Modulation of glucose metabolic by granulocyte colony-stimulating

response to endotoxin factor

CHARLES H. LANG, GREGORY J. BAGBY, CORNEL DOBRESCU, STEVE NELSON, AND JOHN J. SPITZER Departments of Physiology and of Medicine, Section of Pulmonary/Critical-Care Medicine, Louisiana State University Medical Center, New Orleans, Louisiana 70112-1393 Lang, Charles H., Gregory J. Bagby, Cornel Dobrescu, Steve Nelson, and John J. Spitzer. Modulation of glucose metabolic response to endotoxin by granulocyte colonystimulating factor. Am. J. Physiol. 263 (Regdatory Integrative Comp. PhysioZ. 32): R1122-R1129, 1992.-The present study examines whether in vivo administration of granulocyte colonystimulating factor (G-CSF) and the resultant neutrophilia alters basal glucose metabolism or modulates the glucose metabolic response to a subsequent endotoxin [lipopolysaccharide (LPS)] challenge. Rats were injected with human recombinant G-CSF (50 pg/kg SC) twice daily for 2 days preceding an injection of LPS. Animals treated with G-CSF showed an eightfold increase in blood polymorphonuclear leukocytes (PMNs) but no detectable changes in hemodynamics or glucose metabolism. In control animals, LPS transiently decreased circulating PMN number, but by 4 h neutrophils had returned to control levels. LPS produced a greater reduction in circulating neutrophils in GCSF-treated animals, which did not return to pretreatment levels by 4 h. G-CSF also produced marked changes in the glucose metabolic response to LPS. Rates of whole body glucose production and utilization in both control and G-CSF-treated rats were rapidly increased by LPS; however, the increment in glucose flux was 55100% greater in the latter group. The enhanced rate of hepatic glucose production in this group occurred despite lower plasma levels of lactate and glucagon. The elevated rate of whole body glucose utilization was attributable to the G-CSFenhanced LPS-induced increase in glucose uptake by the ileum, spleen, liver, and lung. Furthermore, LPS increased glucose uptake by skeletal muscle in G-CSF-treated rats but not in control animals. The enhanced glucose disposal in G-CSF-treated rats was not mediated by increases in plasma glucose or insulin concentrations. Whereas the elevated glucose uptake by the liver and lung in rats given G-CSF was associated with an increased number of tissue neutrophils, as assessed by myeloperoxidase activity, sequestration of neutrophils could not account for the changes in glucose uptake by skeletal muscle. These data suggest that the chronic in vivo administration of G-CSF alone does not alter cellular glucose metabolism but does ready cells so that the typical LPS-induced increase in glucose uptake is amplified. Thus G-CSF may be an important biological response modifier under in vivo conditions. neutrophils; gluconeogenesis; carbohydrate metabolism; myeloperoxidase; 2-deoxy-D-glucose; tumor necrosis factor COLONY-STIMULATING FACTOR (G-CSF) is a member of a group of glycoprotein growth factors that regulate proliferation and differentiation of hemopoietic progenitor cells (27). Exogenous administration of either purified or recombinant G-CSF has been shown to dramatically increase circulating neutrophil levels in numerous species (5, 30, 37). In addition, this mediator also stimulates several functional activities of mature polymorphonuclear leukocytes (PMNs). These include an enhanced rate of particle phagocytosis, a priming effect on superoxide anion generation, and an GRANULOCYTE

R1122

0363-6119/92

$2.00 Copyright

0

increased chemotactic activity and antibody-dependent cell-mediated cytotoxicity of granulocytes (20, 30, 40, 42). Furthermore, serum G-CSF concentrations are elevated by infection and bacterial products (3,22), as well as by various cytokines released during these insults such as interleukin-1 (IL-l) and tumor necrosis factor (TNF) (14, 32). These findings, coupled with the observation that G-CSF enhances the functional activity of neutrophils, suggest that G-CSF may be an important mediator of host defense against invading microorganisms. One of the hallmark metabolic manifestations induced by sepsis or LPS is the enhanced rate of glucose utilization (15, 16). This response results from an increased uptake of glucose by a variety of tissues, including liver, lung, skin, spleen, and intestine (19, 25). It has been suggested that the stimulation of glucose uptake in these immune-competent and barrier tissues is important for proper function of these tissues during infection and reflects their increased energy demand (7). Thus the glucose metabolic response may be viewed as an important component of the host defense system. The effects of G-CSF on glucose homeostasis have not been investigated, although this factor was shown to be ineffective at stimulating glucose uptake by bone marrow-derived macrophages (11). Therefore, the present experiments were performed to determine whether in vivo administration of G-CSF influenced basal glucose metabolism or modified the LPS-induced changes in glucose flux. MATERIALS

AND

METHODS

Animal preparation. Male Sprague-Dawley rats (340-380 g, Charles River, Wilmington, MA) were housed in a controlled environment, exposed to a 12:12-h light-dark cycle, and provided with standard rodent chow and water ad libitum for at least 1 wk before initiating experimental procedures. Animals were injected twice daily (10 A.M. and 10 P.M.) subcutaneously with human recombinant G-CSF (Amgen, Thousand Oaks, CA) at a dose of 50 pg/kg for 2 days before the experiment. Time-matched control animals were injected subcutaneously with an equal volume (100 pl/lOO g body wt) of sterile saline. On the day before an experiment, animals were anesthetized with an intramuscular injection of ketamine and xylazine (90 and 9 mg/kg body wt, respectively). Using aseptic surgical techniques, we placed a catheter in the arch of the aorta via the left carotid artery and two catheters in the right jugular vein (15). After surgery, rats were housed in individual cages, fasted overnight, and provided with water ad libitum. The rats were conscious and unrestrained throughout the remainder of the experiment. Experimental protocols. Experiments were started at 0600 h the day after surgery, which was ~8 h after the final injection

1992 the American

Physiological

Society

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.237.035.237) on January 20, 2019.

G-CSF

AND

GLUCOSE

of G-CSF. At this time whole body glucose kinetics were measured using the primed, constant intravenous infusion of [ 3-“H] glucose (high-performance liquid chromatography purified, specific activity 13.5 Ci/mmol, DuPont-New England Nuclear, Boston, MA) as previously described by our laboratory (18). Blood samples for basal glucose kinetic determinations were withdrawn from the arterial catheter at 120 and 140 min (0.3 and 1 ml) after the start of the tracer infusion. Plasma glucose and lactate concentrations, as well as glucose specific activity, were determined on both samples; plasma concentrations of insulin and glucagon were only determined on the second blood sample. Total and differential leukocyte counts were also determined on the second sample. Mean arterial blood pressure (MABP) and heart rate were measured (Grass model 79E, Quincy, MA) before the first blood sample by a pressure transducer attached to the arterial catheter. After we obtained samples for basal metabolic determinations, rats previously administered either G-CSF or nonpyrogenie saline received an intravenous injection of either Escherichia coZi LPS (100 pg/lOO g body wt, 10% lethal dose at 24 h, Difco Olll:B4, Detroit, MI) or an equal volume (0.2 ml/l00 g body wt) of isotonic saline. LPS was injected via the second venous catheter so as not to alter the flow rate of the radiolabeled tracer. The lyophilized LPS was reconstituted with isotonic nonpyrogenic saline and filtered through a 0.45~pm sterile filter (Millex-HA) before injection. Arterial blood samples were obtained at various times after the injection of LPS for the determination of glucose, lactate, glucose specific activity, leukocyte counts, and plasma hormone concentrations. Blood pressure and heart rate were also determined at hourly intervals after the administration of LPS. In a separate group of animals, in vivo glucose uptake by individual tissues was determined using 14C-labeled 2deoxy-Dglucose (dGlc), a nonmetabolizable glucose analogue, as described previously by our laboratory (16, 24). Animals were injected with G-CSF (or saline) as described above and then subsequently injected with either LPS or saline. A tracer amount of 2-deoxy-D-[U-14C]glucose (8 &i/rat; specific activity 328 mCi/mmol; Amersham, Arlington Heights, IL) was injected intravenously 200 min after the administration of LPS, and tissues were obtained 40 min later. Serial blood samples (0.3 ml) were withdrawn, plasma was deproteinized, and radioactivity was determined (Beckman, LS7500, Fullerton, CA). An aliquot of the supernatant for each sample was neutralized, and the glucose concentration was determined. Animals were anesthetized with pentobarbital sodium at the end of the 40-min in vivo labeling period (i.e., 4 h after LPS injection). Selected tissues were then excised to determine the intracellular accumulation of phosphorylated metabolites of dGlc (P-dGlc). In this group of animals, serum TNF activity was also determined 90 min after the injection of LPS or saline. Total and differential leukocyte counts were also performed on representative animals from each of the four experimental groups. In a third group of similarly treated rats, selected tissues were obtained for tlhe determination of myeloperoxidase (MPO) activity, percent dry weight, and, where possible, whole organ weight. Plasma TNF levels were also determined at 90 min post-LPS in this series of experiments. Analytic procedures. Plasma glucose and lactate concentrations were determined enzymatically on neutralized supernatants of deproteinized plasma (21). Plasma glucose specific activity was determined as previously described (18). Glucose rates of appearance (R,) and disappearance (Rd) were calculated using non-steady-state equations of Steele (36). The metabolic clearance rate (MCR) for glucose, which measures the avidity of the tissues for glucose, was calculated by dividing the glucose Rd by the prevailing plasma glucose concentration. The glucose metabolic rate (R,) for each tissue examined was calculated

R1123

METABOLISM

based on the in vivo accumulation of P-dGlc by a respective tissue, the integrated dGlc-to-glucose ratio in plasma during the 40-min labeling period, and the lumped constant as described previously (19, 25). Plasma immunoreactive insulin and glucagon concentrations were determined by radioimmunoassay using porcine standards (ICN, Irvine, CA). Total white blood cells were counted in a hemacytometer (Fisher Scientific, Pittsburgh, PA), and differential counts were determined after Wright staining. Serum TNF levels were assessed by a cytotoxicity assay utilizing L929 cells obtained from American Type Culture Collection (Rockville, MD), as previously described (7). Tissue MPO activity was determined by the method of Grisham et al. (10). Experimental values are presented as means t SE. The number of animals per group is indicated in the legends of Table 1 and Figs. l-7. Data were analyzed using analysis of variance followed by Newman-Keuls test to determine treatment effect. Statistical significance was set at P < 0.05. RESULTS

White blood cell kinetics. Figure 1 illustrates the changes in total white cells, granulocytes, and lymphocytes after the administration of LPS in G-CSF-treated and control rats. The twice daily injection of G-CSF for 2 days increased the total white blood cell number by >40,000 cells/pi, which represents a 390% increase above basal control values. The large majority (78%) of this increase was accounted for by the increment in circulating granu.locy tes, although an elevation in plasma lymphocytes was also evident. G- CSF produced a 9.8- and Total

F

WBC

Blood

Granulocytes *

40

I

Blood

Lymphocytes

PRE

SO TIME

POST--LPS

r

SAL/SAL

m

G

120

40

(min) m

SAL/LPS

CSF/SAI

Fig. 1. Total number of white blood cells (WBC), granulocytes, and lymphocytes in control and granulocyte colony-stimulating factor (GCSF)-treated rats after administration of lipopolysaccharide (LPS). Four groups of animals were used in this study: saline (SAL)-saline, saline-LPS, G-CSF-saline, and G-CSF-LPS . Values are means t SE: n = 10-15. * P < 0.05 compared with time-matched saline-saline value; “f P < 0.05 compared with time -matched saline-G-CSF values.

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.237.035.237) on January 20, 2019.

R1124

G-CSF

AND

GLUCOSE

2.4fold increase in the circulating number of granulocytes and lymphocytes, respectively, compared with control values. In control rats, LPS produced a 41 and 52% decrease in the number of leukocytes by 30 and 120 min, respectively. By 240 min after LPS, the leukocyte number had returned to control values. The reduction at 30 min post #-LPS was due to a decrease in both the number of granulocytes and lymphocytes, whereas the leukopenia at later time points was essentially the result of a decreased number of lymphocytes. In G-CSF-treated rats receiving LPS, leukocytes were reduced by 87, 97, and 80% at 30, 120, and 240 min, respectively, compared with G-CSF control values. Although the decrease in leukocytes was primarily due to the reduction in granulocytes, blood lymphocytes were also significantly decreased by LPS at each time sampled. Compared with time-matched values from the saline-LPS group, there were no statistical differences in the number of total white cells, granulocytes, or lymphocytes in G-CSF-treated rats after LPS. Timematched control animals receiving saline in place of LPS showed essentially no alterations in the hematological profile at the times sampled. Whole body hemodynamics and glucose metabolism. In control animals, MABP was decreased 20-40% between 60 and 180 min after the injection of LPS; however, by 240 min blood pressure had returned to control levels (Fig. 2). Animals pretreated with G-CSF showed a similar hypotensive response after LPS. Furthermore, both control and G-CSF-treated rats showed a similar increase in heart rate in response to LPS, which averaged between 10 and 2 7% over the last 2 h of the experiment. In control rats, LPS increased plasma glucose concentrations as early as 30 min after administration (Fig. 3). The hyperglycemia peaked at 60 min and gradually declined, and glucose concentrations were not different from pre-LPS values by 180 min. G-CSF-treated rats showed an essentially identical hyperglycemic response to

METABOLISM GLUCOSE

GLUCOSE G Y >

10

Ra

70 60

s8 & 6

4 I

I

I

I

I

LACTATE

;LUCOSE 21

MCR *t

.\E 8\s

PRE

60 120 TIME POST-LPS

180

240 (min

6-

PRE

60 120 180 240 TIME POST-LPS (min)

Fig. 3. Plasma glucose and lactate concentrations, rate of glucose appearance (R,), and metabolic clearance rate (MCR) in control and GCSF-treated rats after administration of LPS. 0, saline-saline; 0, salineLPS; A, G-CSF-saline; and A, G-CSF-LPS. Values are means t SE; n = 7-8 per group. * P < 0.05 compared with time-matched saline-saline or G-CSF-saline values from same group; t P < 0.05 compared with timematched saline-LPS value.

the injected LPS. A rapid and sustained increase (2- to 6-fold) in plasma lactate values was seen in saline-LPS animals throughout the experiment (Fig. 3). After G-CSF, the hyperlactacidemic response to LPS was present but attenuated. In these animals, plasma lactate levels were significantly lower than saline-LPS values at 30, 180, and 240 min. LPS also produced a rapid increase in the glucose R, in both control and G-CSF-treated rats (Figure 3). The initial response was similar in both groups, but the increment in glucose R, was 55100% greater in rats receiving G-CSF than in control animals 3-4 h after LPS administration. Similarly, calculated rates of whole body glucose disappearance were increased in rats receiving LPS and further enhanced in animals administered G-CSF (data not shown). The late increase in MCR produced by LPS was also enhanced (147-184%) by the administration of G-CSF. Circulating concentrations of insulin, glucagon, and TNF. Table 1 presents the plasma insulin and glucagon

TIME

POST-LPS

(min)

Fig. ‘2. Mean arterial blood pressure (MABP) and heart rate in control and G-CSF-treated rats after administration of LPS. 0, saline-saline; 0, saline-LPS; A, G-CSF-saline; and A, G-CSF-LPS. Values are means k SE; n = 7-8 per group. * P < 0.05 compared with time-matched appropriate control value (i.e., either the saline-saline or G-CSF-saline value).

concentrations in the four groups of experimental animals. Neither G-CSF nor LPS, alone or in combination, altered insulin levels at the times sampled. In control rats, LPS increased glucagon concentrations by 160 and 349% at 1 and 4 h. Twice daily injections of G-CSF did not alter basal glucagon levels. However, G-CSF tended to blunt the LPS-induced hyperglucagonemia, although this change was only statistically significant at 1 h after LPS. Numerous investigators have demonstrated that the peak TNF response occurs -90 min after the injection of LPS (29). Therefore, in the present study blood was collected at this time point to determine whether G-CSF altered the LPS-induced increase in serum TNF levels. In saline-LPS rats, the TNF levels were 25,634 t 4,472

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.237.035.237) on January 20, 2019.

G-CSF

Table 1. Plasma concentrations

AND

GLUCOSE

of insulin and glucagon

in control and G-CSF-treated rats injected with LPS Insulin,

pU/ml

Glucagon,

pg/ml

Group lh

4h

4h

lh

25t2 2723 199t21 171t22 Saline + saline 31t5 22k3 517t24* 768&177* Saline + LPS 21t3 22t4 137t28 153t9 G-CSF + saline 34t6 28t2 396&18*“f 542t53* G-CSF + LPS Values are means t SE; n = lo-14 per group. G-CSF, granulocyte colony-stimulating factor; LPS, lipopolysaccharide. * P < 0.05 compared with time-matched saline-saline or G-CSF-saline values; t P < 0.05 compared with time-matched saline + LPS values.

U/ml. This value was not statistically different from that determined in G-CSF-LPS rats (29,369 t 3,716 U/ml). TNF levels were below the limit of detection of our assay (~40 U/ml) for animals in the saline-saline and G-CSFsaline groups. Glucose uptake by individual tissues. The first series of experiments indicated that G-CSF enhanced the LPSinduced increase in whole body glucose disposal 3-4 h after LPS administration. Therefore, an additional study was performed to determine whether G-CSF produced a generalized increase in glucose uptake by all tissues or only enhanced uptake in selective tissues. In control animals, glucose uptake was not elevated in skeletal muscle 4 h after the injection of LPS, regardless of the fiber type composition (Fig. 4). However, LPS markedly increased glucose uptake in all skeletal muscles sampled from GCSF-treated animals. The increase ranged from a low of 44% in the gastrocnemius to a high of 222% in the white quadriceps. It should also be noted that the twice daily administration of G-CSF alone did not alter glucose uptake in muscle or in any other tissue sampled. There were WHITE

GASTROCNEMIUS

QUADRICEPS

Rll25

METABOLISM

seven tissues where LPS increased glucose uptake in both control and G-CSF-treated rats. Figure 5 illustrates that after LPS, glucose uptake by skin, kidney, and adipose tissue from control rats was increased by 57,49, and 65%, respectively. The magnitude of the LPS-induced increase in glucose uptake by these tissues from G-CSF-treated rats was similar. In contrast, the LPS-induced increase in glucose uptake by ileum, spleen, liver, and lung was greater in rats administered G-CSF than in control animals (Fig. 6). In G-CSF-LPS rats, the increase in glucose uptake was 156, 93, 75, and 15% greater, respectively, for these tissues than in control animals receiving LPS. Cerebral glucose uptake was not altered by administration of G-CSF and/or LPS (Fig. 5). Tissue MPO activity. As described previously, the injection of LPS produced a severe neutropenia. It is possible that the margination or infiltration of large numbers of neutrophils within tissues after LPS may be in part responsible for the enhanced glucose uptake. Therefore, tissue MPO activity was determined to assess the accumulation of granulocytes within specific tissues (Fig. 7). In control animals, LPS did not alter MPO activity in muscle, skin, or intestine; however, it did increase MPO activity in kidney, liver, spleen, and lung (155, 140, 188, and 236%, respectively). The twice daily injection of GCSF for 2 days increased MPO activity only in the spleen and lung (329 and 135%). In G-CSF-treated rats, LPS increased MPO activity in the ileum, kidney, liver, and lung (161, 225, 262, and 140%, respectively). For the kidney, liver, spleen, and lung, the LPS-induced increase in MPO activity was greater in G-CSF-treated rats than it was in animals administered saline. The data in Figs. 4-7 have been normalized to per gram wet weight of tissue. It is possible, however, that LPS and/or G-CSF may have produced fluid extravasation in some or all of the tissues examined. Therefore, the percent dry weight and total organ weight (where possible)

*t

120

KIDNEY

SKIN 100 60 80

1 512

1

60

120

4

*

* 350,

300

*

I

j

250 40

90

4

200

20

150

60 0 1

n

-

100 50

SOLEUS v E 5

I

-I-

250

100

200

80

0 :

\

2

z cr"

60

E c

600

-

100

-60

500

-

A?

400

-

300

-

50

20 0 LrliiL

BRAIN

+

150 1

40

FAT 80

I

40 i 0

CONTROL

G-CSF

CONTROL

G-CSF

Fig. 4. Glucose metabolic rate (R,) of hindlimb skeletal muscles obtained from control and G-CSF-treated rats 4 h after administration of saline or LPS. Values are means f. SE; n = 6-8 per group. Open bars, saline injected; hatched bars, LPS injected (100 pg,/lOO g body wt). * P < 0.05 compared with either saline-saline or G-CSF-saline values from same group; “f P < 0.05 compared with saline-LPS value. Data are expressed per gram wet wt of tissue. LPS and/or G-CSF did not alter the percent dry wt of any tissue examined.

20

0 CONTROL

G-CSF

CONTROL

G-CSF

Fig. 5. R, of skin, kidney, fat, and brain obtained from control and G-CSF-treated rats 4 h after administration of saline or LPS. Values are means t SE; n = 6-8 per group. Open bars, saline injected; hatched bars, * P < 0.05 compared with either saline-saline or G-CSFLPS injected. saline values for same group.

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.237.035.237) on January 20, 2019.

RI 126

G-CSF ILEUM

GLUCOSE

SPLEEN *t

600

*t

1400 1200

500

s \c .-

AND

1000

400

800

300

600

200

400

100

200 0 LIVER

LUNG *t

E 200 -1

600

*t

7

*

500 400

CY

300

100

200 50

100 0 1

0 CONTROL

CONTROL

G-CSF

Fig. 6. R, of ileum, spleen, liver, and lung obtained from control and G-CSF-treated rats 4 h after administration of saline or LPS. Values are means + SE; n = 6-8 per group. Open bars, saline injected; hatched bars, LPS-injected. *‘P < 0.05 compared with either saline-saline or G-CSFsaline values for same group. “f P < 0.05 compared with saline-LPS values.

were determined. Neither LPS nor G-CSF, alone or in combination, altered either of these variables in any of the tissues examined (data not shown). In a small group of animals (n = 3) the G-CSF was heated ( 100°C for 1 h) before administration. Animals injected twice daily for 2 days with the heat-inactivated G-CSF showed no alteration in basal leukocyte number or comparable LPS-induced changes in whole body glucose metabolism to those observed in saline-injected rats. DISCUSSION

The primary physiological role of the various colonystimulating factors (CSFs) is generally considered to be in regulating the proliferation and differentiation of hemopoietic progenitor cells (41). In the present study, where G-CSF was injected for 2 days and the experiment was performed on day 3, it is likely that the peripheral neutrophilia resulted from both the increased proliferation of progenitor cells and the release of preformed cells from bone marrow and tissue storage pools (13). Several reports indicate that the G-CSF-induced increase in total white cell number is due to a selective elevation in granulocytes (5, 37). However, in this and other studies (13, 42), while the large majority (80%) of the increment was indeed the result of a marked neutrophilia, the number of circulating lymphocytes was also moderately elevated. Furthermore, these data confirm that the biological activities of G-CSF are highly species cross-reactive, which is likely the result of the strong conservation of receptorbinding domains between species (27). As described earlier, G-CSF also influences mature cell function (20, 30, 40, 42). The results of the present study extend these observations and indicate the dramatic immunomodulatory action of G-CSF on whole body and tissue glucose utilization. Multiple injections of G-CSF did not produce sustained alterations in glucose metabolism 8-10 h after the final injection, which is the time all

METABOLISM

experiments were performed. Therefore, any relatively transient changes in metabolism produced by the acute administration of G-CSF would not have been detected. The presence of transient changes cannot be discounted because previous work by our laboratory showed that the acute intravenous injection of granulocyte-macrophage (GM)-CSF increases in vivo glucose uptake by a variety of tissues for at least 6 h (34). Alternatively, the intravenous injection of G-CSF at the dose used in the present study may not produce detectable changes in glucose metabolism, despite the presence of elevated circulating levels of G-CSF (38). This latter possibility is consistent with the differential effects of GM- and G-CSF on cell function and metabolism (9, 11). In control animals, LPS produced a transient reduction in mean arterial blood pressure and in the circulating number of neutrophils and lymphocytes. These parameters returned to basal (pre-LPS) levels by 4 h after LPS. In addition, LPS produced sustained changes in carbohydrate metabolism, as reflected by the hyperlactacidemia and the elevated rates of glucose production, utilization, and clearance. The increased whole body glucose consumption observed 4 h after LPS administration resulted from an enhanced uptake by a variety of immunecompetent and barrier tissues, such as the ileum, spleen, liver, lung, and skin. In contrast, skeletal muscle, regardless of fiber type predominance, did not show any detectable change in glucose uptake at this time. In general, these LPS-induced changes in control animals are similar to those reported by our laboratory and others and have been discussed previously (8, 16, 28). An accumulation of neutrophils in the liver and lung after intravenous LPS administration is known to occur (1, 24). However, a systematic examination of LPS-induced neutrophil sequestration in other tissues has not been reported. Several investigators have used tissue MPO activity to quantitate neutrophil accumulation within selected tissues in other pathological conditions (I, 10, 35). Our study demonstrates that 4 h after LPS administration MPO activity was increased in the kidney, liver, spleen, and lung; no change was seen in skeletal muscle, skin, or intestine. The intravascular margination and/or infiltration of neutrophils in the kidney, liver, spleen, and lung during endotoxemia was associated with an enhanced glucose uptake in these tissues. Because endotoxin is known to stimulate glucose uptake by neutrophils (1, 24), it is possible that the sequestered neutrophils are at least partially responsible for the increased glucose disposal. Such a situation has been demonstrated to occur in the liver where sequestered neutrophils have -20% of the increase in total been shown to contribute hepatic glucose uptake (24). However, at this time, there is no direct evidence that the LPS-induced accumulation of neutrophils within the kidney, spleen, and lung contribute significantly to the increase in glucose uptake seen in these tissues. Furthermore, because MPO activity was not elevated in the intestine and skin, the LPS-induced increase in glucose uptake by these tissues appears to be mediated by a PMN-independent mechanism. The dual actions of G-CSF on precursor cells as well as mature neutrophils suggest that this cytokine plays an

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.237.035.237) on January 20, 2019.

G-CSF

-

Red

Quadriceps

White

Quadriceps

AND

GLUCOSE

RI 127

METABOLISM

Skin

Ileum

3.0

.-07 0 ?

2.5 6 2.0 1.5 1 .o 0.5 0.0

v, -w .c 3 -

Kidney

Liver

24

50

1

* t

100

1

*t

* & Control

G-CSF

#* :& 1SP een

Control

180

1

Lung

I

1 *

Fig. 7. Myeloperoxidase (MPO) activity of various tissues obtained from control and G-CSF-treated rats 4 h after administration of saline or LPS. Values are means t SE; n = 6. Open bars, saline injected; hatched bars, LPS injected. *‘P < 0.05 compared with either saline-saline or G-C SF-saline values for same group. with salin .e-LPS tp < 0.05 compared values. # P < 0.05 compared with salinesaline values.

90 60 30

0

G-CSF

Control

G-CSF

important role in host resistance to infection. Studies have now shown G-CSF protects animals from lethal infections produced by a variety of microorganisms (13,23). The ability of G-CSF to attenuate LPS-induced lethality, however, has not been reported. In the present study all LPS-treated rats survived the experimental protocol, but long-term survival was not assessed.On the basis of the slightly smaller hypotensive response and the significantly lower levels of lactate and glucagon in response to LPS administration, G-CSF pretreatment certainly did not exacerbate the insult. These findings are consistent with the results of Havill et al. (I2), who demonstrated that G-CSF did not significantly increase mortality after an LPS challenge. G-CSF functioned as a potent biological response modifier under our in vivo conditions. Compared with salineinjected animals, the LPS-induced granulocytopenia was more severe in G-CSF-treated rats. Whereas the number of neutrophils fell by -2,300 cells/p1 in control animals, there was a 97% decrease in neutrophils in G-CSF rats, which represented a reduction of -34,700 cells/pi. The rapid neutropenia after LPS administration occurs as a result of an increased rate of removal that exceeds an elevated rate of entry into the circulation (4). The enhanced PMN removal in LPS-injected animals is potentially influenced by at least two mechanisms: a reduction in tissue perfusion pressure (39) and an increased neutroPhil-endothelial adhesiveness (33). Although G-CSFtreated rats had a similar hypotensive response as control animals to LPS, changes in cardiac output and regional blood flow were not determined. Therefore, alterations in regional perfusion pressure cannot be excluded as an explanation for the increased neutrophil margination. However, it is more likely that G-CSF increased tissue-associated neutrophils by increasing adherence properties. G-CSF is known to upregulate the expression of CD14 on the surface of neutrophils and increase the adhesive activity of CR3 in addition to increasing neutrophil number (43). On the basis of MPO determinations, increased numbers of neutrophils became sequestered in the intes-

Control

G--CSF

tine, kidney, liver, and lung in G-CSF-treated rats administered LPS. Although G-CSF did not alter the LPS-induced hyperglycemia, it substantially modified other aspects of the glucose metabolic response to endotoxin. The early (O-2 h) increase in glucose production and utilization was similar in G-CSF- and saline-injected animals administered LPS. However, by 4 h the increment in the glucose turnover rate in G-CSF-treated rats was twofold that seen in time-matched control animals injected with LPS. The elevated rate of gluconeogenesis in G-CSF-LPS rats was associated with a smaller rise in the plasma concentration of the major gluconeogenic precursor (e.g., lactate) and a tendency for lower glucagon levels. These changes may indicate that G-CSF increases hepatic sensitivity to glucagon and possibly other gluconeogenic modulators. Thus when stimulated by LPS a smaller increase in glucagon is required to maintain the same or higher rates of gluconeogenesis. Indeed, the smaller increase in plasma lactate may signify an enhanced rate of glucose production from lactate; however, we cannot exclude the possibility that the lower lactate levels seen in G-CSF-LPS rats result from a decreased rate of lactate production by peripheral tissues. The rate of glucose utilization by peripheral tissues is also elevated in G-CSF-LPS rats compared with the saline-LPS group. This increased glucose consumption in G-CSF-treated rats was not mediated by a mass action effect of glucose or by an elevation in plasma insulin levels. Indeed, when the glucose Rd was normalized for the prevailing plasma glucose concentration, G-CSF was shown to enhance the LPS-induced increase in the glucose metabolic clearance rate. This indicates that G-CSF, directly or indirectly, increased the avidity of tissues to take up glucose. The enhanced rate of whole body glucose disposal was not due to a generalized increase in glucose uptake by all tissues. There were two general classesof tissues that were responsive to the actions of G-CSF. Skeletal muscles showed an LPS-induced increase in glucose uptake in G-CSF-treated rats but not in animals

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.237.035.237) on January 20, 2019.

R1128

G-CSF

AND

GLUCOSE

injected with saline before endotoxin. The second group of tissues that responded to G-CSF were immune-competent tissues, including the intestine, spleen, liver, and lung. In these tissues, LPS increased glucose uptake under basal conditions, but a larger increase was seen in tissues removed from G-CSF-pretreated animals. In contrast to skeletal muscle where the enhanced glucose uptake was not associated with an accumulation of neutrophils, the increased uptake seen in these latter tissues was coincident with an increased number of sequestered neutrophils. We have previously shown that other cytokines, such as IL-1 and TNF, can enhance tissue glucose uptake (2, 17). However, treatment of animals with G-CSF did not alter the peak serum TNF response to LPS. In summary, the in vivo administration of G-CSF to rats produces a marked neutrophilia which, in the absence of a secondary stimulus, results in no detectable changes in hemodynamics or carbohydrate metabolism. In contrast, G-CSF primed cells so that the subsequent administration of the potent immunomodulator, LPS, produced a greater increase in glucose utilization than in control animals. This enhanced uptake of glucose was matched by a proportional increase in glucose production, thus maintaining euglycemia. Although the margination and/or infiltration of circulating neutrophils may be an important contributor to the enhanced uptake of glucose by the liver, spleen, intestine, and lung under this condition, it cannot explain the elevated glucose uptake by muscle. Because glucose uptake in neutrophils is coupled with the activity of the hexose monophosphate shunt (6), the ability of G-CSF to enhance glucose uptake by neutrophils and other cells in response to bacterial cell wall products is a potentially important mechanism by which this factor modulates host defense. We thank Jean Carnal and Howard Blakesley for their excellent technical assistance. This work was supported by National Institute of General Medical Sciences Grants GM-32654 and GM-38032. Granulocyte colony-stimulating factor was a generous gift from Amgen. Present address for C. H. Lang and address for reprint requests: Dept. of Surgery, State Univ. of New York at Stony Brook, Nicolls Rd., Health Science Center, T19,020, Stony Brook, NY 11794-8191. Received

3 September

1991; accepted

in final

form

5 May

1992.

REFERENCES 1. Anderson, B. O., D. D. Bensard, J. M. Brown, J. E. Repine, P. F. Shanley, J. A. Leff, L. S. Terada, A. Banerjee, and A. H. Harken. FNLP injures endotoxin-primed rat lung by neutrophil-dependent and -independent mechanisms. Am. J. Physiol. 260 (Regulatory Integrative Comp. Physiol. 29): R413R420, 1991. 2. Bagby, G. J., C. H. Lang, N. Skrepnik, and J. J. Spitzer. Attenuation of glucose metabolic changes resulting from TNF-cu administration by adrenergic blockade. Am. J. PhysioZ. 262 (Regulatory Integrative Comp. Physiol. 31): R628-R635, 1992. 3. Cheers, C., A. M. Haigh, A. Kelso, D. Metcalf, E. R. Stanley, and A. M. Young. Production of colony-stimulating factors (CSFs) during infection: separate determinations of macrophage-, granulocyte, granulocyte-macrophage-, and multi-CSFs. Infect. Immun. 56: 247-251, 1988. 4. Chervenick, P. A., D. R. Boggs, J. C. March, G. E. Cartwright, and M. M. Wintrobe. The blood and bone marrow neutrophil response to graded doses of endotoxin in mice. Proc. Sot. Exp. Biol. Med. 126: 891-895, 1967. 5. Cohen, A. M., K. M. Zsebo, H. Inoue, D. Hines, T. C. Boone, V. R. Chazin, L. Tsai, T. Ritch, and L. M. Souza. In vivo

METABOLISM

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21. 22.

23.

24.

25.

26.

stimulation of granulopoiesis by recombinant human granulocyte colony-stimulating factor. Proc. Natl. Acad. Sci. USA 84: 24842488, 1987. DeChatelet, L. R. Initiation of the respiratory burst in human polymorphonuclear neutrophils: a critical review. J. Reticuloendoth. Sot. 24: 73-91, 1978. D’Souza, N. B., A. P. Bautista, G. J. Bagby, C. H. Lang, and J. J. Spitzer. Acute ethanol intoxication suppresses E. coli lipopolysaccharide enhanced glucose utilization by hepatic nonparenchymal cells. Akohol. CZin. Exp. Res. 15: 249-254, 1991. Feuerstein, G., J. M. Hallenbeck, B. Vanetta, R. Rabinovici, P. Y. Perera, and S. N. Vogel. Effects of gram-negative endotoxin on levels of serum corticosterone, TNF, circulating blood cells, and the survival of rats. Circ. Shock 30: 265-278, 1990. Gabrilove, J. L., and A. Jakubowski. Granulocyte colonystimulating factor: preclinical and clinical studies. Hematol. Oncol. 3: 427-440, 1989. Grisham, M. B., L. A. Hernandez, and D. N. Granger. Xanthine oxidase and neutrophil infiltration in intestinal ischemia. Am. J. Physiol. 251 (Gastrointest. Liver Physiol. 14): G567G574, 1986. Hamilton, J. A., G. Vairo, and S. R. Lingelbach. Activation and proliferation signals in murine macrophages: stimulation of glucose uptake by hemopoietic growth factors and other agents. J. Cell. Physiol. 134: 405-412, 1988. Havill, A. M., J. W. Andresen, E. Korach, and S. Nelson. Enhancement of endotoxin-induced lethality in the mouse by GM-CSF (Abstract). Am. Rev. Respir. Dis. 143: A236, 1991. Iguchi, K., S. Inoue, and A. Kumar. Effect of recombinant human granulocyte colony-stimulating factor administration in normal and experimentally infected newborn rats. Exp. Hematol. 19: 352-358, 1991. Koeffler, H. P., J. Gasson, J. Ranyard, L. Souza, M. Shephard, and R. Munker. Recombinant human TNF stimulates production of granulocyte colony-stimulating factor. Blood 70: 5559, 1987. Lang, C. H., G. J. Bagby, H. L. Blakesley, and J. J. Spitzer. Importance of hyperglucagonemia in eliciting the sepsisinduced increase in glucose production. Circ. Shock 29: 181-192, 1989. Lang, C. H., G. J. Bagby, and J. J. Spitzer. Glucose kinetics following lethal and sublethal doses of endotoxin. Am. J. Physiol. 248 (Regulatory Integrative Comp. Physiol. 17): R471-R478, 1985. Lang, C. H., and C. Dobrescu. Interleukin-1 induced increases in glucose utilization are insulin mediated. Life Sci. 45: 2127-2134, 1989. Lang, C. H., and C. Dobrescu. Sepsis-induced changes in in vivo insulin action in diabetic rats. Am. J. Physiol. 257 (Endocrinol. Metab. 20): E301-E308, 1989. Lang, C. H., and C. Dobrescu. Gram-negative infection increases noninsulin-mediated glucose disposal. Endocrinology 128: 645-653, 1991. Lopez, A. F., N. A. Nicola, A. W. Burgess, D. Metcalf, F. L. Battye, W. A. Sewell, and M. A. Vadas. Activation of granulocyte cytotoxic function by purified mouse colony-stimulating factors. J. Immunol. 131: 2983-2988, 1983. Lowry, 0. H., and J. C. Passonneau. A Flexible System of Enzymatic Analysis. San Diego, CA: Academic, 1972. MacKensen, A., C. Galanas, and R. Engelhardt. Treatment of cancer patients with endotoxin induces release of endogenous cytokines. Pathobiology 59: 264-267, 1991. Matsumoto, M., S. Matsubara, T. Matsuno, M. Tamura, K. Hattori, H. Nomura, M. Ono, and T. Yokota. Protective effect of human granulocyte colony-stimulating factor on microbial infection in neutropenic mice. Infect. Immun. 55: 2715-2720, 1987. Meszaros, K., J. Bojta, A. P. Bautista, C. H. Lang, and J. J. Spitzer. Glucose utilization by Kupffer cells, endothelial cells and granulocytes in endotoxemic rat liver. Am. J. Physiol. 260 (Gastrointest. Liver Physiol. 23): GTG12, 1991. Meszaros K., C. H. Lang, G. J. Bagby, and J. J. Spitzer. Contribution of different organs to increased glucose consumption after endotoxin. J. BioZ Chem. 262: 10965-10970, 1987. Metcalf, D. The granulocyte-macrophage colony-stimulating

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.237.035.237) on January 20, 2019.

G-CSF

AND

GLUCOSE

factors. Science Wash. DC 229: 16-22, 1985. 27. Metcalf, D. The molecular biology and function of the granulocyte-macrophage colony-stimulating factors. Blood 67: 257-267, 1986. 28. Molina, P. E., 6. H. Lang, G. J. Bagby, and J. J. Spitzer. Ethanol attenuates the endotoxin-enhanced glucose utilization. Am. J. Physiol. 258 (Regulatory Integrative Comp. Physiol. 27): R987-R993, 1990. 29 Nelson, S., G. J. Bagby, G. G. Bainton, L. A. Wilson, J. J. Thompson, and W. R. Summer. Compartmentalization of intraalveolar and systemic lipopolysaccharide-induced tumor necrosis factor and pulmonary inflammatory response. J. Infect. Dis. 159: 189-194, 1989. 30 Ohsaka, A., S. Kitagawa, S. Sakamoto, Y. Miura, N. Takanashi, F. Takaku, and M. Saito. In vivo activation of human neutrophil functions by administration of recombinant human granulocyte colony-stimulating factor in patients with malignant lymphoma. Blood 74: 2743-2748, 1989. 31 Proctor, R. A. Endotoxin in vitro interactions with human neutrophils: depression of chemiluminescence, oxygen consumption, superoxide production and killing. Infect. Immun. 25: 912-921, 1979. 29 Rennick, D., G. Yang, L. Gemmell, and F. Lee. Control of hemopoiesis by a bone marrow stromal cell clone: lipopolysaccharide and interleukin1 -inducible production of colony-stimulating factors GM-CSF and G-CSF. Blood 69: 682-691, 1987. 33. Schleimer, R. P., and B. K. Rutledge. Cultured human vascular endothelial cells acquire adhesiveness for neutrophils after stimulation with interleukin 1, endotoxin, and tumor-promoting phorbol diesters. J. Immunol. 136: 649-654, 1986. 34. Schuler, A., Z. Spolarics, C. H. Lang, G. J. Bagby, S. Nelson, and J. J. Spitzer. Upregulation of glucose metabolism by granulocyte-monocyte colony-stimulating factor. Life Sci. 49: 899906, 1991. 35. Smith, J. K., M. B. Grisham, D. N. Granger, and R. J. Korthus. Free radical defense mechanisms and neutrophil infil-

METABOLISM

36. 37.

38.

39.

40.

“a.

41.

42.

43.

R1129

tration in postischemic skeletal muscle. Am. J. Physiol. 256 (Heart Circ. Physiol. 25): H789-H793, 1989. Steele, R. Influence of glucose loading and of injected insulin on hepatic glucose output. Ann. NY Acad. Sci. 82: 420-430, 1959. Tamura, M., K. Hattori, H. Nomura, M. Oheda, N. Kubota, I. Imazeki, M. Ono, Y. Ueyama, S. Nogata, N. Shirafuji, and S. Asano. Induction of neutrophilic granulocytosis in mice by administration of purified human native granulocyte colonystimulating factor (G-CSF). Biochem. Biophys. Res. Commun. 142: 454-460, 1987. Tanaka, H., Y. Okada, M. Kawagishi, and T. Tokiwa. Pharmacokinetics and pharmacodynamics of recombinant human granulocyte-colony stimulating factor after intravenous and subcutaneous administration in the rat. J. Pharmacol. Exp. Ther. 251: 1199-1203, 1989. Thommasen, H. V., B. A. Martin, B. R. Wiggs, M. Quiroga, E. M. Baile, and J. C. Hogg. Effect of pulmonary blood flow on leukocyte uptake and release by dog lung. J. Appl. Physiol. 56: 966-974, 1984. Wang, J. M., 2. G. Cheng, S. Colella, M. Bonilla, K. Welte, C. Bordignon, and A. Mantovani. Chemotactic activity of recombinant human granulocyte colony-stimulating factor. Blood 72: 1456-1460, 1988. Weisbart, R. H., and D. W. Golde. Physiology of granulocyte and macrophage colony-stimulating factors in host defense. Hematol. Oncol. 3: 401-409, 1989. Welte, K., M. A. Bonilla, A. P. Gillio, T. C. Boone, G. K. Potter, J. L. Gabrilore, M. A. S. Moore, R. J. O’Reilly, and L. M. Souza. Recombinant human granulocyte colony-stimulating factor. Effects on hematopoiesis in normal and cyclophosphamide-treated primates. J. Exp. Med. 165: 941-948, 1987. Wright, S. D., R. A. Ramos, A. Hermanowski-Vosatka, P. Rockwell, and P. A. Detmers. Activation of the adhesive capacity of CR3 on neutrophils by endotoxin: dependence on lipopolysaccharide binding protein and CD14. J. Exp. Med. 173: 1281-1286, 1991.

Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.237.035.237) on January 20, 2019.