reviewed in Ref. 8), Harvey, Zerahn, and their associates characterized this new transport system, showing that 1) the midgut secretes K+ at a very high rate (as much as 2 ~eqcm-zomin-l), 2) K+ secretion accounts for almost all of the short-circuit current, 3) K+ transport does not require Na+ or involve Na+-K+-adenosinetriphos- phatase (ATPase), and 4) the transport process is electrogenic. Inten- sive study of this transport system in several laboratories had made it one

D. F. Moffetf and A. Koch are in the Laboratory of A/lolecular Physiology, Dept. of Zoology, Washington State University, Pullman, WA

,992 64-4220, USA.

0886-1714/92 $2.00 0 1992 Int. Union Phvsiol. Sci./Am. Phvsiol. Sot.

of the best understood of inverte- brate epithelia, but recent studies have raised some perplexing new questions.

Cellular basis of active K+ secretion

The midgut of lepidopteran insect larvae (the caterpillar larvae of moths and butterflies) contains two major cell types (Fig. 1). The most numerous are columnar cells pos- sessing a tuft of apical microvilli. In addition, there are goblet cells, in which the apical membrane is inva- ginated to form an apical cavity.

The unusual structure of goblet cells suggested they might be re- sponsible for active K+ secretion (4, 8). The cavity accounts for 40-70% of the volume of the cell and contains a “matrix,” suggested by histo- chemistry to consist of polyanion (reviewed in Ref. 5). Goblet cells from the middle and anterior midgut are distinguishable from posterior midgut goblet cells in having larger cavities that extend further toward the basal pole of the cell.

The interior of the goblet cavity is lined with microvilli that project into the cavity. In the anterior and

middle regions of the midgut, each microvillus contains a mitochon- drion; this close relationship of mi- tochondria with the goblet cell apical membrane (GCAM) does not oc- cur in the posterior midgut. The most apical part of the goblet cavity forms a narrow, tortuous passage surrounded by interdigitated microvilli (Fig. 1). This structure has been termed the apical valve, since in electron micrographs individual passages may appear open or closed (reviewed in Ref. 5). Such valves are not characteristic of other secretory cells with apical crypts or cavities, such as the vertebrate goblet cell and gastric parietal cell or the chloride cell of fish gill. Furthermore, although somewhat similar cells have been reported in the gut and integ- ument of other orders of insects, none of these have apical valves.

Electrophysiology of goblet cells

Penetrations with microelectrodes of the isolated midgut under open circuit showed that the transbasal potential (Vb) is of a magnitude not unexpected for gut epithelial cells (-20 to -40 mV) (Fig. 2). It is not possible to distinguish two modes of Vb, and other evidence suggests that goblet cells are electrically coupled to surrounding columnar cells (9). The potential across the GCAM (Vam) is large (70-140 mV) (Fig. a), suggesting that this is the site of the electro- genie K+ pump. Under short circuit (Fig. 2A), the potential of the goblet cavity relative to the luminal solu- tion (VJ is positive by -50 mV; VP is reduced to a few millivolts under open circuit (Fig. 2B).

Conclusive evidence for active K+ transport across the GCAM was pro- vided by recent studies in our laboratory (9, 10). Goblet cells and goblet cavities in posterior midgut were

Volume 7/Februarv 1992 NIPS 19

a putative role for natriuretic peptides in fish …osmoregulation the sequences of teleost (eel and...

Documents