Variations in ATP7B in cats with primary copper-associated hepatopathy

Abstract

Objectives

Primary copper-associated hepatopathy (PCH) has been reported in young cats. Although our group recently reported a young cat with PCH harbouring single-nucleotide variations in ATP7B, limited information is available regarding its association with the pathogenesis of feline PCH. The objective of this study was to investigate the prevalence of ATP7B variations in cats with PCH.

Methods

Rhodanine staining was performed to detect hepatic copper accumulation (HCA) in intraoperative liver tissue specimens from 54 cats. In cats with HCA, variations in ATP7B and COMMD1 and serum ceruloplasmin activity were analysed.

Results

Based on age, liver histopathological findings and hepatic distribution of accumulated copper, PCH was suspected in 4/54 cats. Sequence analysis of ATP7B and COMMD1 revealed single-nucleotide variations in ATP7B in 3/4 cats with PCH. Among the cats with PCH, one showed remarkably low serum ceruloplasmin activity, while the other three did not.

Conclusions and relevance

The results of this study suggest that some cats with PCH harbour single-nucleotide variations in ATP7B, suggesting that feline PCH is an equivalent disorder to human Wilson’s disease. This study provides basic evidence facilitating further studies of the pathophysiology and treatment of feline PCH.

Introduction

Copper, an essential micronutrient for biological activities, is involved in various metabolic process. It is absorbed from the small intestine and processed in the liver. In hepatocytes, ATP7B, a P-type copper-transporting ATPase, helps secrete copper into bile and blood with COMMD1, copper metabolism MURR domain 1. ¹ During ATP7B-mediated copper secretion in blood, most of the copper is incorporated into ceruloplasmin (Cp), a major copper-transporting protein. ² Impairment of secretion to bile and blood leads to copper accumulation in hepatocytes, and excessive hepatic copper accumulation (HCA) is a potent cellular toxin causing oxidative stress and death of hepatocytes. ¹

Primary copper-associated hepatopathy (PCH) reportedly occurs in both humans and some animal species. In human medicine, Wilson’s disease (WD) is a well-characterised autosomal recessive disorder of copper metabolism caused by familial ATP7B mutations, resulting in copper accumulation in various tissues, such as liver, kidney, brain and cornea.^3,4 WD is diagnosed from clinical signs, clinicopathological and histopathological findings, and ATP7B sequence analysis. ⁴ Although more than 600 pathogenic ATP7B mutations are known among patients with WD, the mutation pattern reportedly varies geographically. ⁵ Furthermore, typical patients with WD reportedly have low serum Cp activity owing to impaired ATP7B function, while 25–36% of children with WD were reported to have normal serum Cp. ⁶ In patients with WD, copper accumulates first in the periportal region of the liver, followed by the centrilobular and intermediate regions. Several rodent models of WD have been described, such as the Long-Evans Cinnamon rat, ⁷ toxic milk mouse ⁸ and the genetically engineered ATP7B^–/– mouse. ⁹

Among dogs, PCH has been frequently identified in Bedlington Terriers, where deletion of exon 2 of COMMD1 reportedly causes PCH. ¹⁰ In dogs with PCH, HCA primarily occurs in the centrilobular region, which is different from human WD. ¹¹ A recent genome-wide association study identified single-nucleotide variations (SNVs) in ABCA12 (a metal ion transporter) in Bedlington Terriers without the COMMD1 deletion, indicating yet another cause of PCH in this breed. ¹² Furthermore, ATP7B variations were recently identified in a Labrador Retriever, suggesting that the breed is a new dog model of naturally occurring WD.^13,14

Recently, PCH has been reported to occur in young cats.^15–17 Cats with PCH present non-specific clinical signs such as vomiting, anorexia, lethargy and weight loss. HCA is detected primarily in the centrilobular regions of the liver.^15–17 Limited information is available regarding the aetiologic or histopathological characteristics of feline PCH. Our group recently reported a young crossbred cat with PCH, which harboured a familial SNV in ATP7B largely affecting protein function. Cats might be a novel potential animal model of WD. ¹⁸ In this study we hypothesised that ATP7B variations are associated with feline PCH. This study aimed to investigate the prevalence of ATP7B variations in cats with PCH.

Materials and methods

Liver specimens and PCH diagnosis

Formalin-fixed paraffin-embedded (FFPE) liver specimens of cats referred to the Veterinary Medical Center, the University of Tokyo, Japan, between April 2013 and September 2018 were assessed. These specimens were surgically resected for treatment or biopsy. Medical records of these cases were also reviewed to obtain the following information: signalment, food, histopathological hepatic diagnosis, concurrent diseases and treatment.

Two 2 µm-thick tissue sections were obtained from the stored FFPE liver specimen of each cat. One section was stained with haematoxylin and eosin, and the other was stained with rhodanine. Rhodanine staining was performed to detect HCA with conventional methods, where accumulated copper granules were stained red to brown. Thereafter, each section was counterstained with haematoxylin for 3 mins. Then, the degree of copper accumulation was scored in 0–5 grades based on a previous report. ¹⁹ Each slide was histologically evaluated by a veterinary pathologist (JKC). PCH was presumptively diagnosed based on age, lack of an evident history of excessive copper intake, results of blood tests, hepatic histopathological diagnosis, HCA detection by rhodanine staining and the distribution pattern of rhodanine-stained copper granules in the liver.

Variation analysis of ATP7B and COMMD1

Genomic DNA was extracted from the FFPE liver specimens of the cats with HCA. Primer pairs for ATP7B (exon 1–21) and COMMD1 (exon 1–3) were synthesised as previously reported. ¹⁸ The DNA samples were amplified by PCR and amplification was confirmed electrophoretically. Then, sequences of the PCR products were analysed by Sanger sequencing. This analysis was subsequently conducted using genomic DNA extracted from peripheral blood samples harvested from 10 healthy cats as normal controls of ATP7B. Thereafter, case-specific variations not observed in the 10 healthy cats were extracted, and their effects on protein function were predicted using PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/). ²⁰

Measurement of serum Cp activity

Serum Cp activity was analysed by a spectrophotometric method in cats with HCA as previously reported. ²¹ In this method, Cp activity was expressed as an increase in absorbance per minute (ΔAbs/min). Serum samples were stored at −20°C until use, as in the previous report. Furthermore, serum samples were harvested from the same 10 healthy cats described above for comparative analysis of serum Cp activity to prepare a reference interval.

Results

Cases and diagnosis of suspected PCH

Liver tissue specimens and medical records of 54 cats were included in this study. Histopathological diagnoses included microvascular dysplasia (MVD) (41%; n = 22), lymphocytic cholangitis (19%; n = 10), hepatocellular degeneration (15%; n = 8), hepatic cyst (9%; n = 5), neoplastic diseases (9%; n = 5), hepatic cirrhosis (4%; n = 2) and hematoma (2%; n = 1). No remarkable finding was observed in four cats (7%). Congenital port-systemic shunt (PSS) was detected in 20/22 cats with MVD, whereas multiple acquired PSS was detected in one cat with MVD.

The representative results of rhodanine staining are shown in Figure 1, and the study results are summarised in Table 1. Upon rhodanine staining, HCA was observed in five cats (9%) including one cat that was described in a previous case report (case 1), ¹⁸ whereas no rhodanine-stained granules were observed in the other 49 cats (grade 0). The grade in each cat with HCA is shown in Table 1. Among the cats with HCA, one cat was an intact male, one was a castrated male, two were intact females and one was a spayed female, all being mixed breed. One cat was 11 years old, whereas the others were 0–2 years of age.

Figure 1

Representative results of histological diagnosis and findings of liver specimens of cats with hepatic copper accumulation (HCA). (a,b) Severe fibrosis with regenerative nodules suggested a diagnosis of hepatic cirrhosis in case 1. Hepatocellular vacuolar degeneration was also observed. Copper granules were diffuse upon rhodanine staining (grade 3). (c,d) Mild hepatocellular vacuolar degeneration was observed in case 4, and HCA was observed in the centrilobular region (grade 3). (e,f) Severe lymphocytic infiltration was observed in most of the portal areas in case 5. Upon rhodanine staining, HCA was observed mainly in the intermediate region (grade 2). Haematoxylin and eosin for (a), (c) and (e), and rhodanine staining for (b), (d) and (f). Bar = 100 µm

Table 1

Summary of results of the present study

Case	Histopathological diagnosis	Hepatic copper distribution	Grade of HCA	SNVs in ATP7B	Serum Cp activity (∆Abs/min)*
1	Hepatic cirrhosis	Panlobular region	3	p.G529R, p.T1297R	0.00061
2	Hepatic cirrhosis	Panlobular region †	4	p.V268I, p.P550L	0.00632
3	MVD	Centrilobular region	1	None	0.00799
4	Hepatocellular vacuolar degeneration	Centrilobular region	3	p.S511G, p.P550L	0.00616
5	Lymphocytic cholangitis	Intermediate region	2	None	NE

*

The median serum ceruloplasmin (Cp) activity in healthy cats was 0.004696 ∆Abs/min (range 0.003139–0.006547 ∆Abs/min)

†

The hepatic copper concentration was 1564 µg/g of dry weight in this case

HCA = hepatic copper accumulation; SNV = single-nucleotide variation; MVD = microvascular dysplasia; NE = not examined

Upon histopathological evaluation, cases 1 and 2 were diagnosed with hepatic cirrhosis with multiple regenerative nodules; case 3 with MVD; case 4 with only mild hepatocellular vacuolar degeneration; and case 5 with severe lymphocytic cholangitis. Hepatocellular vacuolar degeneration occurred in four cats (cases 1–4). Hepatic cirrhosis was not diagnosed in any of the cats without HCA. Furthermore, HCA was observed in the panlobular region in cases 1 (grade 3) and 2 (grade 4), in the centrilobular region in cases 3 (grade 1) and 4 (grade 3), and in the intermediate region in case 5 (grade 2). See the supplementary material for the other results of rhodanine staining in the cats not shown in Figure 1. Hepatic copper concentration was measured as 1564 µg/g of dry weight in case 2. Furthermore, it was confirmed during interviews that the owners of these five cats with HCA fed them general commercial diets without any supplements containing additional copper. Based on these findings, PCH was presumptively diagnosed in cases 1–4. However, secondary HCA was presumptively diagnosed in case 5 based on the cat’s age, histopathological diagnosis and the distribution pattern of rhodanine-stained copper granules in the liver similar to that in cats with secondary HCA in a previous report. ¹⁶

Treatment of cats with PCH

Of the four cats with suspected PCH, case 1 was treated with a penicillamine-based regimen, and clinicopathological abnormalities improved after initial treatment. ¹⁸ Although cases 2 and 4 also received penicillamine, the response could not be evaluated owing to discontinuation of treatment due to poor compliance in case 2 and side effects (anorexia and vomiting) in case 4. No medication was administered in case 3. Survival time could not be evaluated as they were all alive at the time of writing.

Variations in ATP7B and COMMD1

Upon sequence analysis of genomic DNA samples obtained from the five cats with HCA, three cats with suspected PCH harboured SNVs in ATP7B (Table 1), with no variation in COMMD1. Among ATP7B SNVs identified herein, p.T1297R and p.P550L were predicted to have high functional effects. The latter was common in two cats (cases 2 and 4). All SNVs identified herein were homozygous. The region containing p.T1297R was within the copper-transporting P-type ATPase domain and that containing p.P550L was within the copper-binding domain.

Serum Cp activity

Median serum Cp activity in healthy cats was 0.004696 ΔAbs/min (range 0.003139–0.006547 ΔAbs/min). Serum Cp activity was measured in the four cats with suspected PCH (Table 1). Consequently, it was markedly lower in case 1 (0.00061 ΔAbs/min), whereas it was similar to that in 10 healthy cats in the other three cases.

Discussion

Rhodanine staining was performed to detect HCA, using liver tissue specimens from 54 cats, and HCA was observed in five cats (9%). Among the cats with HCA, one had severe lymphocytic cholangitis with rhodanine-stained copper granules primarily in the intermediate region, indicating secondary HCA. This histological pattern of HCA was similar to that in cats with secondary HCA due to cholestatic diseases reported previously. ¹⁶ The absence of SNVs in ATP7B and the age of onset (11 years) in this case might support the diagnosis of secondary HCA. This suggests that some cats with lymphocytic cholangitis may develop secondary HCA, but this needs to be further evaluated.

In this study, PCH was presumptively diagnosed in 4/54 (7%) young cats, with a frequency and age of onset similar to those reported previously. ¹⁶ In the two cats with hepatic cirrhosis, HCA was observed in the panlobular region. This accumulation pattern might be due to the histological progression of PCH, including formation of multiple regenerative nodules. Copper accumulation in the panlobular region was previously reported in a cat with marked architectural remodelling. ¹⁷ Furthermore, the hepatic copper concentration of 1564 µg/g dry weight in one of the two cats with hepatic cirrhosis was consistent with the reported concentration (> 1000 µg/g dry weight) in cats with PCH in a previous report. ¹⁶ Together, considering the absence of hepatic cirrhosis in cats without HCA, hepatic cirrhosis in young cats might be associated with PCH. In the other two cats with suspected PCH, copper accumulation was observed in the centrilobular region, similar to a previously reported pattern in cats and dogs with PCH.^11,15,16 Although secondary HCA cannot be completely excluded, given the age of onset, lack of histological evidence of other cholestatic disease or hepatitis, and the absence of evident history of excessive copper intake, the findings in these four cats were most compatible with PCH rather than secondary HCA. Furthermore, in the four cats with PCH, hepatocellular vacuolar degeneration was evident as reported previously in cats with PCH.^16,17,22 Further studies are required to investigate the association between hepatocellular vacuolation and feline PCH pathophysiology.

Sequence analysis revealed SNVs in ATP7B in 3/4 cats (75%) with PCH. Of the SNVs identified herein, p.T1297R and p.P550L were predicted to markedly influence ATP7B function. Therefore, SNVs in ATP7B might play critical roles in PCH pathogenesis in cats, and examination of SNVs in ATP7B might be a useful method to diagnose feline PCH. Furthermore, P550L was common in two cats with PCH, suggesting that it might be a frequently harboured SNV in cats in Japan. However, no variation was identified in a cat with PCH, thereby indicating that causes other than ATP7B variations might induce PCH in some cases. In canine PCH, a recent study on Bedlington Terriers reported novel SNVs in ABCA12 in cases not harbouring an exon 2 deletion in COMMD1. ¹² Furthermore, in humans, ATP7B mutations were not identified in 7% of patients with WD. ²³ Therefore, PCH might result from variations in genes other than ATP7B in some cats. Although no variation was identified in COMMD1 in this study, further comprehensive studies are required to investigate variations in other genes in a larger number of cats with PCH.

One cat with PCH showed comparatively low serum Cp activity, as reported in some patients with WD. ²⁴ Thus, PCH pathophysiology in the case with low serum Cp activity was suggested to resemble that of typical patients with WD. However, serum Cp activity in the other cats were similar to the range of healthy cats. Some patients with WD were reported to have normal serum Cp activity, ⁶ and it was also reported that some mutations in ATP7B do not impair Cp synthesis. ²⁴ Based on the serum Cp activity determined herein, some PCH-causing variations in feline ATP7B might not impair Cp synthesis. As serum Cp activity was not measured in cats without PCH in this study, further studies are required to investigate the association of serum Cp activity with the pathophysiology of feline PCH.

Regarding the treatment of cats with PCH, one cat responded well to penicillamine-based treatment; however, treatment responses could not be evaluated in the other cats. It was previously suggested that chelation to eliminate copper might be effective in treating PCH in cats. ¹⁶ As some cats with PCH harboured SNVs in ATP7B in this study, a treatment method based on SNVs for patients with WD might be applicable for cats with PCH. Further studies are required to evaluate the efficacy and response rate of penicillamine-based treatment in cats with PCH.

In this study, the dry weight of copper in the liver could only be measured in one cat. Detection of HCA by rhodanine-stained hepatic copper granules without hepatic copper concentration was a limitation as rhodanine staining only detects severe copper accumulation. ²⁵ In humans, <50% of patients with WD have been reported to be positive in rhodanine staining.^26,27 The low positivity rate is due to detoxification of accumulated copper by metallothionein, one of the copper-binding proteins, in hepatocytes. ²⁷ Copper bound to metallothionein is not detected by rhodanine staining. ²⁵ Therefore, there may have been more cases with HCA, which were not identified via rhodanine staining in this study. To evaluate copper accumulation more precisely, measurement of copper dry weight in the liver is needed and it can only be performed on fresh tissue samples. ¹⁶ The lack of data of ATP7B and COMMD1 variations in cats without HCA was another limitation, though those in healthy cats were examined. In addition, the copper content of food was not evaluated in this study. Therefore, further studies with a larger number of cats with PCH are required to investigate clinical features and the genetic background of cats with PCH.

Conclusions

Some cats with PCH harboured SNVs in ATP7B, indicating an association between ATP7B variations with PCH in cats. However, causes other than SNVs in ATP7B are also suggested. The present results suggest that feline PCH is an equivalent disorder to WD and provide basic evidence to facilitate studies on feline PCH pathophysiology and treatment.

Supplemental Material

Supplementary Figure 1

Results of histological diagnosis and findings of liver specimens of cases 1 and 3