Chronic hepatitis is the most common liver disease in dogs.1 The cause of chronic hepatitis in dogs is often unknown (ie, idiopathic chronic hepatitis). Recognized causes of chronic hepatitis in dogs include hepatic accumulation of copper and drug-induced liver injury. Chronic hepatitis is histologically characterized by hepatocellular apoptosis or necrosis, a variable mono-nuclear or mixed inflammatory infiltrate, regeneration, and fibrosis.1 The prognosis for dogs with chronic hepatitis is guarded, with variable progression to end-stage liver disease and fibrosis. Fibrosis is defined by the overgrowth, hardening, and scarring of various tissues and is attributed to excess deposition of extracellular matrix components, including collagen.2 Efforts have been made to replace examination of liver biopsy specimens with measurement of noninvasive markers of liver fibrosis to aid in making a diagnosis and monitoring the treatment response in dogs with liver disease.
Complications of cirrhosis or end-stage liver disease are the primary causes of death related to chronic hepatic diseases. Progressive liver fibrosis resulting in the development of end-stage liver disease is characteristic of chronic liver disease; thus, the assessment of liver fibrosis can be a predictive factor for liver-related fatalities. The stage of liver fibrosis has been correlated with mortality rates and prognosis in chronic hepatic diseases of dogs3,4 and humans.5 Noninvasive markers of liver fibrosis are desired.
Trans-4-hydroxy-l-proline is a nonproteinogenic, nonessential amino acid found in collagen and other extracellular proteins. Trans-4-hydroxy-l-proline is an important constituent of collagen and is a collagen-specific amino acid, with approximately 1% of the amino acid found in elastin.6 Serum trans-4-hydroxy-l-proline concentration has been correlated with the degree of fibrosis in humans with chronic hepatic disease.7,8
The purpose of the study reported here was to develop and validate a simple, efficient, and reliable LC-MS-MS method for quantitative determination of the endogenous trans-4-hydroxy-l-proline concentration in canine serum. The clinical applicability of the method was assessed by analyzing the trans-4-hydroxy-l-proline concentration in the serum of dogs with chronic hepatitis. We hypothesized that dogs with chronic hepatitis would have higher serum concentrations, compared with concentrations in healthy control dogs.
Materials and Methods
Sample
Blood samples previously collected from 20 dogs with histologically confirmed chronic hepatitis and from 20 healthy control dogs that had been examined at the Texas A&M University Veterinary Teaching Hospital from November 2014 through December 2015 were used for the study. Blood samples were immediately centrifuged at 1,800 × g for 10 minutes at 4°C; serum was harvested and stored at −80°C until analysis. For dogs that underwent laparoscopic liver biopsy or exploratory celiotomy with surgical liver biopsy and removal of bile from the gallbladder as part of standard diagnostic testing, a 1-mL sample of bile and 1 additional liver biopsy specimen were collected for an unrelated unpublished study. Four to 6 liver biopsy specimens were acquired during a typical liver biopsy. Of these, 1 biopsy specimen was placed in a sterile plain, evacuated tube and used for tissue copper quantification, 1 biopsy specimen was placed in a sterile specimen cup and submitted for aerobic and anaerobic bacterial culture and susceptibility testing, and 1 to 3 biopsy specimens were fixed in neutral-buffered formalin, processed for routine histologic examination, and embedded in paraffin. Serial sections of liver tissue were stained with H&E, eosin, or picrosirius reda stains or a combination of these stains. Hepatic fibrosis, inflammation, and hepatocellular vacuolation were scored by a single board-certified veterinary pathologist who was not aware of the clinical data; scores were assigned by use of an established 5-point scoring system9 (0 = absent, 1 = mild, 2 = moderate, 3 = marked, and 4 = very marked). Informed client consent was obtained for each dog, and the study was approved by the Texas A&M University Institutional Animal Care and Use Committee (animal use protocol Nos. 2017-0351CA and 2015-0043).
LC-MS-MS
The LC-MS-MS system consisted of a high-performance liquid chromatography systemb and a mass spectrometer.c Sample acquisition and data analysis were performed with commercial software.d A C18 columne was used to achieve chromatography separation. The mobile phase consisted of 0.1% formic acid (solvent A) and 0.1% formic acid in methanol (solvent B). The gradient was as follows: 0 to 5 minutes, 10% solvent B to 40% solvent B; 5 to 7 minutes, 40% solvent B to 95% solvent B; 7 to 9 minutes, 95% solvent B; 9 to 9.1 minutes, 95% solvent B to 10% solvent B; and 9.1 to 13 minutes, 10% solvent B. Injection volume was 10 μL, total assay time was 13.1 minutes, and flow rate was 0.4 mL/min.
The mass spectrometer was operated with electrospray ionization in the positive ion mode. Instrument parameters and transitions were optimized, with flow infusion set at a rate of 10 μL/min. Electrospray parameters used were a static spray voltage of 3,500 V, ion transfer tube temperature of 325°C, and vaporization temperature of 350°C. Sheath gas, auxiliary gas, and sweep gas were 45, 13, and 1 arbitray unit, respectively. Collision gas was at 200 Pa. Injector needles were washed with water and acetonitrile between successive injections. Autosampler carryover was monitored by performing blank injections. Column performance was monitored by measuring retention times.
Standard solutions, calibration, and internal control samples
Trans-4-hydroxy-l-prolinef (purity ≥ 99%) and trans-4-hydroxy-l-proline-2,5,5-D3g (purity > 98%) were used as internal standards. High-performance liquid chromatography–grade acetonitrileh was purchased, and 18 Mohm water was purified with a water purification system.i
Stock solutions (concentration, 0.1 mg/mL) were prepared by dissolving 0.1 mg of D3-trans-4-hydroxy-l-proline in 1 mL of water. Standard working solutions were then prepared by diluting stock solutions with water to obtain concentrations of 0.5, 1, 2.5, 5, 10, 25, 50, 100, and 250 ng/mL. Calibration standards were prepared by spiking 50 μL of blank canine serum with a freshly prepared working solution to achieve concentrations of 0.5, 1, 2.5, 5, 10, 25, 50, 100, and 250 ng/mL; solutions were then extracted with 100% acetonitrile to achieve standard solutions with concentrations of 0.005, 0.01, 0.025, 0.05, 0.1, 0.25, 0.5, 1, and 2.5 ng/mL.
Sample preparation
Canine serum samples (50 μL) were thawed at room temperature (21°C) and placed in a 1.5-mL Eppendorf tube; 200 μL of 100% acetonitrile was added, and the solution was mixed in a vortex device. Samples were then centrifuged at 17,500 × g and 4°C for 5 minutes. An aliquot (200 μL) of supernatant was transferred to a clean 1.5-mL Eppendorf tube. Another 50 μL of supernatant was transferred to a 2-mL glass vial with a glass insert; a sample of 10 μL was removed from this glass vial and injected into the LC-MS-MS system.
Analytic validation
The LC-MS-MS method was analytically validated in accordance with US FDA guidelines10 with respect to linearity accuracy precision, and recovery.
Linearity—Linearity of the assay was tested on 6 replicates of each of 3 concentrations of DQCs (2×, 5×, and 10×); the DQCs were dilutions of a quality control standard solution with a high concentration (5.0 ng/mL). Calculated concentration measurements were compared with the nominal concentration for each DQC.
Selectivity—Selectivity of the method was assessed by comparing chromatogram responses of 6 lots of naïve canine serum with those of canine serum spiked with the internal standard.
Calibration curve and the LLOQ—Calibration standards were prepared by spiking canine serum samples with 0.5 μL of working standard solutions or the internal standard. A calibration curve was constructed with a linear range of 0.01 to 2.50 ng/mL. Calibration curves were prepared for each analytic run by plotting the back-calculated concentration against the nominal concentration. Linearity was assessed by use of linear regression with appropriate weighting. The LLOQ was considered acceptable when 5 replicates of the lowest calibration standard had accuracy and precision deviations < 20%.
Accuracy and precision—Accuracy was evaluated for 4 quality control samples on 5 consecutive days. Accuracy was determined as the percentage difference between the mean of observed concentrations and the theoretical concentration (relative error); this value was required to be within ± 10% for all samples. Intra-assay and interassay precision were determined by assaying 3 samples at each of 3 quality control concentrations in 5 assays and on 5 consecutive days, respectively. The CV was used to express precision; CV values were required to be ≤ 10%.
Recovery—Recovery was determined by comparing the peak areas of extracted endogenous samples with low, mid, and high concentrations (1.8, 2.8, and 4.8 ng/mL, respectively) with the peak areas for postextraction spiked samples. Recovery percentage of the analyte was calculated by dividing peak areas of the analyte obtained for the extracted exogenous samples by those of postextraction spiked samples.
Matrix effect—Matrix effect of canine serum on the analysis of trans-4-hydroxy-l-proline concentrations was determined by comparing peak areas of the analyte in extracted blank serum samples with those obtained for standard solutions at the corresponding concentrations. The matrix effect was evaluated at 3 quality control concentrations (0.01, 0.10, and 0.50 ng/mL).
Statistical analysis
Differences between groups were analyzed for significance with the Mann-Whitney U and Kruskal-Wallis test when appropriate by use of commercially available software.j Significance was set at P < 0.05.
Results
Sample
Serum samples were obtained from 20 dogs with chronic hepatitis and 20 healthy control dogs. The dogs with chronic hepatitis comprised 9 Labrador Retrievers, 4 Doberman Pinchers, 4 crossbreds, 1 Dachshund, 1 Rat Terrier, and 1 Vizla. There were 11 spayed females, 6 castrated males, 2 sexually intact males, and 1 sexually intact female. Fibrosis score was 1 for 3 dogs, 2 for 9 dogs, 3 for 6 dogs, and 4 for 2 dogs. Median age of the dogs was 9 years (range, 3 to 9 years). The healthy dogs comprised 4 Labrador Retrievers, 3 crossbreds, 2 Boston Terriers, 2 Golden Retrievers, and 1 each of Alaskan Malamute, American Bull Terrier, Australian Shepherd, Boxer, Chihuahua, English Cocker Spaniel, Miniature Schnauzer, Schipperke, and Siberian Husky. There were 9 castrated males, 8 spayed females, 2 sexually intact females, and 1 sexually intact male. Median age of the dogs was 3 years (range, 0.5 to 12 years). Dogs with chronic hepatitis were significantly (P = 0.007) older than the control dogs.
LC-MS-MS
Fragment ion spectra, including quantification and confirming ions, of trans-4-hydroxy-l-proline and the deuterated internal standard D3-trans-4-hydroxy-l-proline were 132 m/z and 135.1 m/z, respectively (Figure 1). Trans-4-hydroxy-l-proline and the internal standard D3-trans-4-hydroxy-l-proline were detected under optimized conditions (Figure 2). There were no substantial interfering peaks for blank canine serum at the retention times of trans-4-hydroxy-l-proline and D3-trans-4-hydroxy-l-proline.
Product spectra for LC-MS-MS of trans-4-hydroxy-l-proline (A) and trans-4-hydroxy-l-proline-2,5,5-D3 (B).
Citation: American Journal of Veterinary Research 80, 5; 10.2460/ajvr.80.5.434
The LC-MS-MS chromatogram for canine serum with endogenous trans-4-hydroxy-l-proline (A), canine serum spiked with an internal standard (trans-4-hydroxy-l-proline-2,5,5-D3; B), and the summation of both (C).
Citation: American Journal of Veterinary Research 80, 5; 10.2460/ajvr.80.5.434
Calibration and linearity
The calibration curve for trans-4-hydroxy-l-proline in canine serum was linear over the concentration range of 0.01 to 2.50 ng/mL (r = 0.9997) and had consistent slope values when evaluated by use of weighted (1/2×) linear regression. The equation for the slope of the calibration curve of trans-4-hydroxy-l-proline was y = 9,000,000x − 69,198, and the LLOQ was 0.005 ng/mL
Precision and accuracy
Intra-day and interday precision and interday accuracy were determined for measurement of trans-4-hydroxy-l-proline concentrations in canine serum. Intraday precision (CV) for measurement of trans-4-hydroxy-l-proline concentrations in canine serum ranged from 2.1% to 3.0% (Table 1). Interday precision (CV) for measurement of trans-4-hydroxy-l-proline concentration in canine serum ranged from 3.2% to 5.3%. Interday accuracy (relative error) for measurement of trans-4-hydroxy-l-proline concentrations in canine serum ranged from −2.3% to 7.8% (Table 2).
Precision of LC-MS-MS analysis of the trans-4-hydroxy-l-proline concentration in canine serum.
| Sample | Measured concentration (ng/mL)* | Precision (%) |
|---|---|---|
| Intraday (n = 5) | ||
| 1 | 0.81 ± 0.03 | 3.0 |
| 2 | 1.75 ± 0.04 | 2.1 |
| 3 | 4.26 ± 0.12 | 2.5 |
| Interday (n = 5) | ||
| 1 | 0.81 ± 0.05 | 5.3 |
| 2 | 1.72 ± 0.09 | 4.5 |
| 3 | 4.23 ± 0.15 | 3.2 |
Precision represents the CV.
Mean ± SD.
Accuracy of LC-MS-MS analysis of trans-4-hydroxy-l-proline concentration in canine serum.
| Nominal concentration (ng/mL) | Measured concentration (ng/mL)* | Accuracy (%) |
|---|---|---|
| 0.10 | 0.10 ± 0.001 | –0.5 |
| 0.25 | 0.24 ± 0.010 | 5.5 |
| 0.50 | 0.46 ± 0.010 | 7.8 |
| 1.00 | 1.02 ± 0.010 | –2.3 |
Accuracy was evaluated for 4 quality control samples on 5 consecutive days and represents the relative error.
Mean ± SD.
Recovery
Extraction recovery was calculated by comparing the peak areas of extracted endogenous canine serum samples with low, mid, and high concentrations with those of postextraction spiked samples. Recovery for trans-4-hydroxy-l-proline was 48.9% for 1.8 ng/mL, 55.1% for 2.8 ng/mL, and 47.2% for 4.8 ng/mL (overall mean recovery, 50.4%).
Matrix effect
A possible matrix effect was evaluated for 3 concentrations during method development of the assay (Table 3). Overall mean matrix effect of blank serum was 25.1%.
Matrix effect of canine serum on trans-4-hydroxy-l-proline concentration.
| Observed concentration (ng/mL) | Peak area when dissolved in matrix serum (arbitrary units) | Peak area when dissolved in mobile phase (arbitrary units) | Matrix effect (%) |
|---|---|---|---|
| 0.01 | 93,223 | 116,247 | 19.8 |
| 0.10 | 738,634 | 1,187,546 | 37.8 |
| 0.50 | 3,570,732 | 4,334,020 | 17.6 |
Assessment for clinical application
Median concentration of trans-4-hydroxy-l-proline in the serum of dogs with chronic hepatitis (0.24 ng/mL) was significantly (P = 0.007) lower than the median concentration in the serum of healthy control dogs (0.78 ng/mL; Figure 3). Concentration of trans-4-hydroxy-l-proline of dogs with mild fibrosis (fibrosis score, 1 or 2) was significantly (P = 0.02) lower than that of healthy control dogs (0.20 and 0.78 ng/mL, respectively). There was no significant difference in concentrations between dogs with mild or marked (fibrosis score, 3 or 4) fibrosis and between dogs with marked fibrosis and healthy control dogs (Figure 4).
Serum trans-4-hydroxy-l-proline concentrations in 20 dogs with chronic hepatitis and 20 heathy control dogs. Each circle represents results for 1 dog, and the horizontal line represents the median for the group. *Median value differs significantly (P = 0.007) from the median value for the healthy dogs.
Citation: American Journal of Veterinary Research 80, 5; 10.2460/ajvr.80.5.434
Serum trans-4-hydroxy-l-proline concentrations in 12 dogs with chronic hepatitis and mild fibrosis, 8 dogs with chronic hepatitis and marked fibrosis, and 20 healthy control dogs. *Median value differs significantly (P = 0.02) from the median value for the healthy dogs. See Figure 3 for remainder of key.
Citation: American Journal of Veterinary Research 80, 5; 10.2460/ajvr.80.5.434
Discussion
An LC-MS-MS method for the quantification of trans-4-hydroxy-l-proline in canine serum was developed and analytically validated in the study reported here. Linearity of the calibration curve was indicated by a high correlation coefficient (r > 0.9997), which indicated a strong correlation in the linear range between the peak area ratio and the concentration of trans-4-hydroxy-l-proline. The method described here had a linear range of 0.01 to 2.50 ng/mL, and the LLOQ was 0.005 ng/mL. Selectivity for trans-4-hydroxy-l-proline was indicated by identification of the peak without interference. Intraday and interday precision and accuracy were within 15%, as stipulated by FDA guidelines. Mean percentage recovery was 50.4%; however, recoveries for trans-4-hydroxy-l-proline and the deuterated internal standard were consistent and similar, which allowed adjustment of the low recovery and fulfilled FDA validation guidelines. The method had acceptable performance in terms of sensitivity (ie, LLOQ), precision, and accuracy; a relatively short run time for analysis; and relatively simple sample preparation. Trans-4-hydroxy-l-proline concentrations were significantly lower in the serum of dogs with chronic hepatitis, compared with concentrations in the serum of healthy control dogs.
Hydroxyproline is fairly specific for collagen in mammals, and serum concentrations of hydroxyproline are a reflection of collagen metabolism in tissues with a high metabolic turnover of this protein.11 The quantity of hydroxyproline in the liver of rats,12,13 humans,14 and dogs15 is increased in proportion to the degree of hepatic fibrosis.12,13,15 During conditions of increased fibrogenesis (eg, hepatic fibrosis), serum concentrations of hydroxyproline are increased in humans7 and rats.16
Results of the present study that indicated lower serum hydroxyproline concentrations in dogs with chronic hepatitis than in control dogs were unexpected. The healthy control dogs used in the study were screened on the basis of a client questionnaire, evaluation of serum biochemical values, and results of liver function testing. However, the possibility of occult histologic liver disease cannot be completely excluded without examination of liver biopsy specimens. Although hydroxyproline is relatively specific for collagen, increased tissue concentrations have also been detected in patients with fibrotic pulmonary disease,17 fibrotic cardiac disease,18 and fibrotic renal disease.19 The effect of fibrosis in nonhepatic organs on serum hydroxyproline concentrations is unknown and was not examined in the present study. The possibility of concurrent or occult nonhepatic fibrosis cannot be excluded, and the effect on serum hydroxyproline concentrations could not be determined in the study reported here.
The serum concentration of hydroxyproline in dogs with mild hepatic fibrosis was lower than that in healthy control dogs of the present study; evaluation of a larger group of dogs with a wider range of fibrosis scores may be warranted. Serum hyperprolinemia in humans is a specific feature of alcoholic liver cirrhosis.6,8 Human patients with nonalcoholic cirrhosis or other forms of chronic liver disease have serum proline and hydroxyproline concentrations indistinguishable from those of healthy control subjects.6,8 On the other hand, patients with alcoholic hepatitis have serum proline and hydroxyproline concentrations significantly higher than those in patients with other forms of liver disease.6,8 It has been hypothesized that alcohol inhibits proline metabolism or facilitates its release from hepatocytes.6,8 The pathophysiology of alcohol-induced serum hyperprolinemia in humans with chronic hepatitis likely differs from that of dogs with chronic hepatitis because dogs do not develop alcohol-induced hepatitis.
The trans-4-hydroxy-l-proline concentration was significantly lower in the serum of dogs with chronic hepatitis than in the serum of healthy control dogs (median concentrations of 0.24 and 0.78 ng/mL, respectively). Age-related changes in the liver have been described in humans and include a decrease in volume, decrease in blood flow, mitochondrial dysfunction, and increase in susceptibility to fibrosis.20 Age-related fibrosis has not been characterized in humans. The groups of dogs in the study reported here were not matched on the basis of age. Serum samples available for the control group were obtained from dogs owned by veterinary students and faculty and staff of the veterinary teaching hospital, and these dogs typically were young. Therefore, we cannot rule out the possibility that the difference in the concentration of trans-4-hydroxy-l-proline between groups reflected a difference attributable to age. Age-matched healthy control dogs with histologic confirmation of disease status would be needed to more accurately assess the trans-4-hydroxy-l-proline concentrations in healthy dogs. Therefore, age may have contributed to the finding of a lower serum trans-4-hydroxy-l-proline concentration in dogs with mild fibrosis, which may have included patients of all ages that possibly had occult liver disease, compared with the concentration in healthy control dogs.
Furthermore, total urinary hydroxyproline concentration normalized on the basis of the urine creatinine concentration was found to be a better indicator of collagen metabolism in bone disease than was the serum hydroxyproline concentration.21 To the authors’ knowledge, total urinary hydroxyproline concentration has not been evaluated in dogs with liver disease, and serum proline or trans-4-hydroxy-l-proline concentrations have not been evaluated in dogs with hepatobiliary disease. A larger study may be warranted to determine whether lower serum trans-4-hydroxy-l-proline concentrations are a feature of dogs with chronic hepatitis, are correlated with the fibrosis score, and can be used as a noninvasive biomarker of hepatic fibrosis in this species. A noninvasive biological marker of liver fibrosis would not replace histologic examination of a biopsy specimen, but it could be used to assess the response to therapeutic interventions and as an objective tool for prognostic assessment. Additional research is also warranted to determine the importance of lower serum trans-4-hydroxy-l-proline concentrations for collagen metabolism in canine patients with hepatic fibrosis and the role of this marker molecule in disease progression.
Acknowledgments
This manuscript represents a portion of a thesis submitted by Dr. Lawrence to the Department of Veterinary Small Animal Clinical Sciences in the College of Veterinary Medicine and Biomedical Sciences at Texas A&M University as partial fulfillment of the requirements for a Doctor of Philosophy degree.
Presented in abstract form at the 28th Annual Congress of the European College of Veterinary Internal Medicine–Companion Animals, Rotterdam, Netherlands, September 2018.
ABBREVIATIONS
| CV | Coefficient of variation |
| DQC | Dilution quality control |
| LC-MS-MS | Liquid chromatography–tandem mass spectrometry |
| LLOQ | Lower limit of quantification |
Footnotes
Picrosirius Red staining kit, Polysciences, Warrington, Pa.
UltiMate 3000 HPLC, ThermoFisher Scientific, Waltham, Mass.
TSQ Quantiva triple-quadrupole mass spectrometer, ThermoFisher Scientific, Waltham, Mass.
TraceFinder software, version 3.3, ThermoFisher Scientific, Waltham, Mass.
Synergi 4 μm Fusion–RP, 150 × 2 mm, Phenomenex, Torrance, Calif.
TLC, Sigma-Aldrich Corp, St Louis, Mo.
CDN Isotopes, Pointe-Claire, QC, Canada.
Millipore, Milford, Mass.
EMDMillipore, Burlington, Mass.
GraphPad Prism Software, La Jolla, Calif.
References
1. Lawrence YA, Steiner JM. Canine chronic hepatitis: diagnosis and treatment. Today Vet Pract 2015;5:26–33.
2. McSweeney PM, Martin DD, Ramsey DS, et al. Clinical efficacy and safety of dexmedetomidine used as a preanesthetic prior to general anesthesia in cats. J Am Vet Med Assoc 2012;240:404–412.
3. Escobar A, Pypendop BH, Siao KT, et al. Effect of dexmedetomidine on the minimum alveolar concentration of isoflurane in cats. J Vet Pharmacol Ther 2012;35:163–168
4. Sevelius E. Diagnosis and prognosis of chronic hepatitis and cirrhosis in dogs. J Small Anim Pract 1995;36:521–528.
5. Angulo P, Kleiner DE, Dam-Larsen S, et al. Liver fibrosis, but no other histologic features, is associated with long-term outcomes of patients with nonalcoholic fatty liver disease. Gastroenterology 2015;149:389–397.
6. Mata JM, Kershenobich D, Villarreal E, et al. Serum free proline and free hydroxyproline in patients with chronic liver disease. Gastroenterology 1975;68:1265–1269.
7. Attallah AM, Toson EA, Shiha GE, et al. Evaluation of serum procollagen aminoterminal propeptide III, laminin, and hydroxyproline as predictors of severe fibrosis in patients with chronic hepatitis C. J Immunoassay Immunochem 2007;28:199–211.
8. Kershenobich D, García-Tsao G, Saldaña SA, et al. Relationship between blood lactic acid and serum proline in alcoholic liver cirrhosis. Gastroenterology 1981;80:1012–1015.
9. Lidbury JA, Hoffmann AR, Ivanek R, et al. Interobserver agreement using histological scoring of the canine liver. J Vet Intern Med 2017;31:778–783.
10. US FDA. Guidance for industry: bioanalytical method validation. Washington, DC: US Department of Health and Human Services, 2001.
11. Leroy EC, Sjoerdsma A. Clinical significance of a hydroxyproline-containing proteins in human plasma. J Clin Invest 1965;44:914–919.
12. Rojkind M, Diaz de León L. Collagen biosynthesis in cirrhotic rat liver slices a regulatory mechanism. Biochim Biophys Acta 1970;217:512–522.
13. Chvapil M, Ryan JN. The pool of free proline in acute and chronic liver injury and its effect on the synthesis of collagen and globular proteins. Agents Actions 1973;3:38–44.
14. Kershenobich D, Fierro FJ, Rojkind M. The relationship between the free pool of proline and collagen content in human liver cirrhosis. J Clin Invest 1970;49:2246–2249.
15. Toyoki Y, Sasaki M, Narumi S, et al. Semiquantitative evaluation of hepatic fibrosis by measuring tissue hydroxyproline. Hepatogastroenterology 1998;45:2261–2264.
16. Poole A, Myllyla R, Wagner JC, et al. Collagen biosynthesis enzymes in lung tissue and serum of rats with experimental silicosis. Br J Exp Pathol 1985;66:567–575.
17. Lubec B, Aufricht C, Amann G, et al. Arginine reduces kidney collagen accumulation, cross-linking, lipid peroxidation, glycoxidation, kidney weight and albuminuria in the diabetic kk mouse. Nephron 1997;75:213–218.
18. Osadchii O, Norton G, Deftereos D, et al. Rat strain-related differences in myocardial adrenergic tone and the impact on cardiac fibrosis, adrenergic responsiveness and myocardial structure and function. Pharm Res 2007;55:287–294.
19. Kim IH, Kisseleva T, Brenner DA. Aging and liver disease. Curr Opin Gastroenterol 2015;31:184–191.
20. Guañabens N, Pares A, Alvarez L, et al. Collagen-related markers of bone turnover reflect the severity of liver fibrosis in patients with primary biliary cirrhosis. J Bone Miner Res 1998;13:731–738.
21. George J, Chandrakasan G. Biochemical abnormalities during the progression of hepatic fibrosis induced by dimethylnitrosamine. Clin Biochem 2000;33:563–570.
