Quantification of Protein Metabolism in Vivo for Skin, Wound, and Muscle in Severe Burn Patients
From the * Department of Surgery, The University
of Texas Medical Branch, Galveston, Texas; and Correspondence: Dennis C. Gore, MD, Department of Surgery, The University of Texas Medical Branch, 301 University Boulevard, Galveston, TX 77555-1172. Electronic mail may be sent to dcgore{at}utmb.edu. Background: In response to injury, muscle catabolism can be extensive, and in theory, the wound consumes amino acids to support healing. The purpose of this study is to assess a technique by which in vivo protein kinetics of muscle, wound, and normal skin can be quantified in burn-injured patients. Methods: Study protocol consisting of infusion of d5 phenylalanine; biopsies of skeletal muscle, skin, and donor-site wound on the leg; quantification of blood flow to total leg, wound, and skin; and sequential blood sampling from the femoral artery and vein. Five-compartment modeling was used to quantify the rates of protein synthesis, breakdown, and phenylalanine transport between muscle, wound, and skin. Results: The study results demonstrated a net release of phenylalanine from muscle yet a net consumption of phenylalanine by the wound. Compared with skin, the wound had a substantially increased rate of protein synthesis and a reduced rate of protein breakdown (p < .01). Transport rates into and out of muscle were significantly higher than those for wound (p < .01). Conclusions: This novel methodology enables in vivo quantification of the integrated response of muscle, wound, and skin protein/amino acid metabolism and confirms the long-held theory of a net catabolism of muscle and a net anabolism of wound protein in patients after injury. This methodology can be used to assess the metabolic impact of such measures as nutrition, pharmacologic agents, and surgical procedures. Injury elicits a demonstrable alteration in protein kinetics.1,2 From muscle, amino acids are released at a greatly accelerated rate. In theory, injury elicits this acceleration in muscle protein breakdown to provide abundant amino acids, which supplies the required precursors to support protein synthesis for wound healing. Assuming a metabolic link between the loss of muscle protein and the deposition of protein in the wound, then factors that impede net muscle breakdown might be expected to delay wound healing. In conflict to this assumption, numerous studies have demonstrated that extensive loss of lean body mass correlates with an impairment in immune function, delays in wound healing, and an increase in mortality.3,4 Furthermore, evidence suggest that pharmaceutic agents that are anabolic for muscle accelerate wound healing.5 Therefore, this concept that muscle releases amino acids from storage in support of wound healing may be in error. To address this metabolic dualism between muscle and wound protein kinetics, quantitative measures are needed. Whereas the methodology is well established for assessing the protein kinetic response of muscle,6 experiments quantifying the rates of protein synthesis and breakdown within wound and skin have been limited to animal studies.7 The simultaneous assessment of protein kinetics of muscle, skin, and wound in vivo in humans has not previously been reported. The purpose of this study was to develop and assess a technique in which the in vivo rates of protein synthesis, degradation, and amino acid transport could be determined simultaneously for skin, wound, and muscle in human patients. Skeletal muscle metabolism can be quantified by the infusion of an isotopically labeled amino acid as a tracer combined with the measurement of leg blood flow, sequential sampling of blood from the femoral artery and vein, and periodic biopsy of muscle.6 When an amino acid tracer is used, such as phenylalanine that is neither produced nor oxidized in muscle, this technique allows quantification of not only the rates of muscle protein synthesis and breakdown but also the transport rates of the tracer amino acid into and out of muscle. This methodology assumes that any difference in amino acid balance from femoral artery and vein is related predominantly to the consumption and/or release of that amino acid from muscle. This technique also assumes that the entire blood flow to the leg goes to muscle. However, the net balance of an amino acid across the leg also reflects in part the protein metabolism and blood flow to skin and any wound on the leg. To assess the relative contributions of skin and muscle to arterial-venous balance, Biolo et al8 assumed a model in which these tissues are arranged in parallel with separate blood supplies. By including sequential biopsies of skin to the traditional methodology for determining muscle protein metabolism and assuming the fractional distribution of total leg blood flow to skin and muscle as previously reported, Biolo et al8 calculated that in healthy dogs, skin protein synthesis and degradation accounts for approximately 10%–15% the total leg protein kinetics. We have expanded on this initial report by Biolo et al8 by measuring blood flow to the skin and wound using laser Doppler flowmetry and quantifying the actual volumes of skin, muscle, and wound using dual-energy x-ray absorptiometry and physical measures in human subjects having recently experienced a severe burn. With these direct measurements and using calculations for a 5-compartment model of the leg (ie, artery, vein, muscle, skin, and wound), the in vivo protein kinetics for skin, wound, and muscle were quantified in severely injured human subjects.
Study Subjects A total of 10 patients were recruited for study, of which 2 subjects were women. All patients had burns encompassing over 40% of their total body surface and were admitted to the burn intensive care unit within 2 days of their injury. Subjects ranged in age from 23 to 52 years (mean, 34 ± 7) and in weight from 54 to 130 kg (mean, 85 ± 12). All study subjects had good hemodynamics and adequate urine output on recruitment yet were hypermetabolic, as evidenced by a 41% greater resting energy expenditure above normal and mild tachycardia (mean heart rate 108 ± 12). Seven study subjects required mechanical ventilation at the time of study. Patients were excluded from participation if they were diabetic before their burn injury, had renal and liver dysfunction (serum creatinine >1.5 mg/dL, bilirubin >3.0 mg/dL), were hypoxic (PaO2/FiO2 <150) or had lactic acidosis (pH <7.30, lactate >5 mmol/L). All patients were treated in a similar manner clinically, with early excision of their burn wound and prompt skin grafting. The lower extremity that was subsequently used for metabolic analysis was burned in all subjects, ranging in area from 17% to 72% of the leg. Before study, all full-thickness burn had been excised from the leg, and the majority of any unburned skin on the leg was harvested for autologous skin grafting using a dermatome setting at 0.012 in. (Zimmer Inc, Warsaw, IN). Thereby, approximately 80%–85% of the leg was encompassed by wound for each subject at the time of study. For wound care, meshed autologous skin grafts were placed over areas debrided of burn eschar and then covered with gauze saturated with Polysporin Mycostatin ointment (Bristol-Myers Squibb, New York, NY). Skin-graft donor sites were covered with Scarlet Red dressing. Nutrition support was exclusively enteral (Vivonex TEN, Norvatis Corp, Summit, NJ), initiated within 48 hours after burn injury and delivered continuously via a nasal-jejunal tube. All subjects were receiving enteral nutrition at the time of study at a desired caloric goal of 1.2 times their resting energy expenditure, as determined by previous indirect calorimetry.
Study Protocol This study was approved by the institutional review board of The University of Texas Medical Branch, Galveston, TX. Informed written consent was obtained from the patient and/or a family member before study. All subjects underwent study, as diagramed in Figure 1, between 4 and 9 days after their burn injury and between 2 and 4 days after their operative debridement. Local anesthesia (1% lidocaine) facilitated placement of vascular access catheters into an adjacent femoral artery and vein. Localized blood flow was determined by laser Doppler flowmetry (Laser Flow DPM II, Vasamedics Inc, St. Paul, MN) over normal-appearing skin and over a skin-graft donor-site wound on either the anterior or lateral thigh. Measurements were reported at 1-second intervals and values recorded for 6–8 minutes, then averaged. Care was taken not to press firmly with the probe against the underlying tissue surface in attempt to minimize any decrease in blood flow related to external pressure. The surface area of the leg was determined by integrating several circumference measurements with the length of the calf, thigh, and foot. The volume of leg as skin and wound was then calculated from the relative surface area of each multiplied times the average thickness of 0.27 cm, as previously reported.9 Blood flow measurements of skin and wound were then standardized per their respective tissue volumes. After these blood flow measurements, punch biopsies of normal skin and wound were obtained under local anesthesia (1% lidocaine). These biopsies were immediately frozen in liquid nitrogen. Wound samples included a heterogeneity of tissues (ie, dermis, granulation, neoepidermis) of undetermined proportion. After baseline blood, skin, and wound sampling, a primed continuous infusion of d5 phenylalanine (2 µmol/kg prime; 0.07 µmol/kg/min infusion) was given via central venous access and continued for 5 hours. Total leg blood flow was determined by infusion of indocyanine green dye (ICG, 1 mg/min x 20 minutes) into the femoral artery, with subsequent spectrophotometric analysis of blood drawn simultaneously from the femoral vein and from the central vein access. Total leg blood flow measurements were standardized for leg volume as assessed by the same circumference and length measurements used for determination of leg surface area.10 After infusion of isotopic phenylalanine for 2 hours and immediately after skin and deep infiltration with local anesthesia, a skeletal muscle biopsy was obtained from the Vastus Lateralis muscle using a Bergstrom needle. After 4 hours and 30 minutes, 4 hours and 45 minutes, and 5 hours of isotopic infusion, blood was drawn simultaneously from the femoral artery and vein. After 5 hours of d5 phenylalanine infusion, laser Doppler flowmetry followed by punch biopsies was obtained again from normal-appearing skin and from a wound near sites sampled at the initiation of the study protocol. At 5 hours, repeated percutaneous biopsy of Vastus Lateralis muscle was performed. The isotopic infusion of d5 phenylalanine was then discontinued. After completion of the study protocol, each patient underwent dual-energy x-ray absorptiometry DEXA (Hologic QDR-4500A, Waltham, MA) thereby quantifying muscle mass within the studied leg.
Sample Analysis Biopsies of skin, wound, and muscle were analyzed for both intracellular and protein-bound concentration and isotopic enrichment of phenylalanine using an internal standard method with analysis by GCMS.12 To approximately 20 mg of tissue, 800 µL of 14% perchloric acid and 2 µL of internal standard were added. Samples were homogenized and centrifuged, and the supernatant was collected. This procedure was repeated twice more and the pooled supernatant processed identically to the blood samples above using tert-butyl dimethylsialyl derivatization with analysis by GCMS. This method determined the free intracellular concentration and isotopic enrichment of phenylalanine. The remaining tissue pellet was washed repeatedly with saline and absolute ethanol, dried, and then hydrolyzed with 6-N HC1. The protein hydrolysate was then passed over a cation exchange column (Dowex AG; Bio-Rad Laboratories, Richmond, CA), dried, esterified, heated, and subsequently analyzed by GCMS for determination of protein-bound concentration and enrichment, as previously described.8
Calculations
Phenylalanine was chosen as the amino acid tracer because it is an essential amino acid and oxidized only in the liver.13 Therefore, isotopic phenylalanine provides an index for protein kinetics of muscle, skin, and wound within the leg. The value of blood flow to muscle within the leg of severe burn patients is cited from previous work by Aulick et al14 using 133 xenon clearance.
The fractional synthesis rates of muscle protein (FSRm), skin protein
(FSRs) and wound protein (FSRw) were calculated as
follows15:
Em = average intracellular enrichment for the tissue in question over time t Factors 60 and 100 are required to express FSR in percentage per hour. Muscle, skin, and wound protein kinetics are calculated using equations derived from arterial-venous balance tracer/tracee modeling combined with measurements of the rate of tracer amino acid incorporation into tissue (ie, precursor-product technique).11 This model is pictured in Figure 2 and uses 5 assumptions. One assumption is that the rate at which tracer and/or tracee enters each compartment is equal to the rate at which tracer and/or tracee exits each compartment. The Fick principle supports the validity of this assumption.11 We also assumed that the rate of protein synthesis for each tissue compartment (ie, muscle, skin and wound) is equal to the fractional synthetic rate multiplied times the total mass of that specific tissue within the leg, as indexed by the amount of phenylalanine bound within protein for that tissue compartment.16
Therefore:
The third assumption of this model is that the study is performed during steady-state condition. Diligence was taken to minimize any perturbations in care during the 5 hours of study, with nutrition support provided as a continuous administration and no changes in ventilatory support, fluid administration, or wound dressings. Several previous publications have demonstrated that a 5-hour infusion of d5 phenylalanine is adequate for achieving isotopic steady-state kinetics in burn patients.6,13 This isotopic steady state condition was also verified by quantifying consistent enrichment in serial blood samples obtained from the femoral artery and vein during the final 30 minutes of study. A fourth assumption ascribes that the rate of inward transport of amino acids into skin and wound is relative to the respective blood flows of each tissue. Prior work by Ling et al16 and Flores et al17 using radioactive microspheres to determine tissue perfusion has shown that the exchange of amino acid tracer in plasma with the free intracellular space is dependent on tissue perfusion. This finding supports the validity of this fourth assumption.
Furthermore, this model assumes that the proportion of amino acids
extracted by skin and wound is of similar magnitude, whereas the proportion of
amino acids entering into muscle is 3 times greater than that extracted by
either wound or skin. The validity of this assumption is based on studies
performed in our laboratory by Zhang et
al18 using the ear
of a rabbit, which is devoid of muscle and thus composed predominantly of skin
and cartilage with or without a surface wound. This prior work evidenced that
the extraction of phenylalanine into skin or wound was approximately 17%. By
comparison, whole-leg studies in which the predominant tissue mass is muscle
have generally found an extraction rate of phenylalanine at approximately
50%.6 Thus, 50%
divided by 17% is nearly 3. When normalized for the relative blood flows of
skin, wound, and muscle, then
Under these assumptions, the following parameters are calculated:
The parameters of Fin and Fout are calculated as follows:
Under steady-state conditions, the rate at which amino acids enter the
artery is equal to the rate at which the amino acids are transported into the
intracellular space of muscle, skin, and wound plus the rate at which amino
acids are shunted to the vein.
Furthermore, the rate at which amino acids exit via the femoral
vein is equal to the rate at which amino acids are transported from the
intracellular space of muscle, skin, and wound into the vein plus the rate
that amino acids are shunted to the vein.
Steady-state condition also ascribes that the rate at which amino acids
appear in the intracellular spaces of each compartment via inward
transport and protein breakdown is equal to the rate at which amino acids exit
the intracellular space via outward transport and protein synthesis.
Therefore:
In consideration of the first assumption in which tracer and tracee are
ascribed equal rates of transport, then the rate at which labeled
phenylalanine appears in the intracellular space of each compartment
via inward transport is equal to the rate at which labeled
phenylalanine disappears from the intracellular space via outward
transport and protein synthesis.
Solving these equations for Fvm, Fvs, and Fvw:
Rearranging equation 11 and solving for Fva:
Substituting for equations 19, 20, 21, and 22 into equation 12 and solving
for Fout:
As assumed:
Therefore
Solving this equation for Fsa:
Thus, Fsa, Fwa, Fma, Fva, Fvm, Fvs, Fvw are solved from equations 25, 6, 8, 22, 19, 20, and 21, respectively.
By rearranging equations 13, 14, 15, rate of protein breakdown can be
calculated for each tissue:
Assuming that the net uptake of tracer across the leg is equal to the
incorporation of tracer into muscle, skin, and wound, then:
Therefore, determination of the protein synthesis rate for 1 tissue is not needed. If the synthesis rates for all 3 tissues are measured, then the above equation can be used to assess the accuracy of the measured values and/or estimate the uptake of tracer (ie, synthesis rate) for tissues other than skin, wound, or muscle (ie, bone, fat, or cartilage). Unfortunately, equation 29 cannot be used to determine the actual values for Rws or Rms.
Statistics
All patients were hypermetabolic and hyperdynamic, as evidenced by the >40% elevation in their resting energy expenditure, tachycardia, and the greater than normal blood flow within the study leg (Table I).6 Also noted in Table I are the measures of blood flow to skin, wound, and muscle. Of significance is the increased blood flow to the wound compared with skin. Patients were also catabolic to protein, as evidenced by the negative net balance of phenylalanine from the leg (Table II). Also noted in Table II is the negative net balance of protein from skin and from muscle. In contrast, wound has a positive net balance indicative of net consumption of phenylalanine into the wound tissue.
Fractional synthetic rates for skin, wound, and muscle are presented in Figure 3. Study results demonstrate a greater protein synthetic rate for wound tissue in comparison to skin and muscle. Calculations using the 5-compartment modeling also provide quantification of the rates of protein synthesis for skin, wound, and muscle. When indexed by a set volume for each tissue, this method demonstrates a significantly increased rate of protein synthesis for wound compared with normal skin (Table III). Of the 3 measured tissue compartments, muscle had the slowest rate of protein synthesis. Five-compartment modeling also provided quantification of protein breakdown rates for each tissue. These calculations demonstrated a slightly greater rate of protein breakdown for muscle compared with skin, whereas protein breakdown within the wound was very small. Net balance measurements using this method illustrate a significant net anabolism for protein within the wound and a substantial net catabolism for protein from muscle.
The rates of phenylalanine transport for skin, wound, and muscle are demonstrated in Table IV. These calculations show that when indexed by a set volume of each tissue, the rate of amino acid transport into wound is far greater than muscle, which is in turn only slightly greater than skin. Likewise, transport rates of phenylalanine out of wound are greater than that for muscle, which are greater still than that of skin. These values demonstrate the greater flux of phenylalanine into and out of wound.
Of note, all wound study sites had obvious reepithelialization by 7 days after study, as evident by visual inspection.
This study simultaneously quantifies protein kinetics in vivo in skin, wound, and muscle in patients after severe burn injury. Results demonstrate an extensive net catabolism of muscle postinjury in which the rates of protein breakdown exceed that of protein synthesis despite ongoing nutrition intake. In contrast to muscle, there was a substantial net anabolism within the wound in which the rates of protein synthesis far exceeded the rate of breakdown. Such findings support the longstanding notion that muscle liberates amino acids in response to severe injury, while the wound has a significant consumption of amino acids, in theory to support healing. This study thus illustrates methodology by which the rate of lean body loss can be quantified while simultaneously measuring the rate of protein aggregation within a wound and by inference quantifies healing of that wound. Traditionally, wound healing time is often difficult to quantify in the clinical setting, whereas this new tracer method can be performed in a short time interval and provides quantitative data without the subjectivity inherent in judging wound healing. Thus, an important aspect of this study is that the metabolic influence of different nutrition regimens, various pharmacologic agents, and surgical maneuvers such as the timing of burn eschar removal can now be quantitatively evaluated as to wound and muscle. Our results indicate that the rate of wound protein synthesis far exceeds that of skin. This is an expected result, given the increase in cellular proliferation, reepithelialization, and increased collagen and other protein deposition.19,20 This faster rate of protein synthesis is an integral component for prompt healing of a wound. Surprisingly, the rate of protein degradation in wound was very slow relative to the greatly accelerated rate of synthesis. This suggests a potent mechanism for efficient use of available amino acids for the production of structural proteins. In contrast to wound, the protein flux of skin is slow, thus providing structural stability. This study also demonstrated a faster rate of protein synthesis of skin compared with muscle. Several prior studies support this finding, including work in dogs by Biolo et al8 using similar methodology.21 This study also quantified a negative net balance of phenylalanine from muscle. This finding is consistent with the general theory of muscle as a reservoir for amino acids, which are released after injury or illness in support of functions such as wound healing and immune response. Furthermore, previous measures using similar methodology have noted a significantly lower rate of muscle protein breakdown in healthy volunteers compared with that of burn patients.22 This finding also supports the concept of an injury-mediated release of amino acids from muscle. In this initial work, only isotopic phenylalanine was used as the tracer amino acid. Because phenylalanine is not known to be oxidized or synthesized de novo except in the liver, phenylalanine provides a reliable marker of the overall protein kinetics of tissues within the leg.13 A logical extension from this work is to use amino acids other than phenylalanine as tracers and thus quantify further the metabolic interaction between muscle, skin, and wound in regard to protein and amino acid metabolism. Likewise, a fatty acid tracer could potentially be used to assess the metabolic contribution of adipose tissue within the leg.23 Unfortunately, the procedures involved in this study methodology, as well as the subsequent analysis of tissue and blood samples, are very labor intensive and require sophisticated, expensive equipment. Thus, experience of personnel is important in performing these studies. As noted in the Calculations section, there are 5 basic assumptions underlying the model, namely, (1) the rate at which tracer and tracee enter each compartment is equal to the rate at which tracer and tracee exit each compartment (excluding the bound pools); (2) the rate of incorporation of phenylalanine into bound protein in each tissue is equal to the amount of bound phenylalanine in each tissue times the fractional synthetic rate of that tissue as determined by the direct incorporation method; (3) an isotopic and physiologic steady state exists; (4) the rates of delivery of amino acids to skin and wound capillaries are relative to the respective blood flows of each tissue; and (5) the extraction of amino acids by skin and wound is the same and the extraction of amino acids by muscle is 3 times greater than extraction by skin or wound. Each of these assumptions will be discussed. The first assumption is simply the Fick principle. If this assumption were not true, then tracer or tracee would either accumulate or abate in 1 or more of the compartments. We observed constant enrichments and concentrations in all of the free pools over the 5-hour period of the study, so this assumption is reasonable. The primary problem with the second assumption is that there is still a lingering debate over the proper precursor enrichment for determination of FSR. We have used the respective intracellular free phenylalanine enrichment as the precursor for each of the 3 pools, which we feel is appropriate.24 To ensure that the third assumption holds, we have taken care to minimize the perturbations in the patients by providing nutrition support at a continuous rate and by keeping ventilatory support and fluid administration constant during study. As noted, measured enrichments and concentrations of phenylalanine of plasma from the femoral artery and vein were similar with repeated sampling, thus supporting the assumption of steady-state kinetics. Because the arterial concentrations and enrichments are generally well mixed before entering the femoral artery, the fourth assumption is reasonable. The practical problem with the fourth assumption is measuring the actual blood flow to each of the tissues. In this study, we measured skin and wound blood flow using laser Doppler flowmetry and used a previously published value for muscle blood flow.14 In this case, muscle blood flow amounted to about 50% of the total leg blood flow, as assessed by the ICG infusion method. Measured skin and wound accounted for another 40%, leaving 10% for anatomical shunting. Thus, the assumed muscle blood flow appears reasonable for this study. The fifth assumption is the most troublesome from a theoretical point of view. As noted in the Calculations section, the 3-fold greater inward transport per blood flow for muscle over skin is based on comparison of prior studies, which note the extraction of phenylalanine within the human leg as approximately 50%, whereas extraction within the ear of a rabbit is approximately 17%.6,18 Assuming that the percent extraction of phenylalanine from the skin of a rabbit is equivalent in proportion to the intact normal-appearing skin of a severely burned human, then 50% divided by 17% is nearly 3. Likewise, the equivalent inward transport for skin and wound is suggested from prior work using the isolated rabbit ear model, with no validation in human studies.18 These assumptions have no impact on determination of the rates of protein synthesis, because protein synthesis is determined from FSR. However, the calculation of protein breakdown and net protein balance across each tissue relies on the accuracy of this assumption. One would presume that the relative transport of amino acids in muscle and skin might change in different physiologic states, so protein breakdown or net protein balance may be inaccurately calculated if the same assumed ratio is used in all physiologic conditions. To assess the extent to which this is a practical problem, we examined the effects of changing the relative transport rates (ie, Rms and Rws) on protein kinetics. If one assumes that the relative inward transport for the wound was twice that of skin rather than being the same (ie, Rws = 2 instead of Rws = 1), then the rate of muscle protein breakdown decreases by <1%, with no change in the calculated value for skin protein breakdown. There is with this assumption a 200% increase in the rate of protein breakdown from the wound, yet the net balance of protein from the wound decreases by only 5%. On the other hand, if one assumes that the relative inward transport for muscle is 6 times that of skin rather than 3 times (ie, Rms = 6 rather than Rms = 3), then the rate of protein breakdown from muscle increases by only 2%, with no change in the rate of wound protein breakdown. However, this doubling of the assumed ratio of muscle transport relative to skin does result in a 50% reduction in the breakdown rate for skin and a change in the net protein balance of skin from –0.05 to +0.05 nmol/min 100 g. We caution that the assumed values of Rws and Rms used in this model may vary in different physiologic situations. However, the fact that the relative blood flow to these tissues will often change dramatically will have no direct effect on Rws and Rms, because they rely on the inward transport relative to amino acid delivery. However, the relative extraction might change indirectly as a consequence of a greater or lesser capacity to extract the amino acids that are delivered. Because one would generally encounter situations in which amino acid delivery to all tissues in the leg would deviate in the same direction, the changes in the relative extraction of all tissues will, in general, also deviate in the same direction and thereby minimize alterations in Rms and Rws. Treatments that selectively change inward transport of 1 tissue relative to the other tissues such as an anabolic agent for muscle will also cause potential problems in the interpretation of this model. However, keeping these shortcomings in mind, we believe that this model can be productive in the study of protein metabolism after severe injury. The work is supported by Shriners grant 8590. Received for publication February 2, 2006. Accepted for publication April 6, 2006.
Journal of Parenteral and Enteral Nutrition, Vol. 30, No. 4,
331-338 (2006)
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Top
MATERIALS AND METHODS
RESULTS
