| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Divisions of 1Pulmonary/Critical Care Medicine and 2Cardiology, The Burns & Allen Research Institute, Cedars-Sinai Medical Center, Los Angeles 90048; and 3The David Geffen School of Medicine, University of California, Los Angeles, California 90095
Submitted 4 October 2002 ; accepted in final form 10 April 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
muscle protein turnover; autocrine/paracrine action of IGF-I; diaphragm muscle atrophy; nutritional deprivation; cachexia
The aim of the present study was to evaluate mRNA and protein expression of IGF-I in the costal diaphragm muscle of rats subjected to prolonged periods of malnutrition. In addition, two different levels of malnutrition were compared over the same time period. There is a strong rationale for pursuing such experimental studies. Further understanding of local muscle IGF-I responses to malnutrition are important, because a number of anabolic therapies currently being evaluated may exert their positive effects in muscle by upregulating local IGF-I expression within muscle fibers (2, 26). Furthermore, better understanding of local muscle IGF-I responses to prolonged malnutrition may help better define pathobiological responses to assist in developing more downstream targets with therapeutic potential. We postulated that prolonged nutritional insults severe enough to induce fiber atrophy would result in graded reductions in muscle IGF-I expression. We further postulate that muscle IGF-I mRNA abundance would be preserved, as reported by Gautsch et al. (14), as an adaptive mechanism in an attempt to counteract the negative influences on muscle protein turnover associated with prolonged malnutrition.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
The studies were performed on adult female Sprague-Dawley rats with initial
body weights of 
275 g. Female rats were used because of their relative
stability in body weight compared with male animals. A purified powdered
balanced rodent diet (AIN 93G, Dyets) was manipulated to provide either 60 or
40% of normal intake over a period of 3 wk. The rationale for this is
that there appears to be a critical threshold of energy provision (60% of
normal) above which minimal decrements in serum IGF-I have been reported
(38). These nutritional
paradigms would be expected to produce models of mild or moderate
undernutrition. The duration of the nutritional protocol would also represent
a model of sub-acute/early chronic malnutrition. We have observed no
significant decrement in muscle mineral content or serum electrolytes and
glucose with similar models of undernutrition (Lewis, unpublished data) and
would anticipate no significant vitamin deficiency. In control (Ctl) animals,
the diet was provided ab libitum. In all animals, water was provided ab
libitum. All animals were housed individually in a thermally controlled
environment (22°C) with the 12:12-h light-dark cycle. The experimental
protocol was approved by the Animal Use and Care Committee and the
Cedars-Sinai Medical Center/Burns and Allen Research Institute. The animal
groups were 1) Ctl (n = 8); 2) nutritional
deprivation (ND) with provision of 60% of normal intake (ND60; n =
8); and 3) ND with provision of 40% of normal intake (ND40;
n = 9).
Histochemical Procedures
The entire costal region of both hemidiaphragms was excised and weighed. A small portion of the midcostal region on the right side was used for histochemical studies, and the remaining tissue from this side was used for protein analyses. Midcostal diaphragm muscle segments were stretched to estimated optimal length, mounted on cork, and then rapidly frozen in isopentane, which had been cooled to its melting point by liquid nitrogen. Serial cross sections of the diaphragm segments were cut at 10-µm thickness, by using a cryostat (model 2800E, Reichert-Jung, Nussloch, Germany) kept at -20°C.
Costal diaphragm muscle fibers were classified on the basis of difference in staining intensity for myofibrillar adenosine triphosphatase after alkaline (pH = 9.0) and acid (pH = 4.3 and 4.55) preincubations (5). One additional serial section was fixed in 2% paraformaldehyde at pH = 7.4 for 2 min at room temperature and then preincubated at pH = 10.4 [modification (25) of the method used by Guth and Samaha (17)]. These procedures allow classification of fibers into types I, IIa, IIb, and IIx (16). Fiber type proportions were determined from a sample of 200–300 fibers from each muscle.
Costal diaphragm muscle fiber cross-sectional area (CSA) was determined from microscopic images of digitized muscle sections, by using a computer-based imaging-processing system. The latter is composed of a Leitz Laborlux microscope S (Leica, Deerfield, IL), charge-coupled device video camera system (model VI-470, Optronics Engineering, Goleta, CA), high-resolution Trinitron color video monitor (model PVM-1343MD, Sony, Ichiomiya, Japan), 486 DX 50-MHz personal computer with a Targa+ imaging board (Truevision, Indianapolis, IN), and Mocha image-analysis software (version 1.20, Jandel, San Rafael, CA). A microscope stage micrometer was used to calibrate the imaging system for morphometry. The CSA of individual fibers was determined from the number of pixels within outlined fiber boundaries. The estimated relative contribution of each fiber type to total muscle CSA was determined by multiplying the fiber proportion by the mean fiber CSA for each type and expressed as a percentage of the total costal diaphragm area.
Protein Studies: IGF-I
Serum IGF-I. Serum total IGF-I concentration was determined by RIA from blood samples obtained at the terminal experiments. Before RIA, IGF-I was extracted from the serum and IGFBPs were precipitated by incubation in acidethanol (7). Supernatants were neutralized and the RIA (23) performed with a commercial kit specific for rat IGF-I (DSL-2900, Diagnostic Systems Laboratories, Webster, TX) according to manufacturer's protocol. The intra-assay coefficient of variation is 5.9%, and the interassay coefficient of variation is 9.7%. The sensitivity of the assay allows the detection of IGF-I peptide levels of >21 ng/ml.
Diaphragm muscle IGF-I. Frozen right costal diaphragm tissue was pulverized in liquid nitrogen, IGF-I was extracted twice in acetic acid (1 mg/10 µl), and 100-µl aliquots of the supernatant were lyophilized and stored frozen overnight (9). Aliquots were resuspended in 50 µl of assay buffer and assayed with the same RIA kit, as described above, for the determination of muscle IGF-I. Protein concentration was determined by using a commercial protein assay kit (Bio-Rad, Hercules, CA) on the basis of the Bradford (4) method and measured with a spectrophotometer (SmartSpec 3000, Bio-Rad).
mRNA Studies: IGF-I
Total RNA extraction. Total RNA was extracted from 50-mg samples of the left costal diaphragm by using an RNeasy kit (QIAGEN, Valencia, CA) according to manufacturer's protocol. Quality and concentrations of total RNA were determined with a spectrophotometer (SmartSpec 3000). Samples were stored at -80°C until analysis.
Oligonucleotides. The primers for IGF-I and 
-actin were
designed on the basis of published rat cDNA sequences. Primer sequences for
IGF-I (41; GenBank accession
no. X06107
[GenBank]
) were the following: upstream (5' to 3') AAG CCT ACA
AAG TCA GCT CG (bp 595–614) and downstream (5' to 3') GGT
CTT GTT TCC TGC ACT TC (bp 760–741). Primer sequences for -actin
(37; GenBank accession no.
V01217
[GenBank]
) were the following: upstream (5' to 3') TGA CGT TGA CAT
CCG TAA AG (bp 2,737–2,756) and downstream (5' to 3') ACA
GTG AGG CCA GGA TAG AG (bp 3,054–3,034). The expected lengths of the
RT-PCR products were 114 bp for IGF-I and 194 bp for -actin.
-Actin is a valid housekeeping gene because it is not affected in
catabolic states such as malnutrition and corticosteroid treatments. We have
previously verified the stability of -actin against 18S RNA by Northern
analysis (Lewis, unpublished results).
Semiquantitative RT-PCR. One microgram of total RNA was reverse
transcribed by using random hexamers (Perkin-Elmer) and MMLV-RTase (Life
Technologies) and reactions yielded 20 µl of cDNA. RT-generated cDNA for
both IGF-I and -actin were amplified by using PCR (MJ Research Thermal
cycler) with the following experimental conditions: initial denaturation at
95°C for 4 min followed by 27 cycles for IGF-I or 25 cycles for
-actin (95°C for 1 min, 63°C for 45 s, and 72°C for 2 min).
Ten microliters from each PCR product were loaded on 4% agarose gels and
electrophoresed for separation by using ethidium bromide for visualization
under ultraviolet light. The relative amounts of the PCR products were
measured by densitometry (Kodak Electrophoresis Documentation and Analysis
System 120).
Statistical Analysis
The distribution of all data was tested for normality. Statistical analysis
was then performed by using an ANOVA, with the experimental factors being ND60
and ND40. If a significant interaction was found, post hoc analysis
(Newman-Keuls test) was used to compare differences in independent groups. An
-level of 0.05 was used to determine overall significance. Values
reported are means ± SE.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
The ND protocols resulted in a steady progressive loss of body weight over the 23-day experimental period in both ND groups, with resultant end points being a 16 and 26% reduction in body weight (Fig. 1). By contrast, Ctl animals had a net gain of 5% over the same time frame (Fig. 1). Costal diaphragm muscle weights were significantly reduced in both ND groups compared with Ctl (Fig. 2; P < 0.001). Furthermore, a further reduction in costal diaphragm weight was evident in ND40 (-17%) animals, compared with the ND60 (-27%) group (Fig. 2; P = 0.02).
|
|
Diaphragm Fiber Proportions and CSAs
Both levels of ND had no impact on the proportion of costal diaphragm fibers compared with Ctl animals (Fig. 3A). In ND60 animals, there was a significant reduction in the CSA of both types IIa (-19%) and IIx (-23%) costal diaphragm fibers compared with Ctl (Fig. 3B; P < 0.05) with a trend for the reduction of type I (-16%) CSA (P = 0.059). Atrophy of types I (-24%) (P < 0.02), IIa (-19%) (P < 0.05) and IIx (-6%) (P < 0.01) fibers was evident in ND40 animals compared with Ctl rats (Fig. 3B). In ND40 rats, the degree of atrophy was greater for type IIx costal diaphragm fibers than in types IIa and I (Fig. 3B). As a result of variances in CSA for fiber types within each ND group, no significant differences were observed between ND60 and ND40 animals. However, the estimated relative contribution of type IIx fibers to total costal diaphragm area in ND40 rats (46.6 ± 1.6%) was significantly reduced compared with Ctl animals (52.5 ± 2.2%) (P = 0.03), whereas no significant differences for type IIx fibers contribution to total diaphragm area were observed in ND60 rats.
|
IGF-I Studies
Serum IGF-I. Mean serum IGF-I levels were significantly reduced in both ND60 (-62%) and ND40 (-79%) rats compared with Ctl animals (P < 0.0001; Fig. 4A). Furthermore, the reduction in serum IGF-I of ND40 rats was also significantly greater than that observed in ND60 animals (P < 0.01; Fig. 4A).
|
Muscle IGF-I. IGF-I concentration (ng/g muscle tissue) in the
costal diaphragm was reduced by 36% in ND60 rats (P < 0.001;
Fig. 4B) and by
42% in the ND40 group (P < 0.001;
Fig. 4B), compared
with Ctl animals. The reduction in IGF-I levels in the ND40 costal diaphragm
was also greater than those observed in ND60 animals (P = 0.05).
Muscle protein concentrations were not different between groups (Ctl: 227.7
± 30.8 mg/g; ND60: 234.5 ± 21.9 mg/g; ND40: 207.5 ± 20.5
mg/g). When muscle IGF-I concentrations were normalized for protein content,
the results were the same as those normalized per gram of muscle tissue (i.e.,
significant reductions noted in ND groups). Of interest, in ND animals the
relative reduction in serum IGF-I (Fig.
4A; ND60: -62% and ND40: -79%) was far greater than the
reduction in muscle IGF-I (Fig.
4B; ND60: -36% and ND40: -42%).
Muscle IGF-I mRNA. No significant differences in costal diaphragm muscle IGF-I mRNA abundance were observed between the groups (Fig. 5).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Muscle Protein Turnover
Protein is an integral part of all tissues with the quality and quantity of tissue proteins determined by a remodeling process (turnover) whereby protein stores are continually undergoing removal of old proteins (degradation) and replacement with new ones (synthesis) (36). Malnutrition is associated with disordered muscle protein turnover characterized by impaired muscle protein synthesis with or without enhanced muscle protein degradation (e.g., Refs. 15, 32, 45). With malnutrition, reduction in muscle protein synthesis is related in large part to diminished rates of translation initiation of muscle proteins (40, 45), whereas enhanced protein degradation of muscle proteins, as may be seen in acute severe malnutrition, is likely mediated mainly by the ubiquitin-proteasome proteolytic pathway (18). The pathways or signals mediating disordered muscle protein turnover with malnutrition are not fully known, but alteration in the GH-IGF-I axis is a major contributor. Other contributing factors include reduction in insulin and amino acid pool (and possibly increased endogenous corticosteroids with acute severe undernutrition).
Malnutrition: GH-IGF-I Axis
IGF-I is a polypeptide growth factor through which GH may mediate effects on protein metabolism and growth. IGF-I promotes anabolism by increasing protein synthesis and reducing protein degradation (21). Recently, it has become appreciated that IGF-I is produced in most body tissues, including skeletal muscle, implying important local autocrine/paracrine effects separate from those due to the circulating form produced by the liver (24). Recent gene targeting studies supporting this concept include the liver-specific IGF-I gene-deleted mouse model (43, 49). In these studies, normal growth and muscle mass were observed in liver-specific IGF-I-deficient mice in which circulating levels of IGF-I are reduced by 75–80% compared with wild-type animals (43, 49). By contrast, in the full IGF-I gene-deleted mouse model reported by us and others (12, 33, 39) in which local tissue IGF-I was absent (including muscle), markedly impaired growth (i.e., body weight) together with reduced muscle mass (reduced fiber size and number) was noted in adult animals. Thus it has been proposed that the growth-promoting properties of IGF-I are mainly due to autocrine/paracrine influences, whereas circulating IGF-I has an important metabolic role in glucose homeostasis with IGF-I functioning as an insulin sensitizer (6). In the wild-type animal, there is other evidence supporting a dynamic local IGF-I milieu in muscle. This includes the increase in IGF-I mRNA abundance in muscle with GH administration (19) and the muscle expression of a variety of IGFBPs, which may modulate the ability of muscle IGF-I to bind to its receptor in an autocrine/paracrine fashion (21, 26).
With severe nutritional deprivation, peripheral resistance to the action of GH has been well documented, and circulating levels of IGF-I and IGFBP-3 are reduced (28, 38, 47). However, short-term nutritional deprivation appears to have, in addition, a distinct impact on the IGF system in skeletal muscle. This is supported by studies reporting reduced abundance of IGF-I mRNA in the skeletal muscles of rats subjected to fasting or short-term semistarvation or protein restriction (34, 45, 46, 48). We have also observed reduced diaphragm muscle IGF-I mRNA after 4 days of severe nutritional deprivation (Lewis et al., unpublished data). These reductions in muscle IGF-I mRNA abundance would be expected to influence the anabolic action of muscle IGF-I in vivo due to reduced ligand availability. In one study (34), protein restriction resulted in increased IGF-I binding in some tissues as an adaptive response.
In the present study, the effects of long-term malnutrition of varying severity revealed some interesting data on responses of the IGF system in the costal diaphragm muscle. In contrast to the short-term studies, muscle IGF-I mRNA abundance with prolonged malnutrition was preserved. This was also reported by Gautsch et al. (14) in rats subjected to prolonged malnutrition for either 50 or 120 days. We postulate that preservation of muscle IGF-I mRNA reflects an adaptive response in an attempt to preserve local muscle IGF-I availability and in vivo muscle IGF-I action. The stimulus for this is unclear. In the liver-specific IGF-I gene-deleted mouse model circulating GH levels were elevated, which could theoretically increase muscle IGF-I mRNA levels (43, 49). However, our laboratory and others have shown that with severe malnutrition, peripheral resistance to the action of GH is present, including muscle tissue (28, 47). However, our laboratory observed no resistance to GH action in rats restricted to 50% of normal food intake, a midpoint of the range used in an earlier study (25). Gautsch et al., in their paper on prolonged malnutrition, concluded that muscle atrophy could not be attributable to decreased local autocrine/paracrine production of IGF-I, because skeletal muscle IGF-I mRNA levels were preserved. However, muscle IGF-I protein assays were not performed. Indeed the present study demonstrated significantly reduced protein levels of IGF-I in diaphragm muscle in a graded fashion depending on the severity of the nutritional insult. This is best explained by an impairment of mRNA translation of IGF-I, as has been reported for muscle proteins in cachectic states, including semistarvation (45). Alterations in turnover rate of IGF-I may also contribute to its reduced expression. Because proteolysis tends to be curtailed over time with malnutrition (15), the balance would likely be shifted to reduced synthesis. We propose that the reduced expression of IGF-I in the costal diaphragm muscle would serve to perturb the turnover of muscle proteins, including key contractile proteins such as myosin heavy chains. It is also of interest to highlight other aspects of our data. First, in ND animals, the relative reduction in serum IGF-I was far greater than the reduction in muscle IGF-I. Second, the significantly greater reduction in serum IGF-I levels in ND40 animals compared with ND60 rats was not mirrored by a proportionate reduction in muscle IGF-I in the ND40 group, suggesting differing regulatory mechanisms for circulating and local IGF-I. Lastly, in ND60 animals, the relative levels of circulating IGF-I were twice those reported in liver-specific IGF-I gene-deleted mouse model, yet in our models, body weight loss and muscle atrophy were evident.
Malnutrition: Muscle Fiber Atrophy
In the present study, a graded increase in loss of muscle mass and fiber size was observed with more severe levels of nutritional deprivation. This is supported by a greater reduction in the estimated relative contribution of type IIx fibers to total costal diaphragm area in ND40 animals, compared with the Ctl group, in which type IIx fibers make up the greatest contribution (more than half of total diaphragm CSA). As highlighted above, muscle IGF-I is a major factor responsible for muscle growth and maintenance of fiber size (12). We thus propose that a reduction in muscle IGF-I is an important contributing factor for muscle wasting with ND. It has been shown that the administration of exogenous IGF-I significantly stimulates muscle protein synthesis in the rat hindlimb in a model of semistarvation (45), whereas IGF-I infusion in adolescent rats subjected to short-term moderate undernutrition completely prevented diaphragm muscle fiber atrophy (25). It is of interest, therefore, that in transgenic mice exhibiting low circulating levels of IGF-I (but preserved muscle IGF-I), skeletal muscle fractional synthesis rate with fasting was preserved, suggesting that muscle IGF-I contributed to muscle protein synthesis (46). Thus reduced muscle levels of IGF-I, as noted in the present study, would be expected to negatively impact on the synthesis of muscle proteins. With severe short-term malnutrition, our laboratory observed significant impacts on various signaling pathways downstream of the IGF-I receptor (27). However, no such data exist with prolonged malnutrition. With nutritional deprivation, serum levels of amino acids and insulin are reduced (45, 46), which may compound the problem of disordered muscle protein turnover in addition to the adverse sequelae of low tissue IGF-I levels. Indeed, both insulin and amino acids are thought to exert permissive roles in augmenting IGF-I stimulated muscle protein synthesis (20). Thus reduced substrate availability and/or insulin levels likely further reduce the ability of muscle IGF-I to prevent muscle cachexia with prolonged nutritional insults (20). Curtailment of muscle protein breakdown has been reported as an adaptive process over time with severe short-term ND (15), and hence we propose that muscle proteolysis is not an important factor in our model.
Clinical Implications and Future Perspectives
Reduced muscle IGF-I expression associated with muscle fiber atrophy has a number of implications for future research efforts geared to limit muscle wasting in states of disordered protein turnover. First, increasing substrate availability (e.g., amino acids, either total, essential, or key individual) would likely be important. Second, some anabolic agents, such as anabolic steroids and clenbuterol, may exert their effects by increasing muscle expression of IGF-I (2, 30). This, however, leads to the question of efficacy of these agents in states of severe prolonged malnutrition, as the catabolic milieu as presented in this paper may limit the ability of these agents to augment muscle IGF-I. This appeared to be the case with the administration of nandrolone with prolonged severe malnutrition in emphysematous hamsters (13). Third, other future approaches to promote synthesis of muscle proteins with significant malnutrition would be to augment those signaling pathways distal to the IGF-I receptor that promote muscle protein synthesis (e.g., phosphatidylinositol 3-kinase/Akt) (3). Such approaches could involve both pharmacologic manipulation and even future genetic approaches.
In summary, costal diaphragm muscle autocrine/paracrine expression of IGF-I is significantly curtailed with prolonged malnutrition of varying severity. We propose that this contributes to disordered muscle protein turnover and muscle cachexia with atrophy of muscle fibers. Such influences on the respiratory muscles would reduce force-generating capacity and functional force reserve, leading to task failure under conditions of increased demand.
|
DISCLOSURES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
and diaphragm fiber atrophy with
malnutrition (Abstract). Am J Respir Crit Care Med
165: A258,
2002.
This article has been cited by other articles:
|
|
 |
|
  M. I. Lewis Mechanisms underlying myogenesis: complex and likely to become more so! J Appl Physiol, December 1, 2006; 101(6): 1539 - 1540. [Full Text] [PDF]
|
|
|
|
|
|
M. I. Lewis, S. C. Bodine, N. Kamangar, X. Xu, X. Da, and M. Fournier Effect of severe short-term malnutrition on diaphragm muscle signal transduction pathways influencing protein turnover J Appl Physiol, June 1, 2006; 100(6): 1799 - 1806. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. I. Lewis, X. Da, H. Li, and M. Fournier Tumor necrosis factor-{alpha} and malnutrition-induced inhibition of diaphragm fiber growth in young rats J Appl Physiol, November 1, 2005; 99(5): 1649 - 1657. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. N. Mehiri, E. Barreiro, M. Hayot, M. Voyer, A. S. Comtois, A. E. Grassino, and G. Czaika Time-based gene expression programme following diaphragm injury in a rat model Eur. Respir. J., March 1, 2005; 25(3): 422 - 430. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Dias, C. P. Passaro, V. R. Cagido, M. Einicker-Lamas, J. Lowe, E. M. Negri, V. L. Capelozzi, W. A. Zin, and P. R. M. Rocco Effects of undernutrition on respiratory mechanics and lung parenchyma remodeling J Appl Physiol, November 1, 2004; 97(5): 1888 - 1896. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. I. Lewis, M. Fournier, X. Da, H. Li, Z. Mosenifar, R. J. McKenna Jr, and A. H. Cohen Short-term Influences of Lung Volume Reduction Surgery on the Diaphragm in Emphysematous Hamsters Am. J. Respir. Crit. Care Med., October 1, 2004; 170(7): 753 - 759. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. F. Klawitter and T. L. Clanton Tension-time index, fatigue, and energetics in isolated rat diaphragm: a new experimental model J Appl Physiol, January 1, 2004; 96(1): 89 - 95. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
