Altered Expression of EPO Might Underlie Hepatic Hemangiomas in LRRK2 Knockout Mice
Parkinson’s disease (PD) is a severe neurodegenerative disorder caused by progressive loss of dopaminergic neurons in the substantia nigra pars compacta of the midbrain. The molecular mechanism of PD pathogenesis is unclear. Mutations in the leucine-rich repeat kinase 2 (LRRK2) gene are a common genetic cause of familial and sporadic PD. However, studies on LRRK2 mutant mice revealed no visible dopaminergic neuronal loss in the midbrain. While surveying a LRRK2 knockout mouse strain, we found that old animals developed age-dependent hepatic vascular growths similar to cavernous hemangiomas. In livers of these hemangioma-positive LRRK2 knockout mice, we detected an increased expression of the HIF-2α protein and significant reactivation of the expression of the HIF-2α target gene erythropoietin (EPO), a finding consistent with a role of the HIF-2α pathway in blood vessel vascularization. We also found that the kidney EPO expression was reduced to 20% of the wild-type level in 18-month-old LRRK2 knockout mice. Unexpectedly, this reduction was restored to wild-type levels when the knockout mice were 22 months to 23 months old, implying a feedback mechanism regulating kidney EPO expression. Our findings reveal a novel function of LRRK2 in regulating EPO expression and imply a potentially novel relationship between PD genes and hematopoiesis.
Parkinson’s disease (PD) is the second most common age-related neurodegenerative disorder and the most common movement disorder . PD patients usually exhibit clinical features such as resting tremor, bradykinesia, rigidity, and postural instability as a result of the progressive loss of the dopaminergic neurons located within the substantia nigra pars compacta (SNpc) of the midbrain. Visible loss of the SNpc dopaminergic neurons and subsequent development of dystrophic striatal projections are the hallmarks of PD .
Dopaminergic neurons of PD patients normally contain Lewy bodies and Lewy neurites, intracellular protein aggregates formed primarily by the α-synuclein protein . Mutations in SNCA (a-synuclein) and LRRK2 are the major causes of autosomal dominant forms of PD [1–4]. Loss-of-function mutations in other genes, for example, Parkin, PINK1, and DJ-1, are the causes of autosomal recessive forms of PD [1, 5–7].
LRRK2 (PARK8) was initially identified in autosomal dominant familial PD patients [3, 4]. Subsequent studies identified LRRK2 mutations to be the major cause of familial and sporadic PD . LRRK2 encodes a 2527-amino acid protein with both kinase and GTPase activities and in human is expressed in multiple tissues [3, 4]. Studies of LRRK2 mutant mice (knockins and knockouts) suggest that LRRK2 might be involved in neural process outgrowth, synaptic vesicle dynamics, and the autophagy/lysosomal pathway [9–15]. However, no apparent loss of dopaminergic neurons was detected in these different LRRK2 mutant mice.
Recently, Tong et al. found that the kidneys of a LRRK2 knockout mouse strain exhibit profound α-synuclein aggregation, oxidative stress, inflammation, and cell death . LRRK2 knockout mice also have early-onset increase in the number and size of secondary lysosomes in kidney proximal tubule cells and lamellar bodies in lung type II cells . LRRK2 knockout rats were found to have abnormal morphologies and/or functions in kidney, spleens, lung, and liver [9, 10]. These findings suggest that LRRK2 might have important functions in peripheral organs.
We postulate that phenotypes in peripheral organs of LRRK2 mutant mice might be more obvious in aged animals, similar to the age-dependent degeneration of dopaminergic neurons in PD patients, and these phenotypes could serve as readouts for understanding the molecular functions of LRRK2. By scrutinizing peripheral organs of a LRRK2 knockout mouse line , we found that these mice develop age-dependent hepatic growths similar to hemangiomas. We also detected altered expression of EPO in the livers and kidneys of the knockout mice. Our findings suggest a previously unknown relationship between LRRK2 and hematopoiesis.
2. Materials and Methods
2.1. LRRK2 KO Mice
The LRRK2 KO (LRRK2−/−) mice (stock number 016209)  were purchased from the Jackson Laboratory (Maine, USA). Genotyping was performed as described . F1 wild-type or LRRK2−/− homozygous mice of the B6/129 genetic background were obtained by crossing P0 LRRK2−/+ heterozygous mice. Littermate F1 wild-type mice were crossed to generate F2 wild-type controls and littermate F1 LRRK2−/− homozygous were crossed to generate F2 LRRK2−/− mice. Analysis was performed with F2 mice. This study was conducted in compliance with the regulations of the Institutional Animal Care and Use Committee of the State Key Laboratory of Medical Genetics of CSU in China.
2.2. Quantitative RT-PCR
Total RNA was extracted from kidney and liver using TRIzol (Invitrogen) and treated with RNase-Free DNase I (New England Biolabs). First-strand cDNA was synthesized using the Superscript III First-Strand Synthesis Kit (Thermo Scientific). ~1 μg total RNA from each sample was used for RT experiments. qPCRs were performed in triplicate for each sample using the Maxima SYBR Green qPCR Master Mix (Thermo Scientific). Fluorescence signals were detected using Bio-Rad CFX96 real-time cycler with the following cycling conditions: 10 min at 95°C, followed by 40 two-step cycles of 15 s at 95°C, and 20 s at 60°C. GAPDH is the reference gene. The expression level of each gene was quantified using Bio-Rad CFX Manager 3.0 and ΔΔCt method. PCR primers are listed in supplementary Table S2 in Supplementary Material available online at http://dx.doi.org/10.1155/2016/7681259.
2.3. Western Blot
Kidneys and livers were homogenized in homogenization buffer (10 v/w) containing 50 mM Tris (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mM sodium orthovanadate, and 25 mM sodium fluoride, supplemented with protease inhibitor mixtures (Thermo Scientific) and phosphatase inhibitor mixtures (Thermo Scientific). 20 μg proteins were loaded onto 10% SDS-polyacrylamide gels. Following electrophoresis, proteins were transferred to an Immobilon -P membrane (Millipore, USA). Membranes were blocked for 1 hour with 5% nonfat milk in Tris-buffered saline containing 0.1% Tween 20 (TBST). Membranes were probed with the primary antibodies diluted in 5% BSA in TBST. The dilution for each antibody was as follows: rabbit anti-LRRK2 mAb (Epitomics), 1 : 3000; rabbit anti-GAPDH pAb (Santa Cruz), 1 : 2000; and rabbit anti-HIF-2α mAb (Santa Cruz), 1 : 2000. After washing with TBST, the blots were incubated with secondary antibodies (Jackson ImmunoResearch) conjugated to horseradish peroxidase. Immunoreactive bands were detected by the chemiluminescence reagent (ECL) (GE Healthcare, UK).
2.4. Histological and Immunohistochemical Analysis
Mice were anesthetized by intraperitoneal injection of 10% chloral hydrate and perfused with 0.9% saline. The kidneys and livers were dissected out and postfixed in 4% paraformaldehyde in PBS (pH 7.4) at 4°C and then processed for paraffin embedding. Kidneys and livers were cut as 5 μm thick sections on a microtome (Leica). For histopathological analysis, tissue sections were stained with hematoxylin and eosin (H&E). For immunohistochemical analysis, tissue sections were subjected to antigen retrieval by microwaving for 15 min in 10 mM sodium citrate buffer, pH 6.0. Endogenous peroxidase activity was quenched by incubating in 3% H2O2. After blocking, sections were incubated with primary antibodies overnight at 4°C followed by 30 min incubation with HRP-conjugated secondary antibodies (ZSGB-BIO, Beijing, China) and then developed using chromogenic DAB substrate (ZSGB-BIO, Beijing, China).
2.5. Hematological Examination
Hematological parameters were evaluated on animals of different ages and analyzed on Beckman Coulter Ac.T 5diff AL automated hematology analyzer (Beckman Coulter, USA). Blood samples were collected via the submandibular vein. Potassium EDTA was used as anticoagulants.
2.6. Statistical Analysis
Statistical analysis was performed using Prism 5 (GraphPad Software) and Excel (Microsoft). Data are presented as means ± SEM. Statistical significance was determined by values of Student’s -test.
3.1. LRRK2 Knockout Mice Had Hepatic Hemangiomas
The LRRK2 knockout (LRRK2−/−) mice had grossly normal dopaminergic neurons at 2 years of age . Interestingly, these LRRK2−/− mice exhibit age-dependent renal atrophy, increased α-synuclein aggregation, impaired autophagy-lysosome pathway, and increased apoptosis and inflammation in kidneys .
We examined other peripheral organs of old LRRK2−/− mice and found that a proportion of these mice exhibited hepatic growths that resembled vascularized blood vessels (Figure 1(a)). 12 of 36 LRRK2−/− mice (33%) had such growths (Figure 1(b)), the youngest of which was 19 months old. No such growths were found in 35 wild-type controls. The sizes of the growths vary from those obvious with naked eyes (Figure 1(a), left) to those of microscopic scales (Figure 1(a), right). Among the 12 mice with the growths, five were males and seven were females, suggesting that sex might not affect the phenotype.
We examined liver sections of the LRRK2−/− mice and found that the growths contained tangles of small and large blood vessels partially or completely filled with red blood cells (Figure 1(c)). This phenotype was not observed in the livers of control mice (Figure 1(c), left) or LRRK2−/− mice without the liver growths (BW and LM, unpublished observations). The gross morphology and the microscopic structure of the liver growths are similar to the characteristics of hepatic hemangiomas , a benign tumor composed of hepatic endothelial cells .
In humans LRRK2 is expressed in multiple tissues including the brain, heart, lung, kidney, and liver [3, 4]. In mice LRRK2 is abundantly expressed in the brain, lung, heart, kidney, spleen, and lymph node and weakly in liver, eye, skeletal muscle, smooth muscle, thymus, stomach, and small intestine [19, 20]. We examined LRRK2 expression in the kidneys and livers at two different ages using Western blot (Figures 1(d) and 1(e), upper panels). As expected, LRRK2 is abundantly expressed in the kidneys and the LRRK2−/− mutation completely abolishes the expression (Figures 1(d) and 1(e)). However, no obvious LRRK2 expression was detected in the livers by Western blot. We further examined LRRK2 mRNA expression by qRT-PCR (Figures 1(d) and 1(e), lower panels). Here we detected a residual LRRK2 expression (12% and 10%, resp.) in the livers at both ages, consistent with the weak expression of LRRK2 reported previously [19, 20].
3.2. Increased Erythropoietin Expression in the Livers of HA-Positive LRRK2−/− Mice
The genetic cause of hepatic hemangiomas is unclear. Previous studies found that the signal pathway involving the von Hippel-Lindau (VHL) factor, the HIF-2α transcriptional factor, and HIF-2α target gene erythropoietin (EPO) could affect the formation of cavernous hemangiomas in mouse livers [17, 21].
We examined HIF-2α expression by immunohistochemistry and found an apparent increase of HIF-2α signal in the nuclei of numerous cells surrounding the vascularized vessels in the livers of hemangioma- (HA-) positive LRRK2−/− mice (Figure 2(a), HA+). However, no visible HIF-2α signal was detected in livers of wild-type or HA-negative LRRK2−/− mice (Figure 2(a), WT and HA−, resp.). Similarly, no visible HIF-2α signals were detected in the livers of either wild-type or LRRK2−/− mice of younger ages (supplementary Figure S1).
To determine whether the increased expression of HIF-2α has any consequences, we examined the transcript level of its target gene erythropoietin  by qRT-PCR. In HA-positive LRRK2−/− mice, we detected an apparently increased (reactivated) EPO expression (Figure 2(b), HA+), which was not detected or was barely detected in the livers of either wild-type or HA-negative LRRK2−/− mice (Figure 2(b)). In addition, no obvious EPO signals were detected in the livers of both wild-type and LRRK2−/− mice of younger ages (BW and LM, unpublished observations).
To identify other factors that might be involved in the formation of hemangiomas in LRRK2−/− mice, we examined the transcript level of the vascular endothelial growth factor (VEGF) gene . Similar levels of VEGF were detected in the livers of 22-month-old to 23-month-old mice of both wild-type and LRRK2−/− genotypes (Figure 2(c)). The presence of hepatic hemangiomas did not alter the expression either (Figure 2(c)). This similarity was also found in wild-type and LRRK2−/− mice of younger ages (Figure 2(d)). Therefore, LRRK2 might not affect the expression of VEGF in the liver.
3.3. Altered EPO Expression in the Kidneys of LRRK2−/− Mice
During embryonic development, the liver is the major site of EPO production, which is taken over by the kidney after birth . It is possible that the altered expression of EPO in the livers of LRRK2−/− mice might also be related to a change of EPO expression in the kidneys.
We measured EPO expression in the kidneys of mice of different ages. At the ages of 2 and 12 months, the expression was similar between wild-type and LRRK2−/− mice (Figures 3(a) and 3(b)). However, in the kidneys of 18-month-old to 19-month-old LRRK2−/− mice, EPO expression was reduced to ~20% of the wild-type level (Figure 3(c)). This reduction was restored in LRRK2−/− mice at 22 to 23 months of ages (Figure 3(d)). We found that five of the six HA-positive LRRK2−/− mice had apparently increased EPO levels compared to wild-type mice (supplementary Table S1). However, no significant difference was detected between these groups by statistics probably due to the wide variations of the values for the HA-positive group (Figure 3(d) and supplementary Table S1).
We tried but failed to detect an apparent HIF-2α signal in the kidneys of both wild-type and LRRK2−/− mice at different ages by Western blot (BW and LM, unpublished observations) or immunohistochemistry (supplementary Figures S2A and S2B). In the kidneys of 22-23-month-old LRRK2−/− mice, we observed apparent histochemical signals that were not caused by HIF-2α immunostaining (supplementary Figure S2A). These signals might be similar to the increased autolysosome and lipofuscin signals described previously .
Since an altered EPO expression was detected in the LRRK2−/− mice, we examined whether this change would be linked to any alterations in the hematological profiles of these mice. In all groups at different ages, we failed to detect a significant difference in red blood cell count (RBC count), hemoglobin concentration, and hematocrit value (Figures 4(a), 4(b), and 4(c)). Therefore, a more extensive study is warranted to determine whether any EPO-related processes are altered in these mice.
In this study, we report the identification of hemangioma-like growths in the livers of old LRRK2−/− mice and found that LRRK2 could affect the expression of EPO in the livers and kidneys of these mice.
LRRK2 mutations are the most frequent genetic causes of familial and sporadic PD . To date, no apparent dopaminergic neuron degeneration was observed in various LRRK2 mutant mice [13–16], raising the question that unknown genetic differences might be responsible for the discrepancy between human and mouse. Interestingly, LRRK2 mutant rodents, including mice and rats, had physiological defects in peripheral organs such as kidney, lung, and liver [9–11, 13, 14, 16]. Considering the current lack of understanding on PD-related genes, a detailed analysis of LRRK2 in peripheral organs not only is meaningful but also could provide novel insights into PD pathogenesis.
We found that 33% of LRRK2−/− mice after the age of 19 months develop liver growths that are similar to hepatic hemangiomas, suggesting that LRRK2 might have an age-dependent function in affecting either erythropoiesis or angiogenesis in mice. The incomplete penetrance of this phenotype implies that either unidentified genetic modifiers or environmental factors are involved.
The genetic cause of hepatic hemangiomas remains unclear. Loss-of-function mutations in the von Hippel-Lindau tumor suppressor gene (VHL) could lead to age-dependent hepatic hemangiomas in mice . Inactivation of HIF-2α can suppress the formation of hepatic hemangiomas in VHL mutant mice , suggesting that HIF-2α might promote the formation of hepatic hemangiomas. Consistent with this notion, we found that cells lining the vascularized vessels in the livers of HA-positive LRRK2−/− mice exhibited apparently increased expression of HIF-2α. This increase was not found in the livers of wild-type and HA-negative LRRK2−/− mice. Similarly, the expression of the HIF-2α target gene EPO was dramatically increased (reactivated) in the livers of HA-positive LRRK2−/− mice, suggesting a close correlation of EPO expression and the formation of hemangiomas in LRRK2−/− mice. Furthermore, the implication of LRRK2 in this process provides a new insight into the genetics of hepatic hemangiomas.
VEGF is a signal protein that promotes vasculogenesis and angiogenesis . We measured VEGF expression in mice at four different ages (2, 12, 18-19, and 22-23 months) but failed to detect an apparent change in either the kidneys or livers of LRRK2−/− mice with or without hepatic hemangiomas. Hence, the altered expression of EPO but not VEGF in LRRK2−/− mice suggests a special relationship between LRRK2 and EPO, which does not simply result from a general defect caused by the LRRK2 mutation.
In embryos, the liver is the major site for EPO production, while in adults the kidney takes over this role . We found that EPO expression is significantly reduced in the kidneys of 18-19-month-old LRRK2−/− mice. Interestingly, the reduction was restored when the mice became 22 to 23 months old, implying a feedback mechanism. The reduced EPO expression might be caused by impaired kidney functions, since these mice exhibited renal atrophies by the age of 7 months and severe renal injuries at 20 months of age [11, 14]. Alternatively, this change might suggest a regulatory role of LRRK2 in the expression of EPO. Future studies might reveal the underlying mechanism that can distinguish these two possibilities.
Our findings provide a first line of genetic evidence that LRRK2 might be intrinsically related to EPO expression. Following the line, other PD genes might also be involved in EPO expression in adult kidneys. Indeed, a recent epidemiological study found that PD patients tend to have anemia years before developing PD symptoms . Our findings also provide supporting evidence for the notion that the neural protective function of EPO could be applied for treating PD and other neurodegenerative diseases [26–28].
In short, we found that LRRK2 could affect EPO expression in the kidneys and livers of aged mice, which might be related to the formation of hepatic hemangiomas. Future investigation into this phenomenon might provide new molecular details into the function of LRRK2 in peripheral organs and PD pathogenesis.
The funding sources had no roles in the design of experiments, collection/analysis/interpretation of data, the writing of the article, and the decision to submit the article for publication.
The authors declare that there are no competing interests.
Ben Wu performed the experiments and collected and analyzed the data. Kaifu Xiao collected the data. Ben Wu, Zhuohua Zhang, and Long Ma wrote the manuscript.
The authors thank J. Shen for suggestions. This study was supported by a National Basic Research Program of China grant (2011CB510005) and a Natural Science Foundation of China grant (31371253) to Long Ma.
HIF-2a immunohistochemistry, EPO levels in individual mouse and list of PCR primers.
Y. Tong, H. Yamaguchi, E. Giaime et al., “Loss of leucine-rich repeat kinase 2 causes impairment of protein degradation pathways, accumulation of α-synuclein, and apoptotic cell death in aged mice,” Proceedings of the National Academy of Sciences of the United States of America, vol. 107, no. 21, pp. 9879–9884, 2010.View at: Publisher Site | Google Scholar
V. H. Haase, J. N. Glickman, M. Socolovsky, and R. Jaenisch, “Vascular tumors in livers with targeted inactivation of the von Hippel-Lindau tumor suppressor,” Proceedings of the National Academy of Sciences of the United States of America, vol. 98, no. 4, pp. 1583–1588, 2001.View at: Publisher Site | Google Scholar
A. Toro, A.-E. Mahfouz, A. Ardiri et al., “What is changing in indications and treatment of hepatic hemangiomas. A review,” Annals of Hepatology, vol. 13, no. 4, pp. 327–339, 2014.View at: Google Scholar