Secondary Logo

Journal Logo

Original Articles: Hepatology

Abundant Expression of Lysyl Oxidase-like 2 Protein in Intrahepatic Bile Ducts of Infants With Biliary Atresia

Honigbaum, Stefany; Zhu, Qingfeng; Layman, Andrew; Anders, Robert A.; Schwarz, Kathleen B.

Author Information
Journal of Pediatric Gastroenterology and Nutrition: September 2019 - Volume 69 - Issue 3 - p 344-350
doi: 10.1097/MPG.0000000000002414


What Is Known/What Is New

What Is Known

  • Biliary atresia is a rapidly fibrosing infantile liver disease without definitive therapy.
  • Lysyl oxidase-like 2 is a profibrotic enzyme.
  • Anti-lysyl oxidase-like 2 antibody has shown antifibrotic effects in animal models of liver fibrosis.

What Is New

  • Biliary atresia versus nonbiliary atresia livers exhibited increased intensity of staining for lysyl oxidase-like 2, particularly in bile ducts.
  • Extra-hepatic pediatric tissues rich in collagen exhibited minimal staining for lysyl oxidase-like 2.

Biliary atresia (BA) is an idiopathic neonatal liver disease characterized by extrahepatic biliary tree obstruction, inflammation, and rapidly progressive fibrosis (1–3). Incidence varies by geographical region but estimated to be anywhere from 1:5000 to 1:20,000 live births (1,4). Despite surgical palliation with a Kasai hepatoportoenterostomy (HPE), ongoing biliary fibrosis and attendant complications usually ensue. In fact, 1 cross-sectional study of 219 infants with BA determined that only 11% of children were without signs of chronic liver disease 10 years post HPE (5,6). At another facility, nearly 50% of BA adult survivors had already developed cirrhosis and complications of liver disease by age 20 years. There are currently no accepted antifibrotic treatments to prevent or delay the disease progression of fibrosis.

Lysyl oxidase (LOX) is a copper-dependent amine oxidase able to modify the extracellular matrix by initiating the cross-linkage of collagen and elastin (7–12). It is thought to originate from hepatic stellate cells and portal fibroblasts (10). Previous investigators have identified several closely related lysyl oxidase-like proteins, which have a conserved carboxyl-terminal end. Currently, the LOX family is made up of 5 members: LOX, LOXL1, LOXL2, LOXL3, and LOXL4 (12). Lysyl oxidase-like 2 (LOXL2) has been specifically implicated in the pathological environment of fibrosis because of its role in the cross-linking of fibrillar collagen (8–11,13). In adults, the expression of LOXL2 was found to be confined to fibrotic organs and absent in healthy tissue (13,14). It is upregulated in many fibrotic liver diseases (8,13). Likewise, LOXL2 has a hypothesized role in promoting invasion of tumors and malignant transformation (15). Accordingly, LOXL2 has become a promising target for future antifibrotic and anticancer therapies.

An anti-LOXL2 antibody may reverse hepatic fibrosis and increase survival in animal models (13). There were recently several phase 2 clinical trials investigating the use of an anti-LOXL2 therapy in adult patients with fibrosing liver diseases (16,17). Unfortunately, these trials failed to demonstrate reversal of hepatic fibrosis. Altinbas (14) has argued that anti-LOXL2 may still be our most effective antifibrotic therapy and that the initial trials failed because the patients had irreversible hepatic fibrosis and cirrhosis. He recommended application of the drug to earlier stages of fibrosis.

To our knowledge, there are no previous investigations regarding the presence of LOXL2 in livers of patients with BA or other pediatric liver diseases. Additionally, there is a paucity of data regarding the presence of LOXL2 in extrahepatic collagen-rich tissue from growing children. Given that organs of healthy growing infants and children contain abundant collagen and given the intimate relationship between LOXL2 and fibrillar collagen, it is important to ascertain the absence of LOXL2 in pediatric extrahepatic tissue.

The primary objective of this pilot study was to investigate the presence of LOXL2 protein in the hepatic tissue of patients with BA at the time of the Kasai procedure. The secondary objective of this study was to investigate the presence LOXL2 protein in pediatric extrahepatic collagen-rich tissue.


Patient Cohort

This study was submitted to and approved by the Johns Hopkins School of Medicine Institutional Review Board and was exempt from informed consent. We utilized tissue from the Johns Hopkins University Department of Pathology archives, a computerized tissue database with associated specimens dating back to 1986. We selected the 20 most recent BA patients (extracted from years 2003–2012) to use for our study. The diagnosis of BA had been previously confirmed for each of these patients via intraoperative cholangiogram. All BA tissue used was formalin-fixed, paraffin-embedded tissue obtained from wedge biopsies of the liver taken at the time of HPE. For non-BA pediatric (ages 0–21 years) hepatic tissue, we selected 20 patients extracted from years 1999 to 2008 who had hepatic biopsies or resections for a multitude of reasons, including: 7 cases of malignancy (primary or secondary tumors), 4 cases of vascular disease, 3 cases of metabolic disease, 2 cases of trauma, and 3 cases of nonspecific neonatal hepatitis and 1 common bile duct stricture (ages are shown in the Supplemental Table, Supplemental Digital Content, If tissue was used from a patient with a hepatic tumor, we assessed the nontumor tissue in the specimen.

For all of our BA and non-BA patients, relevant clinical data was obtained as close to the time of tissue extraction as possible and no more than 1 month before the procedure. This included general patient data (age, date of surgery), total bilirubin, direct bilirubin, alanine aminotransferase (ALT), albumin, and gamma glutamyl-transpeptidase (GGT).

For nonhepatic tissue, we chose 3 tissue types that we hypothesized may express LOXL2 because they are collagen-rich. Previous investigators have found LOXL2 mRNA in collagen-rich tissue with expression highest in reproductive tissues, uterus, prostate, and placenta (11). In contrast, other investigators have not found LOXL2 to be expressed or found it only to be minimally expressed in healthy tissue (8,13). The nonhepatic pediatric tissue we chose to investigate included skeletal muscle (n = 5), bone (n = 10), and heart (n = 5). The tissue for this group came from children ages 5 weeks to 18 years. The skeletal muscle was obtained from children ages 5 weeks to 14 years with biopsies obtained for the following reasons: 1 toe amputation, 1 resection of right axillary mass, 1 rib resection, 1 hemangioma resection, 1 left calf mass resection. We assessed the skeletal muscle on the periphery of these samples surrounding the resections. The tissue for the heart specimens came from children ages 7 months to 18 years. The heart tissue was obtained for the following indications: 1 mitral valve explant, 1 stillborn fetal heart, 1 restrictive cardiomyopathy, and 2 dilated cardiomyopathies. The tissue from the bone was from children ages 5 weeks to 18 years including 7 accessory digit removals, 1 xiphoid, 1 femur, and 1 spine.


Liver biopsies were scored for liver fibrosis, LOXL2 intensity, percent of bile duct LOXL2 staining, and copper deposition (given that LOXL2 is a copper-dependent enzyme). Extrahepatic control tissue was stained only for LOXL2 intensity. The tissue fibrosis was evaluated using hematoxylin and eosin-stained tissue and/or Masson/trichrome histochemical stain if available. The fibrosis scoring system used was the Ishak fibrosis score (18) (0–6) using the following scoring system (18): 0 = normal (no fibrosis), 1 = fibrous expansion of some portal areas (+/− ) short fibrous septa, 2 = fibrous expansion of most portal areas (+/−) short fibrous septa, 3 = fibrous expansion of most portal areas with occasional portal to portal (P-P) bridging, 4 = fibrous expansion of portal areas with marked bridging (P-P) as well as portal to central (P-C), 5 = marked bridging (P-P and/or P-C), with occasional nodules (incomplete cirrhosis), 6 = cirrhosis, probable or definite.

Immunohistochemistry for Lysyl Oxidase-like 2

All tissues were fixed in 10% neutral-buffered formalin. Five-micron sections of the paraffin-embedded tissue were stained with a commercially available anti-LOXL2 polyclonal antibody (Abcam; dilution 1:100) followed by secondary anti-rabbit antibody (Dako, Carpinteria, CA) and detected using the diaminobenzidine substrate kit (Dako) per instructions from manufacturer. A hepatic pathologist (R.A.A.) who was blinded to the diagnosis did all tissue grading. For LOXL2 intensity and LOXL2 distribution of bile duct staining, each specimen was evaluated at ×20 fields with 3 to 5 portal areas per specimen graded. LOXL2 intensity (0–2) staining of bile ducts was evaluated according to the following semiquantitative scoring system: 0 = none, 1 = light staining, 2 = heavy staining. The distribution of bile ducts staining (0–3) was then performed to assess the percentage of bile ducts staining for LOXL2. This was assessed with the following scoring system: 0 ≤25%, 1 = 26% to 50%, 2= 51% to 75%, 3 ≥ 75%). Additionally, we noted and commented upon the staining of other portal constituents (portal vein, artery, mesenchyme) as well as the hepatocytes. The LOXL2 staining of the nonhepatic tissue was initially assessed for the presence or absence of LOXL2. If present, the intensity (0–2) was scored via the scoring system mentioned above.

Copper Deposition (Rhodanine) Staining

Following the LOXL2 staining of the tissue, we stained the same formalin-fixed, paraffin-embedded tissue with 5-(4-Dimethylaminobenzylidene) rhodanine 98% (Alta Aesar, Johns Matthey Company, England) to detect the presence of copper. Staining was done as a modification of the Johns Hopkins Hospital rhodanine staining protocol and as previously described (19). Tissue copper scoring (0–3) was done using the following scoring system: 0 = negative, no rusty red cytoplasmic granules in hepatocytes, 1 = isolated periportal cells containing small sparse granules (not seen at low magnification, 10×, and involving less than 2 lobules/nodules), 2 = moderate to numerous periportal hepatocytes containing copper granules (seen at low magnification, 10×, mainly in zone 1), 3 = widespread and heavy deposition of granules throughout the lobule.

Statistical Analysis

For the comparison of BA versus non-BA tissue, statistical analysis was performed using the Wilcoxon rank sum test for independent variables and reported using the median/interquartile range (p25–p75) for nonnormally distributed continuous variables. Differences were considered to be significant at P < 0.05. For our nonhepatic tissue, numbers are presented as median/interquartile ranges (p25–p75).


Lysyl Oxidase-like 2 Expression in Hepatic Tissue: Biliary Atresia Versus Nonbiliary Atresia

Twenty wedge biopsies from time of Kasai procedure were obtained from children with BA. Twenty biopsies (wedge, needle, and resections) from children without BA were obtained. Both pediatric cohorts had similar degrees of fibrosis (BA 4 [3–5] vs non-BA 3.5 [1–5], P = 0.13). The 2 pediatric populations varied in terms of age, ALT, direct bilirubin levels, and GGT (Table 1).

Clinical and laboratory profiles of biliary atresia and nonbiliary atresia patients

Bile duct staining of LOXL2 was noted in both BA and non-BA livers (Fig. 1). The intensity was, however, uniformly higher in patients with BA (Fig. 2A). BA patients exhibited a median intensity score of 2 (2–2) versus a non-BA median of 1.4 (0.3–2) (P < 0.001; Wilcoxon rank sum test). The percent bile duct staining was also significantly different between test groups (Fig. 2B) with BA samples having a median score of 3 (3–3) and non-BA patients having a median of 2.8 (0.4–3) (P = 0.001; Wilcoxon rank sum test). We also compared the control group (n = −7) resected for malignancy versus the remainder of the controls and found no difference in LOXL2 intensity (0.7568) or distribution (0.8740), despite a difference in fibrosis (0.0009).

Lysyl oxidase-like 2 staining in biliary atresia (A and B) and nonbiliary atresia liver tissue (C and D). (C) Giant cell hepatitis ages 0.1 years; (D) HCC (nonmalignant part) ages 18 years. Arrows highlight strongly stained bile ducts. Arrow heads highlight minimal patchy staining of hepatocytes. Scale bars = 200 μm. HCC = hepatocellular carcinoma.
Lysyl oxidase-like 2 staining in bile ducts in biliary atresia and nonbiliary atresia liver tissue. (A) Intensity of bile duct staining. (B) Percentage of bile duct staining.

We also noted minimal patchy staining of hepatocytes and other portal constituents in both samples (perivascular areas and mesenchyme). When only the infant control non-BA tissue (n = 5) was compared with the BA tissue (n = 20) as a means to control for age variation, there remained a difference in intensity of LOXL2 staining (BA 2 [2–2] vs non-BA infants 0.2 [0–0.6], P < 0.001) and in percent of bile duct LOXL2 staining (BA 3 [3–3] vs non-BA infants 0.2 [0–1.7], P < 0.001). The hepatic fibrosis scoring remained similar between the 2 groups (BA 4 [3–5] vs non-BA infants 4 [2–5], P = 0.67).

Copper Deposition in Hepatic Tissue: Biliary Atresia Versus Nonbiliary Atresia

Despite LOXL2 being a copper-dependent enzyme, we found no significant correlation between areas of copper deposition and/or increased copper deposition in tissue that strongly expressed LOXL2 and tissue that did not strongly express LOXL2. The copper staining for the BA tissue was 0 (0–1) and for the control was 0 (0–0) (P = 0.07). When we compared just the infant controls (n = 5) to the BA tissue, we also found no significant difference in copper deposition (BA 0 [0–1] vs non-BA 0 [0–0.8], P = 0.61).

Lysyl Oxidase-like 2 Staining in Nonhepatic Pediatric Tissue

We stained tissue from pediatric bone (n = 10), heart (n = 5), and skeletal muscle (n = 5). For all tissue, there was absent to minimal staining of LOXL2 (Fig. 3). The median for our bone tissue was an intensity scoring of 0 (0–0). The heart specimens had a median score of 1 (1–1) and the skeletal muscle had a median score of 1 (1–1). Given the importance of studying tissues from young infants we examined the LOXL2 intensity scores according to age. Six controls were <6 months of age. There were 5 muscle samples: ages were 5 weeks; 7 months; 16 months; 8 years; and 14 years. There were 10 bone samples: ages were 5 weeks; 3 months; 4 months; 5 months (n = 2); 7 months; 2 years (n = 3); and 18 years. There were 5 heart samples: ages were 7 months; 8 months; 5 years; 16 years; and 18 years. LOXL2 intensity scores were as follows: muscle: 4 were intensity score of 1; 1 was an intensity score of 2. In the 5-week olds, there was LOXL2 staining (intensity score 1) in the sarcomeres of muscle bundles. Bones: 9 bones had intensity score of 0; 1 bone (in the 5 months old) had an intensity score of 2 at the physis. Heart: 4 had intensity score of 1. The youngest (7 months) had an intensity score of 0 in the muscle bundles. Trichrome staining was performed on the nonliver tissue of the infant controls; all of the tissues had physiologic collagen. There was only mild pathologic fibrosis in the endocardium of the heart. The subcutaneous tissue of the extra-numerary digit also had mild pathologic fibrosis.

Lysyl oxidase-like 2 staining in skeletal muscle (A), bone (B and D), and heart (C). Arrows highlight weak staining in skeletal muscle only. (A and C) Taken at 10× (B and D) taken at 20×.


Our study shows that LOXL2 is overexpressed in intrahepatic bile ducts in liver biopsies from infants with biliary atresia, which is a novel finding. Although LOXL2 was also present in our controls with hepatic fibrosis, it was expressed significantly less than in our BA tissue. Furthermore, comparison of those controls resected for malignancy versus the remainder of the controls showed no difference in LOXL2 staining despite differences in fibrosis between the 2 types of controls. We believe that such a large discrepancy between the 2 populations is because of the intense staining of the bile ducts in patients with BA, a result especially notable as bile duct proliferation is a pathological hallmark of this disease.

In addition to the very intense and uniform staining of the bile ducts in patients with BA, we also noted minimal patchy cytoplasmic staining of hepatocytes (which could either be nonspecific vs simply minimal expression) as well as prominent staining of the endothelial cells of the portal vessels. When we compared our infant controls (n = 5) to our BA patients (n = 20) to control for age variation, we found an even greater discrepancy in the intensity of LOXL2 staining (BA 2 [2–2] vs non-BA infants 0.2 [0–0.6], P < 0.001) and in percent of bile duct LOXL2 staining (BA 3 [3–3] vs non-BA infants 0.2 [0–1.7], P < 0.001).

There are several limitations to our study. A much larger number of non-BA infant controls with hepatic fibrosis would need to be compared with a larger number of BA infants to determine with more certainty that LOXL2 is over-expressed in BA versus controls with hepatic fibrosis because of other causes. Furthermore, although we found minimal expression of LOXL2 in pediatric heart, skeletal muscle, and bone, it would be necessary to examine LOXL2 expression in a much larger number of extrahepatic tissues from infants, children, and adolescents before more firm conclusions about the probable safety of anti-LOXL2 therapy in the pediatric age group could be considered.

It should be noted that the ability to age-match controls was limited by the tissue available in our archives and the fact that all BA tissue was obtained at a very young age because of the nature of the course and treatment of the disease. It is also important to note that obtaining pediatric non-BA control hepatic tissue was not possible given the ethical prohibition of obtaining healthy pediatric hepatic tissue. Though the control patients did not have biliary atresia, they were not representative of healthy hepatic tissue, as they all had other existing medical conditions.

We believe that the failure of anti-LOXL2 therapies in adults with advanced liver disease should not dissuade investigators from examining this type of therapy in infantile BA. The drug trials may have failed because the liver fibrosis was too advanced (14), as the drug was not delivered to the target tissues, because of inadequate dosage, or other reasons. Brenner (20) has shown that hepatic fibrosis can be a reversible process. Unfortunately, the drug trials failed to demonstrate that anti-LOXL2 can achieve this in humans with advanced liver disease. Anti-LOXL2, however, does reverse biliary fibrosis in experimental animals (21). Our data suggest that the next step in investigating a role for anti-LOXL2 therapy in infants with BA would be to expand our observations to a larger cohort of infants with BA or other causes of cholestasis as well as to nonhepatic control tissues rich in collagen. Furthermore, antifibrotic drugs that may have clinical applications in preventing fibrosis in BA may also have roles in other pediatric models of liver fibrosis, such as cystic fibrosis-associated liver disease. Future studies could examine LOXL2 in the serum of healthy individuals and those with hepatic fibrosis to see if there are correlations between LOXL2 in the circulation and that in the liver. Focusing on therapies that counteract enzymes known to be overexpressed in models of fibrosis but are absent in healthy tissue provide a promising new approach to treatment.


We acknowledge Alison Klein, MD for statistical support.


1. Wildhaber BE. Biliary atresia: 50 years after the first kasai. ISRN Surg 2012; 2012:132089.
2. Santos JL, Carvalho E, Bezerra JA. Advances in biliary atresia: from patient care to research. Braz J Med Biol Res 2010; 43:522–527.
3. Schwarz KB, Haber BH, Rosenthal P, et al. Childhood Liver Disease Research and Education Network. Extrahepatic anomalies in infants with biliary atresia: results of a large prospective North American multicenter study. Hepatology 2013; 58:1724–1731.
4. Bessho K, Bezerra JA. Biliary atresia: will blocking inflammation tame the disease? Annu Rev Med 2011; 62:171–185.
5. Ng VL, Haber BH, Magee JC, et al. Childhood Liver Disease Research and Education Network (CHiLDREN). Medical status of 219 children with biliary atresia surviving long-term with their native livers: results from a North American multicenter consortium. J Pediatr 2014; 165:539.e2–546.e2.
6. Shinkai M, Ohhama Y, Take H, et al. Long-term outcome of children with biliary atresia who were not transplanted after the Kasai operation: >20-year experience at a children's hospital. J Pediatr Gastroenterol Nutr 2009; 48:443–450.
7. Smith-Mungo LI, Kagan HM. Lysyl oxidase: properties, regulation and multiple functions in biology. Matrix Biol 1998; 16:387–398.
8. Vadasz Z, Kessler O, Akiri G, et al. Abnormal deposition of collagen around hepatocytes in Wilson's disease is associated with hepatocyte specific expression of lysyl oxidase and lysyl oxidase like protein-2. J Hepatol 2005; 43:499–507.
9. Reiser K, McCormick RJ, Rucker RB. Enzymatic and nonenzymatic cross-linking of collagen and elastin. FASEB J 1992; 6:2439–2449.
10. Perepelyuk M, Terajima M, Wang AY, et al. Hepatic stellate cells and portal fibroblasts are the major cellular sources of collagens and lysyl oxidases in normal liver and early after injury. Am J Physiol Gastrointest Liver Physiol 2013; 304:G605–G614.
11. Jourdan-Le Saux C, Tronecker H, Bogic L, et al. The LOXL2 gene encodes a new lysyl oxidase-like protein and is expressed at high levels in reproductive tissues. J Biol Chem 1999; 274:12939–12944.
12. Nishioka T, Eustace A, West C. Lysyl oxidase: from basic science to future cancer treatment. Cell Struct Funct 2012; 37:75–80.
13. Barry-Hamilton V, Spangler R, Marshall D, et al. Allosteric inhibition of lysyl oxidase-like-2 impedes the development of a pathologic microenvironment. Nat Med 2010; 16:1009–1017.
14. Altinbas A. A quick overview to the early phase clinical trials of Simtuzumab®: are we losing the most promising anti-fibrotic product? Med Hypotheses 2017; 108:159–160.
15. Wu L, Zhu Y. The function and mechanisms of action of LOXL2 in cancer (review). Int J Mol Med 2015; 36:1200–1204.
16. Muir AJ, Levy C, Janssen HLA, et al. GS-US-321-0102 Investigators. Simtuzumab for primary sclerosing cholangitis: phase 2 study results with insights on the natural history of the disease. Hepatology 2018; 69:684–698.
17. Harrison SA, Abdelmalek MF, Caldwell S, et al. Simtuzumab is ineffective for patients with bridging fibrosis or compensated cirrhosis caused by nonalcoholic steatohepatitis gastroenterology 2018; 155:1140–1153.
18. Ishak K, Baptista A, Bianchi L, et al. Histological grading and staging of chronic hepatitis. J Hepatol 1995; 22:696–699.
19. Mounajjed T, Oxentenko AS, Qureshi H, et al. Revisiting the topic of histochemically detectable copper in various liver diseases with special focus on venous outflow impairment. Am J Clin Pathol 2013; 139:79–86.
20. Brenner DA. Reversibility of liver fibrosis. Gastroenterol Hepatol (N Y) 2013; 9:737–739.
21. Ikenaga N, Peng ZW2, Vaid KA, et al. Selective targeting of lysyl oxidase-like 2 (LOXL2) suppresses hepatic fibrosis progression and accelerates its reversal. Gut 2017; 66:1697–1708.

antifibrotic therapy; hepatic fibrosis; neonatal cholestasis

Supplemental Digital Content

Copyright © 2019 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition