Metabolism of glucose and glycogen in Limulus polyphemus in vivo

Comp. Biochem. Physiol. Vol. 73B, No. 4, pp. 803 to 813. 1982 Printed in Great Britain.

0305-049l 82 I20803-I1503.000 Pergamon Press Ltd

METABOLISM OF GLUCOSE AND GLYCOGEN IN LIMULUS

POLYPHEMUS

IN

VIVO

MARJORIE R. NTETTEN National Institute of Arthritis, Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD 20205 and the Marine Biological Laboratory, Woods Hole, MA 02543, USA

(Reeeired 4 May 1982)

Abstract 1. As much as 70!',. of Tritium administered on C-6 of glucose is excreted as 3H20 within 3 hr, showing an extensive, rapid glycolytic degradation of glucose in Limulus. About 80",, of the ~'*C found in Limulus blood one hour after injection of glucose->*C is in the anionic fraction. 2. Preferential loss of carbon from C-1 of glucose prior to incorporation into glycogen in paired in rit:o experiments indicates a significant occurrence of the oxidative (pentose phosphate) pathway of glucose metabolism. 3. Glucose is more rapidly and extensively incorporated into glycogen of muscle than into that of hepatopancreas, indicative of the relative importance of muscle in metabolic turnover in Linmlus. Maximum incorporation into leg muscle glycogen of a large individual Limulus was seen at about 4 6 hr after a single injection of glucose-~'*C. 4. Little gluconeogenesis from lactate or pyruvate was observed. 5. The glucose of Limulus blood, occurring normally at levels of only about 3 5mg per 100ml, undergoes rapid metabolic changes and utilization.

INTRODUCTION Despite the great current interest in Limulus in studies of the electrophysiology and chemistry of vision and as a source of a blood cell coagulation factor used in detecting traces of endotoxin (Cohen, 1979)~ little is known specifically about the intermediary metabolism of this interesting, ancient species. A m o n g the few studies reported are an early extensive survey of digestive enzymes of the intestine and digestive glands (Schlottke, 1935), a study of hydrolytic enzymes found in lysosomal subfractions of the digestive gland (Brown & Whitegiver, 1971), and the finding of a vertebrate type of glucose-6-phosphatase in the hepatopancreas and coxal glands (Stetten & G o l d s m i t h , 1976). This paper reports the results of some studies of carbohydrate metabolism in Limulus, using injected glucose, pyruvate or lactate variously labelled with t'*C or Tritium [3HI.

glucose, D-[1-~4C], 0.05 mCi/1.3 mg; glucose, D-[6-~'~C], 0.05 mCi/1.6 mg; glucose, D-[U-14C], 2.2 mCi,m mole; pyruvic acid, Na salt [2-~4C], 0.05 mCi/0.7 mg: lactic acid, Na salt, L,-[U-I~'c], 0.05mCi/0.032mg: and HzO[3H], l mCi/g were obtained from New England Nuclear. Boston, Mass. Except as described in Tables 2 and 3 no carrier compounds were added.

MATERIALS AND METHODS

Animals Specimens of intermolt LimMus polyl~hemus, varying in size between 4 and 20cm across the widest part of the carapace, were obtained from the supply department of the Marine Biological Laboratory. Woods Hole, Mass. They were maintained either in running sea water at the Marine Biological Laboratory or in a solution of synthetic sea salts (Instant Ocean. Aquarium Systems, Inc., Eastlake, OH) in an aquarmm kept at 16~17 at the National Institutes of Health, and were fed a diet of squid muscle about every other day.

Isotopic compounds Radioactive samples used: glucose, D-[6AH], 1.0mCi/0.006mg; glucose, D-[5-3H], 1.0mCi/0.012mg., © 1982. US Govt, ( f~.p.73,41~ i

803

Experimental procedm'es To remove possible contaminating 3HzO or 3H-containing ionic decomposition products from the commercial glucose-[6AH] a solution of the glucosc was passed through a microcolumn of mixed be ion exchange resin Amberlite MB-3 (Bio Rad Laboratories, Richmond, CA) and the eluate Iyophilized and redissolved in water. Radioactive compounds, dissolved in a small volume of isotonic salt water, were injected at time zero into the heart at the central point of the juncture between the cephalothorax and abdomen. When needed, samples of blood were obtained from the same point. Details of each experiment are given in the captions to the tables and figures. Animals were killed by exsanguination and dissection and muscle and hepatopancrcas glycogen isolated. Glycogen was isolated by ethanol precipitation following digestion of hepatopancreas and muscle tissue in hot 30!~; KOH, and purified by reprecipitation after resolution twice in water, once in cold 5'~,, trichloroacetic acid, and finally again several times from water, removing any water insoluble material by centrifugation at each purification step (Stetten & Stetten, 1951). Glycogen was usually determined by isolation, giving minimum quantitative wdues because of unavoidable losses occurring during the extensive isolation and purification steps. In some cases glycogen was determined by glucose analysis of acid hydrolysates of glycogen precipitated from KOH digests of muscle or hepatopancreatic tissue. Glucose was determined colorimetrically by oxidation with specific glucose oxidase (Bergmeyer & Bernt, 1974). Isolation of glucose and of cationic and anionic compounds derived in civo from the injected radioactive glucose was carried out on aliquots of blood and hepatopan-

M ARJORIE R. STt TTEN

804

creas homogenates by a modification of the method of Katz et al. (1975l. Protein was precipitated with perchloric acid and, after removal of perchloric acid by neutralization with KOH, an aliquot of supernatant solution was passed through 3 small ion exchange columns arranged in tandem. Column 1 contained AG50W-XS-H * form to retain cations, column 2 AGI-X8- acetate form to retain anions, and column 3 AG1-X8-borate form to adsorb glucose. Glycogen and 3H20 formed during metabolism were contained in the solution and wash water which passed through all 3 columns. Glucose was recovered by elution of column 3 with 0.5 N acetic acid or 6 N formic acid. Combined acidic metabolites were obtained by elution of column 2 with 6 N formic acid. Less than 0.5", of the radioactive material put on the columns usually remained uneluted on the columns. Radioactivity was determined by liquid scintillation counting using Hydrofluor solution and a Beckman LS-233 or LS-9000 counter.

RESULTS

Glucose of blood and ,qlyeoqen qf hepatopancreas and nll~,St'le Prior to the isotope experiments, analyses were done on fed and fasted Limuli to establish the range of normal levels of glucose in the blood and of glycogen in muscle and hepatopancreas (Table 1). The glucose level in Limulus blood, determined on freshly drawn samples was found to be low, usually between 3 and 5 mg2{;, and does not appear to vary greatly with the state of nutrition. Blood glucose levels between 1 and 95 rag% have been reported for a variety of Crustaceans (Hohnke & Scheer, 1970}. On standing, incubation, or freezing and thawing, blood of Limulus may show a 50 1003~; increase in glucose concentration, presumably due to hydrolysis of glucose-containing oligosaccharides or polysaccharides. All of the Limuli used in the present studies were, so far

as could be determined, in an intermolt stage. No attempt was made to make the stage of the molt cycle the variable. In several Crustacean species where this was done the concentration of blood glucose has been found to vary with the stage of the molt cycle (Hohnke & Scheer, 1970). Both muscle and hepatopancreas had relatively high concentrations of glycogen, the level varying with the nutritional state. Hepatopancreas and muscle of non-fed animals from the M B L stock, whose nutritional state was uncertain, contained less than 17,, of glycogen (1 and 2, Table 1), but glycogen levels of 3 4°o of the wet weight of hepatopancreas were found in animals which had been well fed for weeks. In paired animals (3 6. Table 1) it was found that the level of b o t h hepatopancreas and muscle glycogen could be greatly reduced by prolonged fasting. Short. 6-day, fasting of well fed animals (7 and 8, Table l) resulted in rapid depletion of the hepatopancreas glycogen while the muscle glycogen level was unaffected. In this regard changes in the glycogen of Limulus do not differ qualitatively from those of the rat in which fasting does not deplete muscle of glycogen to the same extent as the liver (Cori & Cori, 1926). The relatively large and variable content of glycogen of Limulus hepatopancreas and muscle was amply confirmed in the quantities of glycogen isolated in the subsequent isotope experiments.

Rc~te ofincorporation (?],qluco,~e into le(t muscle A large Limulus, measuring about 20 cm across the widest part of the carapace was obtained from the stock supply of the Marine Biological Laboratory and maintained unfed in a tank of fresh flowing sea water. A solution of glucose [U-b*C] was injected into the heart at time zero. After each successive time interval between 1 and 48 hr one of the six identical appearing walking legs was removed for glycogen isolation. The

Table I, Blood glucose and glycogen content of hepatopancreas and muscle of Limulus Limulus Size cm

Glycogen

Weight

Condition*

mg%

1 2

25.4 16.5

M.B.L. M.B.L.

3 4

12.0 11.4

5 6

16.5 15.9

7 8

i0.0 9.0

134 I00

8.9

68 + 7

9-11

Blood Glucose

g

stock, not fed stock, not fed

Hepatopancreas

Muscle

% of wet wt

4.6 4.7

0,39 0.22

0.38 0.65

Fed 7 weeks Fasted 7 weeks

4.3 3.2

3.02 0.42

---

Fed 8 weeks Fasted 8 weeks

3.7 4.6

4.12 0.19

2.38 0.23

4.0 3.9

2.68 0.29

1.08 1.29

---

3.3 + 1.3

1.8 + U,3

Fed 9 months Fed 9 months, fasted 6 days Fed 9 months

then

* 1 6 were kept in running sea water at the Marine Biological Laboratory. 7 I 1 were maintained in Instant Ocean at about 16'C at the National Institutes of Health. i For Limuli I 8 portions of the hepatopancreas and of the body muscle were used for glycogen analysis and t\~r Limuli 9-11 glycogen was isolated and purified as described in the experimental section. Glucose was determined by the glucose oxidase method.

Metabolism of glucose and glycogen in Limulus l

glycogen was replaced by non isotopic precursors. About half of the maximum glucose-l~C incorporated remained in the glycogen after 48 hr, The progressive decrease in the quantity of glycogen isolated from each leg for the first day indicates that during this period of fasting and of increased muscular exertion degradation of muscle glycogen exceeded synthesis.

12

10

8

Comparison of incorporation of blood ,qlucose into muscle and hepatopancreas glycogen

6

'5 4

oc

805

2-~

Y .....

1JO

i

t

J

i

20

30

40

50

Time {hrs)

Fig. 1. Radioactivity in glycogen isolated from leg muscle after a single injection of glucose-[U-~4C] into Limulus heart. A solution of 126 mg of [~4C]glucose containing 0.05/LCi/mg in 0,7 ml was injected into the heart at time zero. The quantity of purified glycogen isolated from each leg was 30.2rag at 1 hr, 29.4mg at 2hr, 15.6rag. at 4hr, 13.3 mg at 12 hr, 7.8 mg at 24hr and 14.7 mg at 48 h. after glucose injection. Each point represents the muscle glycogen of an individual walking leg.

first and last specially adapted pairs of legs were not used. The animal became excessively agitated for a short time and considerable bleeding occurred after each operation but, due to the very efficient blood coagulation mechanism of Limulus, the bleeding soon stopped and the animal appeared normal in the salt water tank. Each leg was immediately put into a small amount of boiling water to coagulate the protein and glycogen was isolated from the combined coagulate and extract as described in the experimental section. The results are given in Fig. 1. Significant glucose had been incorporated into leg muscle glycogen in l hr, the shortest time interval studied, and the amount of radioactivity in the glycogen increased rapidly to a maximum at about 4~6 hr. After this time the concentration of 14C in glycogen decreased slowly as muscle

To get a preliminary indication of the rate of incorporation of glucose into hepatopancreas and muscle glycogen, identical amounts of a solution of glucose-l'~C were injected into the hearts of three small fed Limuli which had been matched as nearly as possible to be equivalent in size, nutrition and stage of life cycle. Carrier glucose was injected with the isotopic glucose to favor deposition rather than degradation of glycogen depots during the short time of the experiment. The specific activities of the glycogen samples isolated from the hepatopancreas and pooled muscles of each animal sacrificed 1. 2 and 3 hr after a single injection of glucose are given in Table 2. Glucose was rapidly incorporated into muscle glycogen in these fed small animals, with the specific activity probably not yet at a maximum by 3 hr as had been observed with the muscle glycogen of the large fasted Limulus (Fig. 13. Hepatopancreas glycogen had incorporated less glucose than had muscle glycogen at each time interval in this experiment as in all subsequent experiments.

Incorporation of specifically labeled ,qlucose into ylyco,qen in paired experiments In an attempt to evaluate the relative contributions of glycolysis and the oxidative (pentose phosphate) pathway to the metabolism of glucose in Limulus, glucose labeled uniformly with '4C or specifically with '4C in the C-! or C-6 position was injected (Table 3). Within each of the 3 groups of animals studied the conditions of the experiment, nutritional state and body weight were maintained as nearly as possible the same so that 1 would be comparable with 2, 3 with 4, and 5 with 6 and 7, the only difference being in the location of the >*C-label in the glucose

Table 2. Rate of incorporation of glucose-[~4C] into glycogen in Limulus in t:ivo

Glycogen isolated Limulus

We ight g

Time hr

Hepatopancreas mg cpm/mg

mg

Muscle cpm/mg

1

6.3

1

1.2

740

34

i 900

2

6.8

2

0.3

1,190

1.7

5 250

3

7.5

3

6.7

i, 270

9.7

15,600

Fed Limuli, maintained for several months in an Instant Ocean bath at 16-17:'C, were injected with a single dose of a solution of glucose-[-U-~4C] into the heart. All radioactive values have been normalized for an equivalent of 150,000 cpm in the glucose injected per g of Limulus. About 0.74 mg of carrier glucose was injected per g Limulus I and 2.

MM'..JORIt R. St't] r t N

g06

Table 3. Incorporation of specifically labeled glucose -~4(" into kimulus glycogen kimulus

(:lycugcn

(;lucost, Time

Weight

Nutri t ion

I~.7

fasted 4 day's

2.

7.5

fasted

4 days

Muse t ,

tlepatopancrcas

1eg

body

i n j ected~'¢

g

I.

isolatud

hc

mg

c pro/rag

mg

cpm/mg

i

2~J

20

i0

%()9

1

7

15

5

<)9

;

3 3

7 47

i 270 24

I() 36

13,61)0 1 400

14 glucose-[U-14 C] glucose-[l-iAC] glucose- [6-~-C]

i I 1

92 185 108

24 4 12

17 213 13

58 43 70

glucose-[U-14C] ~,ith carr{er 14( :] g' l u c o s e - [ ] -

mg

cpm/mg

with c a r r i e r 3. 4. 3.

(~. 7.

7. 5

fed

14C glucose- U -14

11.7

fed

glucose'- I C -

70 h4 64

fud

fed fud

32 33 24

239 177 i 240

* All radioactivity values have been normalized to an equiwflent of 150,000 cpm m the glucose rejected per g Limulus. The 0.74 mg of carrier glucose given per g of Limulus in Experiment 1 and 2 is estimated to have ele,,ated the glucose concemration of the blood more than 20 fold v,hile the trace doses given to Limuli 3 7 did not greatly change the 1o~ glucose levels of the blood. All animals were fed squid muscle 2 hr before injection of the isotopic glucose, For isolation of muscle glycogen, the leg and body muscles of small animals (1 4} were pooled while those of large animals ~;ere isolated separately.

injected. With the first pair of small Limuli, which had been fasted for four days and injected with carrier as well as radioactive glucose, the specific activity of the isolated glycogens of both hepatopancreas and muscle was higher from [U-~4C]-than from [1-~4C]glucose. Similar results were obtained with the second pair of Limuli which were well fed and received only trace doses of radioactive glucose. Much higher levels of radioactivity were observed in the glycogen from the 3 hr than from 1 hr experiment and in the muscle than in the hepatopancreas glycogen samples, In experiments 5, 6 and 7, with somewhat larger fed animals killed one hour after receiving radioactive glucose, the quantities of glycogen isolated from individuals were more nearly the same than was the case with the small Limuli. Here, as in the previous groups, less radioactivity was incorporated into glycogen from the glucose labeled only on C-1 than from that labeled on C-6 or on all 6 C atoms. Quantitative estimation of the relative contributions of the glycolytic and the oxidative pathways cannot be made from such in vit:o experiments. Insofar as the paired animals may be assumed to be in similar steady states of deposition and utilization of glycogen these results are compatible with the assumption that an appreciable preferential loss of carbon from C-1 occurs via the operation of the oxidative pathway of glucose metabolism before utilization of glucose for glycogen synthesis.

Estimation of ocerall degradation qt,qluco.se in vivo in Limulus using glucose doubly labeled with 3H and >*C To try to estimate the extent of recycling via degradation products compared with direct incorporation of blood glucose into glycogen, glucose labeled with Tritium at 0 6 together with variously labeled glucose-~4C was injected in trace amounts into various sized fed Limuli. The amount of 3H- and 14C-excreted

into the bath was monitored and glycogen was isolated from the hepatopancreas and muscles at the end of the 3hr experiments. Glucose-[6-3H] was used because hydrogen at C-6 has been shown to be more stable metabolically than that at any other position of the glucose molecule (Clark el al, 1974) aim release of 3H20 from glucose-[6-3H] has been used to estmmte over all glycolytie flux (Katz & Rognstad, 1975; Hue. 1976). Reincorporation of 3H20 into glucose is neglible, it was shown that all of the tritium in the bath water was in ~H20 and all of the radioactive carbon in ~4CO2. No radioactive glucose or organic degradation products could be detected in the bath when glucose was given in trace amounts. Even when the level of blood glucose was increascd about 100 fold by glucose injection no unchanged glucose was excreted, but under these conditions about 1"0 of the dose was excreted as anions, presumably lactate. The quantities of 3H- and t4C-excreted into the bath water, expressed as percent of the administered dose. are plotted as a function of time in Fig. 2. -~H3) derived from ~H on C-6 of glucose appeared surprisingly rapidly in the bath water. Normally active animals metabolized and excreted about 30", of the total 3H in l hr and between 40 and 70". m 3 hr. One animal, number 4, Fig. 2, which was inert for about 2 hr after being injected, showed a correspondingly slow initial rate of 3H20 excretion. Considerably less ~*CO2 than 3H20 was excreted, but the somewhat higher levels of 14CO2 derived from glucose-[1-1aC] than from glucose-[U-I~C] is in accord with the conclusions drawn from the data of Table 3 about the probable operation of the oxidative pathway of glucose metabolism. In Experiment 5 essentially no 14CO 2 from pyruvate [2-14C] was excreted, while glucose degradation, as monitored by the excretion of aH20 from glucose-[6-3H], was proceeding normally. A large and variable amount of ~4CO2 and 3H20 is

Metabolism of glucose and glycogen in Limulus I

I

I

I

E~p

1

70

6O

3

.....

i

I6 - 3HI gl . . . . .

807

from carbon-6 of glucose than is indicated by the aH in the excreted water, and the 44~70% values shown in Fig. 2 are minimal estimates of the glucose degraded.

Excretion of 3H20 after injection o[ glucose-[5-3H], giucose-[6-3H] and 3H20 4O .~.

20 14

,0

C trom

//// _--'5~_-~-*'~:.-~-_-..oa Y/.=' ~ - - ' ~ ' ~ . ~ -" ~10'ooe'...,

~'~-~------,

---~--~_-L_

1 Time

lhrs~

2

~

3

Fig. 2. Excretion of [3H]H:O and [>*C]CO 2 by' Limuli after injection of a single dose of radioactive compounds into the heart. Limuli 1 {O) and 2 (×) were given [1-1"~C, 6-3H]glucose, Limuli 3 [O) and 4 (B) received [UJ'*C, 6-3H]glucose and Limulus 5 (A) received [6-3H]glucose and [2J~C]pyruvate. Each animal was kept in a known volume of "lnstam Ocean" water (pH = 8.5, 200ml for 1, 3. 4, 400ml for 2, and 1000 ml for Limulus 5) and aliquots were taken every half hour for analysis. Limulus 4 was very inactive for the first 2 hr while all others were normally active. For comparison between the experiments which were done at different times and with animals of varied size, all isotope values were normalized to what would have been obtained had each been given 100 × 103 cpm/g animal of each isotopic compound. Thus, as plotted, the ordinate values give directly the percent of injected dose excreted. Trace quantities of isotopic glucose were given with no carrier added. Actual amounts given and the weight of the animal in each experiment were: To Limulus 1 (9.5 g). 1030 × 103 cpm in [114C]glucose and 2335 x 103cpm in [6-3H]glucose. To Limulus 2 (47g1 6390 x 103 in [1-1~C]glucose and 22,490 × 103 cpm in [6JH]glucose. To Limulus 3 (26g) 5090 x 10~cpm in [U-X4C]glucose and 15,214 × 103cpm in [6-3H]glucose. To Limulus4 [14.6g)3570 × 103cpm in [U-14C]glucose and 10,780 × 103cpm in [6-3H]glucose. To Limulus 5 (124g), 13,825 × 103cpm in [2-14C pyruvate] and 76,750 × 103 in [6-3H]glucose. The temperature of the bath was maintained at 16 17C. Solid lines indicate 3H-excreted as H20 and broken lines ~aC-excreted as CO2, retained in the blood. The rate of removal of these end products of metabolism presumably depends upon the rate of respiration, as roughly indicated by the beating movement of the gill lamellae which causes water to circulate over the lamellae and pumps blood in and out of the gills (Barnes, 1980). For a rough estimate of the amount of isotope retained in the CO2 and H2O of the blood 3 hr after injection of radioactive glucose in two of the experiments, aliquots of blood supernatant were acidified with HC1, evaporated to dryness and recounted. About 20% of the 14C and 620o of the 3H of the blood in Experiment 2 and 25"~i of the ~4C and 71°, of the 3H in Experiment 3 (Fig. 2} were eliminated by this procedure. Thus an even larger percentage of glucose has been metabolized in such a way as to lose stable H

To aid in the interpretation of the observed rapid metabolism and excretion of hydrogen on C-6 of glucose, several control experiments seemed necessary. The possibility was entertained that H on C-6 might be rapidly and selectively lost in the course of a conversion of glucose to glucuronic acid in Limulus. Since the rate of appearance of excreted 3H20 into the bath water from glucose-[3H] is a resultant of the rate at which glucose is metabolically degraded and the rate at which body water is exchanged with the environment, it was of interest to have a measure of the normal rate of excretion of 3 H : O per se after injection into Limulus. Glucose-[5-3H] was chosen to compare with glucose-[6-3H] because it would be expected to produce 3H20 more rapidly in the course of normal operation of glucose degradative and synthetic pathways (Hue, 1981). The rates of excretion of 3H20 after injection of 3H20 or specifically labeled glucose into two fed Limuli are plotted on Fig. 3. 3H20 injected into the heart of a 43 g Limulus (Fig. 3B) was so rapidly excreted and exchanged with 300 ml of bath water that a state of equilibrium was reached by about 40 rain. When the animal was transferred to a fresh bath water every 45 rain essentially all of the body 3H20 had been excreted in about 3 hr. From the initial rate of excretion following each change of bath water a first order rate constant of about 1.0 rain-1 and a to.5 = 38 rain were calculated. This is, of course, a minimum rate of exchange of Limulus body water because of the limitations of the methodology used. 3H20 would have been lost more rapidly to an infinitely large surrounding pool of water. From a comparison of the rate of excretion of 3H20 derived from glucose labeled with tritium with that of 3H20 injected as such, it is clear that the rate limiting step is the metabolism of glucose. In each of the three paired glucose experiments plotted on Fig. 3 tritium is metabolically much more rapidly lost from C-5 than fiom C-6 of glucose. In the duplicate experiments with Limulus-1 given glucose-[S-3H] (Fig. 3A) the rate and extent of catabolism of glucose were nearly identical. In the duplicate experiments with Limuhls-2 given glucose-[6-3H], while both initial rates of catabolism were rapid, the extent of overall degradation of glucose was much lower in one of the experiments. This is a good example of the difficulty in establishing and maintaining comparable conditions of glucose metabolism and glycogen deposition even when identical external conditions are kept constant. In this case the excretion of 3lt~20 was followed for 48 hr while the animal was not fed (Fig. 3A) and the results indicate the gradual mobilization and degradation of glycogen depots which had been formed from glucose-14C early in the experiment. When the experiment was repeated with the Limuli reversed (Fig. 3B) the results were the same, i.e. much more rapid elimination of the 31t from C-5 than from C-6 of glucose. Selective rapid oxidation at C-6 to

808

MARJ()Rlli

100

80

[ A

I

I

I

I

'

R. STI TTEN

"1 100

Glucose_[5_3H] ~ "~. Limu,us 1, . ~ J b ~ - ' ~ ' ~ ~- - ~ ,~" ~ GI ...... [6-'H] ? 80

40 20

2 t

2

L

4

,

I

6

,

L

8

,

J

0

48 1 2 TIME-HOURS

~

4

6

8

Fig. 3. Excretion of 3H20 by Limuli after injection of glucose-[5-3H], glucose-[6-3H] and 3H20. Tracer doses of radioactive compounds containing 40/~Ci in 0.2 ml of isotonic saline were injected at time zero into the hearts of two fed intermolt animals, Limulus-I weighing 36.5 g and kimulus 2, 43 g. Each animal was maintained in a bath of 300 ml of Instant Ocean and aliquots of bath water were taken for analysis at timed intervals. The Limuli were transferred to a fresh bath at 3 hr intervals when glucose-[3H] was given, and at 45 min interwds in the 3H20 experiment. Compounds injected-A-glucose-[5-3H] ; Q-glucose-[6-3H]: O-3H20. (A) Limulus I received glucose-[5-3H] and Limulus 2 glucose-[6AH] in two paired experiments. The data indicated by the broken lines ( ) were obtained at the same time using the same pair of animals and radioactive solutions, after restoration of background radioactivity one week after the lirst experiment ( ). (B) the Limuli were reversed after restoration of background activity. Limulus 1 received glucose-[6--~H] and Limulus 2 glucose-[5-3H]. The ~H20 experiment was done with Limulus 2.

form glucuronic acid is therefore eliminated. 3H on C-5 is readily exchanged with water during glucose metabolism chiefly by the action of transaldolase and triose phosphate isomerase (Katz & Rognstad, 1975) in the course of operation of the glycolytic and pentose phosphate pathways. 3H is lost from C-6 of glucose chiefly at the later stages of metabolism of lactate and pyruvate and in complete oxidation to CO2 and H20. Thus blood glucose of Limulus is rapidly and extensively metabolized to three carbon intermediates which are further totally degraded.

Hepatopanereas and mu.wle glycogen ji'om douhh' labeled ~dlucose Although most of the blood glucose is rapidly broken down and utilized as an energy source, a significant and variable amount is incorporated directly into glycogen. Table 4 gives the results obtained from glycogen isolated from the hepatopancreas and muscles of the .'fed Limuli whose excretory rates are shown on Fig. 2. From the quantity and radioactivity of the glycogens isolated, rough estimates could be made that 3hours after injection of glucoseJ'~C between 0.1 and 0.6°4, of the injected dose had been incorporated into hepatopancreas glycogen of the various animals. If it is assumed that muscle comprises about 10",; of the total weight of the Limuli, between 1 and 5°,0 of the injected dose of glucose had been incorporated into muscle glycogen in the same time. During the same time interval a minimum of 70°; of the glucose had been used as an energy source and totally broken down to CO2 and H20. All the values in Table 4 have been normalized for a ratio of 3H/~'*C = 1 in the glucose injected. Therefore, insofar as glucose is incorporated into glycogen without metabolic degradation and resynthesis the ratio of 3H to 14"Cwould remain 1. The extent of the

decrease in 3H/14"C ratio in the isolated glycogen may be considered a rough indicator of the amount of metabolic change which the glucose had undergone before being incorporated into tissue glycogen. The specific activity of the glycogen derived from radioactive glucose is enormously variable and depends not only on the rate of glycogen synthesis but also upon such relatively uncontrollable factors as the quantity of preexisting glycogen in each tissue, the state of nutrition of the animals and the extent of their physical activity during the experiment all contributing to either an overall net synthesis or net utilization of molecules from the reservoirs of glycogen. The same amount of radioactive glucose incorporated into a small reservoir will yield isolated glycogen of higher specific activity than if incorporated into a large reservoir. The 3H/14c ratios in the two cases would be identical however. In the short term experiments reported in Table 4, in which glycogen was isolated 3 hr after injection of glucose labeled with 3H and ~'*C the quantity of glycogen and the absolute amounts of 3H and ~4C in the glycogens vary greatly, but for each animal the 3H/14c ratio is higher in muscle than in hepatopancreas glycogen. This may be taken as an indication that the pathway for incorporation of glucose into glycogen is more direct in muscle than in hepatopancreas. In some cases glycogen was isolated separately from the leg muscles and from the other body muscles while in others total combined muscle glycogen was isolated. Incorporation of radioactivity from glucose varied in the different samples but in all cases overall synthesis of glycogen was much greater in muscle than in hepatopancreas, in agreement with earlier results (Tables 2 and 3). In contrast with the results of Table 3, where strictly parallel experiments were run at about the same time with all conditions kept as

47

26

14.6

124

90

372

2.

3.

4.

5.

3

6.

7.

3

3

3

glucose[U-14C,63H]

lactate[U-14Cl

days

6

3

pyruvate[2-~14 C] 3 glucose-[6-~H]

[U-14C,63H]

glucose-

[U-14C,63H]

glucose-

glucose[I-14C,6-3H]

glucose[1-14C,6-3H]

0.2

3.6

0.i

0.8

i.i

2.8

0.i

480

--

24

41

56

106

1830

148

1

ii

52

85

155

3190

14 C

tt/14C

--

2.2

0.79

0.68

0.68

0.57

3

1.9

0.4

---

1.0

1.7

---

7,' of w¢~t wt

--

950

1210

1230

1250

5580

3H

5

22

1280

1480

1500

9330

14 c

;

174

1050

2730

3H

Leg muscle cpm/mg

. . . .

43

0.95

0.83

0.83

0.60

3II/14(.

Body muscle glycogen cpm/mg

3

8

1420

3340

--

22

0.74

0.82

14 C 3ti/14(

glycogen

* All radioactivity values have been normalized for a ratio of 3H/14c = I in the compounds given and an equivalent of 100,000cpm injected per g Limulus. N o carrier glucose was added. All animals were fed squid muscles before but not during the experiments. For Limuli l and 4 leg and body muscles were combined for glycogen isolation. Limuli 1 5 are the same animals for which 3H and ~4C excretion are plotted on Fig. 2.

9.5

i.

wet wt

3H

cpm/mg

Hepatopancreas glycogen

g

hr

Time

% of

Injected*

Compound

wt

Limulus

Table 4. Glycogen synthesized in t:iro from doubly labeled glucose

t"

O

o

o

MARJORIE R. STI!I-TtN

8 [0

nearly as possible constant, conclusions cannot be drawn here (Table 4) comparing the results of specifically labeled glucose-14C. Variations due to different metabolic conditions made the location of the ~'*C-label irrelevant. The Limuli used in the experiments recorded in Table 4 varied greatly in size, and very, different amounts of isotopes were administered so, for rough comparisn, the absolute isotope concentrations of the samples isolated have been normalized to 100.000cpm injected per g of Limulus. Such a normalization is not optimal because metabolism is no doubt related to the relative size of the metabolizing organ which is not strictly proportional to the weight of the animal. Thus the amounts of isotope incorporated are not as comparable between animals as were those in experiments using animals of about the same size {Tables 2 and 3). In one experiment (Table 4, Limulus 7) glucose-[U-~'~C. 6 P H ] was injected into the heart of a very large fed Limulus and glycogen samples were isolated after 6 days of fasting. The quantity of glycogen which could be isolated from the hepatopancreas was very low. about ().2°~i of wet weight, while that of muscle was not significantly reduced (1.2°,,), in agreement with the results shown in Table 1, Experiment 8. The specific activity of the small amount of residual glycogen of the hepatopancreas was high and had a -~H/14 c ratio about 1/3 as high as in the injected compound, indicating that a portion of the glycogen initially synthesized and deposited was retained during the 6 day fasting period. The muscle glycogen had a somewhat higher specific activity and a much larger total quantity of the injected glucose was retained in the muscle than in the hepatopancreas glycogen. As was the case with the short time experiments, 3H/14c

ratios were higher for muscle, indicative of less extensive metabolic changes before deposition.

Ghwose and anionic dericotices.from inlected 1*C-compounds The distribution of radioactive compounds in blood drawn at time intervals after injection of ~4C-labeled substrates was studied by separation of glucose, anionic, cationic and unabsorbed compounds using three columns in tandem. Similar studies were done with supernatant solutions of hepatopancrcas homogenates in several of the experinaents. Preliminary trials established that glucose could bc absorbed and eluted from the borate column and pyruvate fl-om the acetate anion exchange column and that glycogen passed unabsorbed through all three columns (data not shown). Results, shown in Table 5, arc expressed as percent of the total radioactivity of deproteinized tissues which was recovered in each fraction. Even in the shortest time interval studied, one half hour. the bulk of the glucose-~4C had been metabolized to anionic compounds. Only between 10 and 26"}, of the radioactivity was found in glucose while between 86 and 62~Ii, appeared as anions in the same animals. With the passage of time the rehttive proportions mcreased in anions and decreased in the blood glucose, only about 2~',, remaining in the blood ghlcose aflcr 3 hr. These results are in good general agreement with the conclusion of rapid glucose metabolism drawn from the results given in Figs 2 and 3. Hepalopancreas supernatants always showed somewhat more isotope in the glucose and less in the anionic compounds than was found in the blood at the same time interval. While no cations derived from the glucose were detected in any of the blood supernatants, up to 5?, of the radioactivity in the hepato-

Table 5. Distribution of ~4C in proteinq'rce supernatants of kimulus blood and hepatopancreas* Radioactive compound injected

Sample analyzed

Limulus i. Blood-½ hr Blood-i hr Hepatopancreas-I Limulus 2.

%

(;lucose (eluate, Col. 3) %

Anions (eluate, Col. 2)

Cat ions

(retained,

Col.

1)

glucose- [i-14C]

hr

6.4 2.2 5.1

26.2 6.4 16.0

62.3 9'5.4 63.2

<

(J.I O.i 4.9

12.6

18.1

67.6

<

0.1

8.8 7.7

11.1 17.6

76.2 56.8

<

II,l 1.5

4.0 0,9

i0.5 2.1

86.5 87.9

< <

0.I 0.1

0

O

95.0

0.I

0.i

0

77.0

2.6

glucose-16-14C]

Blood-½ hr Blood-i hr Hepatopancreas-I

Not absorbed

hr

Limulus 3. Blood-½ hr Blood-3 hrs

glucose-[U-14C]

Limulus 4. Blood-3 hrs Hepatopancreas-i

lactate-[U-14C] hr

* Blood samples, drawn at time intervals after injection of the radioactive compound, the homogenatcs of hepatopancreas taken at the end of the experiment were deproteinized with perchloric acid. Portions of the protein-free supernatants were passed through 3 tandem columns to remove cations, anions and glucose respectively. (See Experimental). Results are expressed as percent of the total put on the columns that was recovered in each fraction.

Metabolism of glucose and glycogen in Limulus pancreas supernatants appeared in positively charged organic derivatives. Glycogen could readily be isolated from the aqueous solution which had passed through all three columns. Glycogen isolated from Limuli 1 and 2 one hour after receiving glucose [1-14C] and glucose-[6-14C] respectively contained 23 and 21 cpm/mg. Thus the location of the ~4C-label is not significant in these short experiments in which glucose is largely degraded than deposited into glycogen. Gluconeoyenesis fi'om lactate and pyrm:ate To get an indication of the extent of the occurrence of gluconeogenesis, trace amounts of labeled pyruvate and lactate were administered to Limuli in experiments parallel to those done with radioactive glucose. Pyruvate-[2-14C] along with glucose-[6-3H] was injected in Experiment 5 shown in Table 4 and Fig. 2. Glycogen isolated from muscle after 3 hr had incorporated very little glucose derived from the pyruvate- ~4C while appreciable synthesis of glycogen from glucose was occurring as indicated by the relatively large incorporation of glucose-[6-3H]. When a trace amount of e-lactate-[U-~4C] was given in a similar experiment insignificant amounts of 14C were incorporated into hepatopancreas or muscle glycogen in 3 hr (Table 4, Experiment 6). Fractionation of the blood of this Limulus (Table 5, Limulus 4) showed that essentially all of the ~4C from lactate remained in the anionic portion 3 hr after injection. No detectable radioactivity was found in the blood glucose. These results indicate that gluconeogenesis from the glucose metabolic intermediates pyruvate and lactate is minimal and that the so-called Cori cycle (muscle glycogen ~ blood lactate---, liver glycogen ~ blood sugar) is probably not of importance in Limulus.

DISCUSSION

In an evaluation of the relative quantitative importance of hepatopancreas and muscle in the storage of glycogen and its utilization for energy production, the relative size of the organs must be taken into consideration. The hepatopancreas of Limulus is a large, moist, friable organ which occupies much of the dorsal space under the carapace of the cephalothorax. The actual percentage of total body weight of the excised hepatopancreas of the 18 small Limuli used in the isotope experiments reported in this paper varied from 7.8 to 18.5~o (12,2 4- 2.8°~i). This is a minimum estimate since 100'~~, of the organ was not recovered in the rapid dissections. The hepatopancreas of large well fed Limuli may exceed 20')Jl] of the total body weight. Limulus muscles are more compact and dense and so distributed that it is more difficult to estimate total quantity. Leg muscle glycogen was obtained from KOH digests of whole legs including the shell. while portions of easily accessible muscles within the abdominal cavity were weighed and used for "body" muscle glycogen isolation. No attempt was made to obtain the small, presumably metabolically very active muscles controlling gill action. The total weight of muscle is certainly of the same order of magnitude, roughly perhaps 50-100~o, of that of hepatopancreas. Thus the relative specific activities of the two tissues

811

may be taken as a first approximation of their relative total involvement and importance. With every Limulus studied, regardless of size, state of nutrition or position of the label on the compound given, the muscle glycogen isolated had incorporated a much greater amount of the glucose than had the hepatopancreas. In some cases, where very little glucose-14C had been incorporated into hepatopancreas glycogen, 10-20times as much had been used in 1-3 hr for glycogen synthesis in muscle. These facts, together with the finding of a more rapid decrease in the quantity of hepatopancreas glycogen in relatively short periods of fasting, are taken to indicate the relative importance of muscle in overall energy metabolism in Limulus. It seems probable that the utilization of glucose for energy production and for metabolic interconversions, including transient storage as glycogen, is primarily a function of muscle. Net glycogen deposition and storage in hepatopancreas occurs only when the availability of nutrients exceeds the immediate needs of the animal and this glycogen can be readily mobilized and nearly depleted leaving the muscle as the major site of continuing carbohydrate metabolism. Molecules of glycogen are inhomogeneous metabolically as well as structurally (Stetten & Stetten, 1960). Glycogen regeneration in the mammal is not a process wherein a newly synthesized glycogen molecule replaces a preexisting one. Even when the quantity of glycogen in a tissue remains relatively constant, it is in a state of dynamic equilibrium. Glycogen labeling, indicating glycogen turnover, results from continual addition and removal of glucosyl residues at the nonreducing peripheral surface of the polysaccharide together with a process of branching, or transglucosylation, which moves the labeled glucosyl residues toward the center of the molecule (Stetten & Stetten, 1954, 1955). A similar process may be assumed to be occurring in the tissues of Limulus but at a slower rate, in keeping with the more sluggish metabolism of this cold blooded invertebrate. In the long term experiment reported in Table 4, number 7, a large well fed Limulus incorporated a relatively large amount of glucose into both hepatopancreas and muscle glycogen following the single injection of glucose [U~4-C, 6-3H]. In the course of the subsequent 6 day fast most of the hepatopancreas glycogen had been utilized but the small amount of glycogen remaining retained a significantly high concentration of isotope. This and the larger quantity of labeled glucose retained in the muscle glycogen probably represents glucosyl residues added at the non-reducing termini of the chains, moved by branching activity toward the central core of the molecules, and not removed by phosphorlytic degradation occurring at the periphery. Hydrogen on C-6 of glucose is lost metabolically at a slower rate than from any other C atom. In perfused rat liver and isolated hepatocytes loss of 3H from glucose-[6-aH, U14-C] has been shown to be only slightly greater than loss of I'*C (Katz & Rognstad, 1975). Here the loss was chiefly pictured as following triose phosphate isomerase catalyzed equilibration between glyceraldehyde-3-P and dehydroxyacetone phosphate followed by proton exchange in the hexose phosphate isomerase reaction after recycling to 6-C sugars. The catabolism of glucose in vivo occurs

812

M ,\RJ(}RII P~. S ll!l I IN

mostly in extrahepatic tissue (Hue. 1981), so the appearance of 3 H 2 0 from tritiated glucose-[6-3H] is more a reflection of metabolism by the whole body than specitically of the liver. In whole animal studies much of the +H on C-6 is lost to 31120 during resynthesis of glucose in the liver from lactic acid produced by glucose breakdown in muscle and also. overwhelmingly by total breakdown of intermediates to COe and HeO. Simultaneous metabolism of glucose dot~bly labeled with 3H and >*C on C-6 has been used in rats to estimate the extent of the occurrcnce of the Cori cycle, i.e. the reutilization of muscle lactate in hepatic glucose synthesis (Dunn et al., 19671. The findings in our double label experiments with glucose-E6-3H] and glucose-t'*C, that the 3H'14c ratio in the Limulus hepatopancreas glycogen is always somewhat lower than that in muscle glycogen, shows that some of the glycogen of the hepatopancreas has been synthesized from glucose degradation products. Measurement of the rate of gluconeogenesis in rats is usually made by injecting labeled lactate and isolating glucose and glycogen (Clark et al., 1974}. Although the quantity of incorporation wtried with the state of nutrition and the a m o u n t of lactate given, significant amounts of lactate were incorporated into rat liver glycogen in 30 min in all cases studied. In contrast, in our experiments with fed Limuli, very little ~4C-injected as pyruvate or lactate was incorporated into glycogen in 3 hr and no *+C from lactate appeared in blood glucose. It thus appears that gluconeogenesis flom pyruvate and lactate are probably not important in Limulus. The use of the term "'hepatopancreas'" for the rnidgut gland of the higher mollusks and a r t h r o p o d s has been questioned (Van Weel. 19741. This multifunctional structure is clearly not a "liver" in the sense of the discrete organ of vertebrates. A careful study of the ultrastructure of the hepatopancreas of Limulus (Herman & Preus, 19721 has revealed three major cell types only one of which, the intertubutar R-cells, contains stored glycogen. The midgut gland has important functions which are carried out by the intestine or pancreas in higher animals. In addition, as discussed by Vonk {1960}, in many cases it is the site of metabolic processes characteristic of liver. Since gluconeogenesis is an important function of liver and kidney in higher animals, evidence for or against its occurrence in invertebrate organs should shed light on a possible liver function. Results of relevant studies are ambiguous. Our finding of little incorporation of ~aC from injected lactate or pyruvate in Limulus is in agreement with the observation of Phillips et al. {1977) that lactic acid, accumulating in the h e m o l y m p h of crayfish and lobsters as a result of forced muscular exercise, disappeared relatively slowly and that labeled lactate, propionate and malate are not incorporated into glucose or glycogen in in t?itro studies of the midgut gland. These authors interpreted their results as arguing against the use of the term "hepatopancreas". O n the other hand, in the same paper Phillips e t a / . (1977) reported the appearance of ~'*C from lactate in blood sugar and the occurrence in the midgut gland of craytish of pyruvate carboxylase and phosphoenolpyruvate carboxykinase, enzymes characteristic of gluconeogenesis. They also cited earlier findings of these enzymes as well as of fructose, 1,6-diphospha-

tase in the midgut glands of crustaceans. O u r tindings of a consistently lower 3H: 14 c ratio is hepatopancreas glycogen than in muscle glycogen from Limulu given doubly labeled glucose is also in accord with the assumption of a liver-like function for the hepatopantreas. Stetten & Goldsmith {1976} found in the hepatopancreas and m coxal glands (analogous to the kidney of vertebrates) of Limulus a particulate glucose-6phosphatase with activities characteristic of the enzyme of vertebrate liver, kidney, and intestinal mucosa. This glucose-6-phosphatase does not occur in other organs nor h;,ts it been found in most lower aninrals. In that paper we speculated on the possible appearance of the gluconeogenic enzyme glucose-6phosphatase, at the evolutionary level of the Arthropods. in those animals which utilize free glucosc as an important circulating form of energy. Perhaps it is in these same animals that cells with true liver-like functions appeared in the midgut gland.

RI':FFRENCI':S Barx'~s R. D. (1980) /mer:eh/'a:~, Zo+do:l.v. 4{h cdn., pp

592 593. Saundcrs College, Philadelphia. Bll~{iM~'~lR H. U. & B~rxl E. (1974) D-Glucose determination with glucose oxidasc and peroxidasc. In Alt,thods (!fEncy,lati¢ Analysis {Editcd by B~R{,i;~t','l]~ H. l,l.k Vol. 3. pp 1205 1215. Academic Press, New York. BI,:owN S. C. & WHIII GlWiI~ E. R. {1971} Cell fractionalion studies on the digestive glands of Limtdus Polyphe,lus. Comp. Biochem. Physiol. 40B, 637 650. Ct,ARK M. G., BL{}XHAMD, P.. H{}IA,ANI) P. C. & LARI}'~ H. A. {1974}. Estimation of the fructose 1,6-diphosphatase-phosphofructokinasc substrate cycle and its relationship to gluconeogenesis in rat liver i, ciPo. J. Biol. Chem. 249, 279 29(/. COHI:N E. (Editor} (19791 Biomedical applications of the horscshoc crab (Limulidae}. Proceedings of a symposium hcld at the Marine Biological Laboratory, Woods Hole. Mass. October. 1978. Alan R. Liss. New York. Cot{I C. F. & Col,~I G. T. (19261 Fate of sugar in the aniinal body tl. The relationship between sugar oxidation and glycogen formation in normal and insulinized rats during the absorption of glucose..I, biol. Chem. 70, 557 576. Dt x:', A.. CHINOVv'IIH M. & S{'HAIFFI R L. D. (19671 Estimation of glucose turnover and the Cori Cycle using glucose-6-t-t'*C. Biochemistry 6, 6 11. HERMAN W. S. & PrH:S D. M. (19721 Ultrastructure of the Hepatopancreas and Associated Tissues of the Chelicerate Arthropod. Limulus polyphemus. Z. Zellfi,'~ch. 134, 255 271. HOHNKI k. & S('Hllir B. T. 119701 Carbohydrate metabolism in Crustaceans. In Chemical Zoolo:ty {Edited by FLOrKIN M. & S('H~I~ B. T.J, Vol. 5. pp 147 166. Academic Press. New York. Hut L. (19761 Use of [14"C. 3HI glucose in the study of so-called futile cycles in liver. Biochem. S+~c. Trans. 4, 994 998. Htr} L. (1981)The role of futile cycles in the regulation of carbohydrate metabolism in the liver. Adr. Encymo/. 25, 247 331. KATZ J. & ROGNSTAD R. {19751 Futile cycles in the metabolism of glucose. Ct,'rent Topics in Celhdar R('~lulation 10, 237 289. KATZ J.. WALS P. A. GOLDliN S. & ROGNSrAD R. {19751 Recycling of glucose by rat hepatocytes. Ear. J. Biochem 60, 9l 10l.

PttlLLIPS, J. W.. M('K1NNIY, R. J. W., HIRD, F. J. R. & MAcMH,t.AN. D. L. {1977} Lactic acid formation in Crus-

Metabolism of glucose and glycogen in Limulns taceans and the liver function of the mid gut gland questioned. Comp. Biochem. Physiol 56B, 427 433. SCHLOTTKI~ E. (1935) Biologische, physiologische and histologische Untersuchungen tiber die Verdauung yon Limulus. Z. t'er~ll. Physiol. 22, 359 413. STETTEN M. R. & GOLDSMITH P. K. (1976) A Limulus glucose-6-phosphatase with phosphotransferase activity characteristic of vertebrate liver microsomes. Bioehim. hiophys. Acta 444, 835- 852. STi!TH-N M. R. & STIiT'rEN D. Jr, (19511 Metabolism of sorbitol compared in normal and alloxan-diabetic rats. J. biol. Chem. 193, 157 165.

813

S'rETTEN M. R. & STtiTTIiN D. Jr. (1954] A study of the nature of glycogen regeneration in the the intact animal. J. biol. Chem. 207, 331 340. STJ!TTiiN M. R. & STI!TTtN D. Jr, (19551 Glycogen regeneration in ~'it'o. J. biol. Chem. 213, 723 732. ST~-ZTTEN D. Jr & ST~iTTt!N M. R. (1960) Glycogen metabolism. Physiol. Ret,. 40, 505 537. VAN WEI{L P. B. (19741 "'Hepatopancreas"?, Comp. Biochem. Physiol. 47A, 1 9, VONK H. J. (1960) Digestion and Metabolism, In The Physioloqy (~f the Crustaeea (Edited by WATERMAN T. H.), Vol 1. Academic Press, New York.