Liver changes in severe Plasmodium falciparum malaria: histopathology, apoptosis and nuclear factor kappa B expression
© Viriyavejakul et al.; licensee BioMed Central Ltd. 2014
Received: 19 December 2013
Accepted: 11 March 2014
Published: 17 March 2014
Liver involvement in severe Plasmodium falciparum infection is commonly a significant cause of morbidity and mortality among humans. The clinical presentation of jaundice often reflects a certain degree of liver damage. This study investigated the liver pathology of severe P. falciparum malaria as well as the regulation and occurrence of apoptosis in cellular components of formalin-fixed, paraffin-embedded liver tissues.
The liver tissues used in the study came from patients who died from P. falciparum malaria with hyperbilirubinaemia (total bilirubin (TB) ≥ 51.3 μmol/L or 3 mg/dl) (12 cases), P. falciparum malaria without hyperbilirubinaemia (TB < 51.3 μmol/L) (10 cases); and patients who died due to accidents, whose liver histology was normal (the control group) (10 cases). The histopathology of the liver tissue was studied by routine histology method. Caspase-3 and nuclear factor kappa B (NF-κB) p65 expressions were determined using immunohistochemistry.
The severity of liver histopathology, occurrence of apoptosis and NF-κB p65 activation in P. falciparum malaria were associated with higher TB level. Significant correlations were found between NF-κB p65 expression and apoptosis in Kupffer cells and lymphocytes in the portal tracts.
Hyperplastic Kupffer cells and portal tract inflammation are two main features found in the liver tissues of severe P. falciparum malaria cases. In addition, NF-κB is associated with Kupffer cells and lymphocyte apoptosis in severe P. falciparum malaria.
Plasmodium falciparum malaria is a life-threatening infectious disease that remains a major global health problem. The severe manifestations often present clinically as cerebral malaria, pulmonary oedema, acute kidney injury, hypoglycaemia, lactic acidosis, anaemia and liver involvement. Plasmodium falciparum malaria causes clinical jaundice in 2.5-5.3% of cases in endemic areas[2, 3]. The liver is an important organ involved during the hepatic stage of the malaria parasite’s life cycle, where malaria sporozoites develop into merozoites. The merozoites are then released into the circulation and enter the erythrocytic stage. In the erythrocytic stage, parasitized red blood cells (PRBCs) become sequestered in small blood vessels. The degraded haemozoin pigment is then engulfed by local tissue macrophages, such as Kupffer cells and alveolar macrophages. Common histopathological findings of the liver in P. falciparum malaria include reactive Kupffer cells, retention of haemozoin pigment and minimal PRBC sequestration[4, 5]. An ultrastructural study reported an association between high PRBC load in the livers of malaria patients with jaundice, hepatomegaly and liver enzyme elevation.
Apoptotic changes occur in a variety of cellular systems and involve both physiologic and pathologic changes. While apoptotic change in the liver have not been documented in human malaria, changes have been reported in animal models during the erythrocytic stage in hepatocytes[7, 8] and during the hepatic stage in Kupffer cells. This process of programmed cell death can be mediated by various stimuli, including hormones, cytokines, growth factors, bacterial or viral infections and the immune response. Cell apoptosis is regulated via two major pathways: the intrinsic or mitochondrial pathway and the extrinsic or death-receptor pathway. Initiator caspases, such as caspase-8 or -9, play a regulatory role by activating downstream effector caspases, such as caspase-3, -6, or -7. NF-κB has been shown to regulate the apoptotic program in various cell types, either as an up-regulating response or as an apoptosis blocker. Evidence of NF-κB regulating apoptosis was found in the brain endothelial cells and intravascular lymphocytes in cerebral malaria. However, no linkage between NF-κB and apoptosis has been reported in the livers of P. falciparum malaria patients. This study evaluated the liver pathology of severe P. falciparum malaria in association with total bilirubin (TB) level. The occurrence of apoptosis and its relation to a signaling molecule (NF-κB) in liver tissues was investigated.
Liver tissue specimens from malaria patients and controls
Specimens were classified into two groups according to the level of total bilirubin (TB): TB ≥ 51.3 μmol/L (3 mg/dl) (12 cases) and TB < 51.3 μmol/L (10 cases), based on laboratory data. Liver specimens with normal histology obtained from fatal accident cases (10 cases), served as the control group. The specimens were obtained from autopsied cases at the Department of Tropical Pathology, Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand. Control cases came from the same region. Patients with hepatitis B co-infection, and patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency, were excluded from the study. The use of left-over liver specimens and the study protocol were reviewed and approved by the Ethics Committee of the Faculty of Tropical Medicine, Mahidol University (MUTM 2011-025-01 and MUTM 2011-025-02).
Liver tissue preparation
Left-over liver tissues in paraffin-embedded blocks were re-embedded with new paraffin medium using standard histological techniques. Liver tissues were sectioned at 4 μm thickness and placed onto glass slides for histopathological examination and onto positively charged slides for immunohistochemistry study against cleaved caspase-3 and NF-κB p65.
Histology of liver in severe P. falciparum malaria
Histopathological changes and grading schemes used to evaluate liver tissues
No fatty change
Portal tract inflammation
< 5% of portal tract area
5-15% of portal tract area
16-30% of portal tract area
> 30% of portal tract area
Bile duct proliferation
Immunohistochemical staining of cleaved caspase-3 and NF-κB p65
The 4 μm paraffin sections were deparaffinized and rehydrated by sequential immersion in a graded series of alcohol, then transferred into water for 5 mins. To inhibit endogenous peroxidase activity, the sections were incubated with 3% H2O2 for 5 mins. The sections were then heated in a microwave oven (in 0.1 M sodium citrate buffer, pH 6.0 for cleaved caspase-3 and 0.1 M Tris–HCl buffer, pH 9.0 for NF-κB p65) for 20 mins for epitope retrieval. After washing with phosphate buffered saline (PBS), pH 7.4, sections were incubated for 30 mins with normal serum as blocking solution to reduce the non-specific background, then cooled to room temperature while still immersed in buffer. The following protocol was realized using avidin-biotin alkaline phosphatase complex (VECTASTAIN® ABC-AP kit (Rabbit IgG) # AK-5001) for cleaved caspase 3 and avidin-biotin peroxidase complex (VECTASTAIN® ABC kit (Mouse IgG) # PK-4002) for NF-κB p65 (Vector Laboratories, Inc., USA) according to the manufacturer’s protocol. The sections were incubated with primary antibody; rabbit polyclonal anti-cleaved caspase-3 (Asp175) antibody (1:200 dilution) (Cell Signaling Technology, USA) and mouse monoclonal anti-NF-κB p65 (1:50 dilution) (Santa Cruz Biotechnology Inc., Santa Cruz, CA) and incubated overnight at 4˚C in a humidity chamber. The following day, sections were washed three times with PBS, and incubated with anti-mouse/rabbit biotinylated secondary antibody (Vector Laboratories, Inc., USA) for 30 mins at room temperature and reacted with avidin-biotin complex (ABC) conjugated with alkaline phosphatase (AP)/horseradish peroxidase (HRP) (Vector Laboratories, Inc., USA) for 30 mins. After washing, enzyme activity was visualized by Vector® Red substrate kit (Vector Laboratories, Inc., USA) for cleaved caspase-3, resulting in the formation of a red colour at the antigen sites and by 3,3′- diaminobenzidine (DAB) (Vector Laboratories, Inc., USA) for NF-κB p65, presenting as a brown colour. Subsequently, sections were counterstained with haematoxylin for 1 min, and mounted with a coverslip.
Cleaved caspase-3 and NF-κB p65 analysis
The presence of immunopositive cells for cleaved caspase-3 and NF-κB p65 was recorded as percentages. In addition, immunostaining intensity was scored from 0–3 (0-negative staining, 1- mild, 2- moderate, and 3- strong immunostaining). Total score was calculated by multiplying percentage immunopositive cells and intensity, a method used by Punsawad et al. 2013.
Measurement of Kupffer cell length
Ten representative views of H&E-stained liver sections were randomly photographed at 400x magnification using an Olympus Bx41 light microscope (Olympus, Tokyo, Japan) connected to an Olympus DP20 digital camera (Olympus, Tokyo, Japan). Kupffer cell length was measured with the UTHSCSA Image Tool program (developed at the University of Texas Health Science Center at San Antonio, TX; freely available from the Internet).
Data were expressed as mean ± standard error of the mean (SEM). The normality of distribution was determined by the Kolmogorov-Smirnov test. Differences between groups were analyzed by Mann Whitney U- test. In addition, the correlations of each variable within groups and pertinent clinical data were calculated using Spearman’s rank correlation (r s ). Statistical analysis was performed using SPSS version 17.0 software (SPSS, IL, USA). A p value < 0.05 was considered significantly different.
Clinical and laboratory parameters of P. falciparum malaria patients
Non-hyperbilirubinaemia (TB < 51.3 μmol/L) (n = 10)
Hyperbilirubinaemia (TB ≥ 51.3 μmol/L) (n = 12)
Age (years) (p = 0.974)
25.8 ± 3.95
26.17 ± 4.28
Days of fever (p = 0.095)
4.1 ± 0.82
5.83 ± 0.77
Parasitaemia (/μl) (p = 0.619)
303,186.67 ± 151,070.30
391,501.40 ± 183,362.30
Albumin (g/L) (p = 0.006)
33.2 ± 0.17
25.3 ± 0.13
AST (U/L) (p = 0.006)
72.00 ± 21.41
266.33 ± 64.93
ALT (U/L) (p = 0.049)
49.17 ± 8.23
126.79 ± 29.22
Alkaline phosphatase (U/L) (p = 0.049)
8.42 ± 2.62
20.07 ± 4.33
Total bilirubin (μmol/L) (p = 0.001)
30.61 ± 0.31
441.60 ± 5.39
Direct bilirubin (μmol/L) (p = 0.001)
8.55 ± 0.15
217.17 ± 2.73
Histopathology of liver in severe P. falciparum malaria
Histopathologic grading of liver tissues in P. falciparum malaria
Non-hyperbilirubinaemia (TB < 51.3 μmol/L)
Hyperbilirubinaemia (TB ≥ 51.3 μmol/L)
Kupffer cells hyperplasia
Portal tract inflammation
Bile duct proliferation
Total histological score
Quantification of fatty changes, hyperplastic Kupffer cells and portal inflammation in the livers of P. falciparum malaria patients compared with the control group
Normal liver (n = 10)
Non-hyperbilirubinaemia (TB < 51.3 μmol/L) (n = 10)
Hyperbilirubinaemia (TB ≥ 51.3 μmol/L) (n = 12)
Fatty changes (%/HPF)
0.59 ± 0.25
1.71 ± 0.46
1.97 ± 0.67
Hyperplastic Kupffer cells (count/HPF)
9.65 ± 0.67
32.05 ± 3.34a
52.21 ± 2.32a,b
Portal tract inflammation (%)
5.49 ± 1.20
17.1 ± 1.62a
32.79 ± 2.48a,b
Occurrence of apoptosis in the livers of severe P. falciparum malaria cases
NF-κB p65 expression in the livers of severe P. falciparum malaria
Correlation between apoptosis and NF-κB p65 expression
Liver pathology in severe P. falciparum malaria
The present study shows a rise in liver transaminases and alkaline phosphatase in the malaria group with hyperbilirubinaemia (TB ≥ 51.3 μmol/L). Clinical jaundice in P. falciparum can be caused by several factors, i.e. intravascular haemolysis from parasitized red blood cells (PRBCs), G6PD deficiency-related haemolysis or anti-malarial drugs, disseminated intravascular coagulation (DIC) or co-existing septicaemia-induced hepatitis[14, 15]. The histopathology of the liver in severe malaria has been previously studied. However, the present study demonstrated certain morphological variations in the liver from other reports, such as an abundant chronic inflammatory cell response and an absence of liver cell necrosis. Liver changes in severe malaria often include hyperplastic Kupffer cells[4, 5, 16–18], fatty change[16, 17], portal tract inflammation, cholestasis[16, 17], liver cell necrosis[4, 16, 18], sequestration of PRBCs and the deposition of haemozoin pigment[4, 5, 16, 18]. The present study documented hyperplastic Kupffer cells with scattered haemozoin deposition and portal inflammation as the most common histological changes in the livers of severe P. falciparum malaria cases. The enlarged Kupffer cells were confirmed quantitatively using Image Tool software. The immune response in the liver to PRBCs primarily involves the activation of Kupffer cells. The recruitment and activation of Kupffer cells and macrophages in the spleen and bone marrow are important for the clearance of malaria parasites. Portal tract inflammation consists mainly of lymphocytes and a few plasma cells (Table 4), in contrast with mild inflammation (portal and lobular lymphocytic infiltrates) reported earlier. An acute inflammatory process involving neutrophils is not seen. The minimal fatty change noted here was similar to a previous finding. Liver cell necrosis in P. falciparum was not a striking finding in this study. It has also been reported as a rare event in some other studies[5, 18]. However, the incidence of hepatic necrosis may be as high as 41% and severe cases of centrizonal necrosis have been documented. This change has been reported to be secondary to suppression of bilirubin excretion by PRBCs or metabolic acidosis rather than hepatitis per se. Sequestration of PRBCs is a common finding and depends on malaria parasite load.
Apoptosis and NF-κB p65 in the liver of severe P. falciparum malaria cases
Among various cells evaluated in the liver tissue, Kupffer cells and inflammatory cells show significant apoptotic changes in severe P. falciparum malaria. Hepatocytes, bile ducts and ECs failed to show significant apoptotic changes. P. falciparum has been shown to induce apoptosis in human cells, such as in lymphocytes[13, 20–22], neurons, glial cells, brain ECs and lung ECs. This may be responsible for clinical manifestations and progression to severe disease. In animal models, malaria-induced apoptosis was evident in astrocytes, lymphocytes, liver and spleen[8, 25]. During the liver stage, however, hepatocytes are usually spared from the apoptotic process, allowing merozoites to be released into the circulation, suggesting that malaria sporozoites can block pro-apoptotic pathways. The present study focuses on liver changes in the erythrocytic stage, which is far beyond the liver stage of parasite development. Nevertheless, hepatocytes remain morphologically unaffected and protected from apoptosis. Hepatocytes are defended by a barrier of Kupffer cells, endothelial cells, stellate cells, space of Disse (perisinusoidal space) and Kupffer cells. Moreover, PRBCs localized within the sinusoidal area are in close contact with Kupffer cells, the first-line immune defense in the liver. Recognized as foreign bodies, PRBCs and haemozoin are primarily engulfed by the Kupffer cells. In contrast, apoptosis in the hepatocytes has been reported in animal models, linked to activation of the mitochondrial pathway, release of reactive oxygen species[7, 8] and induction by glycosylphophatidylinositol (GPI), a major membrane-associated protein of P. falciparum.
Apoptosis in the Kupffer cells was evidenced by strong caspase-3 expression. The loaded haemozoin within the cytoplasm of Kupffer cells can be toxic to these immune cells. In humans, during the erythrocytic stage, malaria parasites degrade haemoglobin to produce haem and haemozoin which are harmful. Haemozoin can be deposited in the liver and primarily phagocytized by Kupffer cells, where it can induce oxidative stress, a possible mechanism for the induction of apoptosis in Kupffer cells. During the liver stage, Kupffer cell apoptosis has also been detected in a murine model after incubation with Plasmodium yoelii sporozoites. Apoptosis in malaria is reportedly mediated by the Fas-ligand in lymphocytes and in murine astrocytes.
NF-κB has been shown to regulate various cellular processes such as inflammation, immunity, cell proliferation and apoptosis. A previous study documented NF-κB activation and pro-inflammatory response in human brain ECs exposed to PRBCs. A recent study of human brain tissues has demonstrated that NF-κB is one of the signaling molecules that modulates apoptosis in brain ECs and intravascular leukocytes in fatal cerebral malaria. The present study documented NF-κB mediating apoptosis in Kupffer cells and lymphocytes within the portal tract in severe P. falciparum infection.
Histopathological changes in the livers of severe P. falciparum malaria cases are associated with total bilirubin levels. Apoptosis of Kupffer cells and portal tract lymphocytes is a significant finding and is related to NF-κB activation.
The study was funded by research grants from the Office of the Higher Education Commission and Mahidol University under the National Research Universities (NRU) Initiative, Thailand (by PV) and the Faculty of Tropical Medicine, Mahidol University, year 2011 (by VK). We thank Dr. Mario Riganti for his generous guidance and fruitful suggestions. Help from the staff at the Department of Tropical Pathology, Faculty of Tropical Medicine, Mahidol University, Thailand is highly appreciated.
- White NJ: The treatment of malaria. N Engl J Med. 1996, 335: 800-806. 10.1056/NEJM199609123351107.View ArticlePubMedGoogle Scholar
- Anand AC, Ramji C, Narula AS, Singh W: Malarial hepatitis: a heterogeneous syndrome?. Natl Med J India. 1992, 5: 59-62.PubMedGoogle Scholar
- Mehta SR, Naidu G, Chandar V, Singh IP, Johri S, Ahuja RC: Falciparum malaria–present day problems. An experience with 425 cases. J Assoc Physicians India. 1989, 37: 264-267.PubMedGoogle Scholar
- Rupani AB, Amarapurkar AD: Hepatic changes in fatal malaria: an emerging problem. Ann Trop Med Parasitol. 2009, 103: 119-127. 10.1179/136485909X385054.View ArticlePubMedGoogle Scholar
- Whitten R, Milner DA, Yeh MM, Kamiza S, Molyneux ME, Taylor TE: Liver pathology in Malawian children with fatal encephalopathy. Hum Pathol. 2011, 42: 1230-1239. 10.1016/j.humpath.2010.11.019.PubMed CentralView ArticlePubMedGoogle Scholar
- Prommano O, Chaisri U, Turner GD, Wilairatana P, Ferguson DJ, Viriyavejakul P, White NJ, Pongponratn E: A quantitative ultrastructural study of the liver and the spleen in fatal falciparum malaria. Southeast Asian J Trop Med Public Health. 2005, 36: 1359-1370.PubMedGoogle Scholar
- Dey S, Guha M, Alam A, Goyal M, Bindu S, Pal C, Maity P, Mitra K, Bandyopadhyay U: Malarial infection develops mitochondrial pathology and mitochondrial oxidative stress to promote hepatocyte apoptosis. Free Radic Biol Med. 2009, 46: 271-281. 10.1016/j.freeradbiomed.2008.10.032.View ArticlePubMedGoogle Scholar
- Guha M, Kumar S, Choubey V, Maity P, Bandyopadhyay U: Apoptosis in liver during malaria: role of oxidative stress and implication of mitochondrial pathway. FASEB J. 2006, 20: 1224-1226. 10.1096/fj.05-5338fje.View ArticlePubMedGoogle Scholar
- Klotz C, Frevert U: Plasmodium yoelii sporozoites modulate cytokine profile and induce apoptosis in murine Kupffer cells. Int J Parasitol. 2008, 38: 1639-1650. 10.1016/j.ijpara.2008.05.018.PubMed CentralView ArticlePubMedGoogle Scholar
- Thompson CB: Apoptosis in the pathogenesis and treatment of disease. Science. 1995, 267: 1456-1462. 10.1126/science.7878464.View ArticlePubMedGoogle Scholar
- Thornberry NA: Caspases: key mediators of apoptosis. Chem Biol. 1998, 5: R97-103. 10.1016/S1074-5521(98)90615-9.View ArticlePubMedGoogle Scholar
- Shishodia S, Aggarwal BB: Nuclear factor-kappaB: a friend or a foe in cancer?. Biochem Pharmacol. 2004, 68: 1071-1080. 10.1016/j.bcp.2004.04.026.View ArticlePubMedGoogle Scholar
- Punsawad C, Maneerat Y, Chaisri U, Nantavisai K, Viriyavejakul P: Nuclear factor kappa B modulates apoptosis in the brain endothelial cells and intravascular leukocytes of fatal cerebral malaria. Malar J. 2013, 12: 260-10.1186/1475-2875-12-260.PubMed CentralView ArticlePubMedGoogle Scholar
- Anand AC, Puri P: Jaundice in malaria. J Gastroenterol Hepatol. 2005, 20: 1322-1332. 10.1111/j.1440-1746.2005.03884.x.View ArticlePubMedGoogle Scholar
- World Health Organization: Severe falciparum malaria: Communicable Diseases Cluster. Trans R Soc Trop Med Hyg. 2000, 94 (Suppl 1): S1-S90.View ArticleGoogle Scholar
- Srivastava A, Khanduri A, Lakhtakia S, Pandey R, Choudhuri G: Falciparum malaria with acute liver failure. Trop Gastroenterol. 1996, 17: 172-174.PubMedGoogle Scholar
- Chawla LS, Sidhu G, Sabharwal BD, Bhatia KL, Sood A: Jaundice in Plasmodium falciparum. J Assoc Physicians India. 1989, 37: 390-391.PubMedGoogle Scholar
- Joshi YK, Tandon BN, Acharya SK, Babu S, Tandon M: Acute hepatic failure due to Plasmodium falciparum liver injury. Liver. 1986, 6: 357-360.View ArticlePubMedGoogle Scholar
- Chua CL, Brown G, Hamilton JA, Rogerson S, Boeuf P: Monocytes and macrophages in malaria: protection or pathology?. Trends Parasitol. 2013, 29: 26-34. 10.1016/j.pt.2012.10.002.View ArticlePubMedGoogle Scholar
- Riccio EK, Junior IN, Riccio LR, das Gracas Alecrim M, Corte-Real S, Morgado M, Daniel-Ribeiro CT, de Fatima Ferreira-da-Cruz M: Malaria associated apoptosis is not significantly correlated with either parasitemia or the number of previous malaria attacks. Parasitol Res. 2003, 90: 9-18.PubMedGoogle Scholar
- Balde AT, Sarthou JL, Roussilhon C: Acute Plasmodium falciparum infection is associated with increased percentages of apoptotic cells. Immunol Lett. 1995, 46: 59-62. 10.1016/0165-2478(95)00017-Y.View ArticlePubMedGoogle Scholar
- Toure-Balde A, Sarthou JL, Aribot G, Michel P, Trape JF, Rogier C, Roussilhon C: Plasmodium falciparum induces apoptosis in human mononuclear cells. Infect Immun. 1996, 64: 744-750.PubMed CentralPubMedGoogle Scholar
- Pino P, Vouldoukis I, Kolb JP, Mahmoudi N, Desportes-Livage I, Bricaire F, Danis M, Dugas B, Mazier D: Plasmodium falciparum–infected erythrocyte adhesion induces caspase activation and apoptosis in human endothelial cells. J Infect Dis. 2003, 187: 1283-1290. 10.1086/373992.View ArticlePubMedGoogle Scholar
- Potter SM, Chan-Ling T, Rosinova E, Ball HJ, Mitchell AJ, Hunt NH: A role for Fas-Fas ligand interactions during the late-stage neuropathological processes of experimental cerebral malaria. J Neuroimmunol. 2006, 173: 96-107. 10.1016/j.jneuroim.2005.12.004.View ArticlePubMedGoogle Scholar
- Helmby H, Jonsson G, Troye-Blomberg M: Cellular changes and apoptosis in the spleens and peripheral blood of mice infected with blood-stage Plasmodium chabaudi chabaud i AS. Infect Immun. 2000, 68: 1485-1490. 10.1128/IAI.68.3.1485-1490.2000.PubMed CentralView ArticlePubMedGoogle Scholar
- van de Sand C, Horstmann S, Schmidt A, Sturm A, Bolte S, Krueger A, Lutgehetmann M, Pollok JM, Libert C, Heussler VT: The liver stage of Plasmodium berghei inhibits host cell apoptosis. Mol Microbiol. 2005, 58: 731-742. 10.1111/j.1365-2958.2005.04888.x.View ArticlePubMedGoogle Scholar
- Wichmann D, Schwarz RT, Ruppert V, Ehrhardt S, Cramer JP, Burchard GD, Maisch B, Debierre-Grockiego F: Plasmodium falciparum glycosylphosphatidylinositol induces limited apoptosis in liver and spleen mouse tissue. Apoptosis. 2007, 12: 1037-1041. 10.1007/s10495-006-0033-9.View ArticlePubMedGoogle Scholar
- Li Q, Verma IM: NF-kappaB regulation in the immune system. Nat Rev Immunol. 2002, 2: 725-734. 10.1038/nri910.View ArticlePubMedGoogle Scholar
- Tripathi AK, Sha W, Shulaev V, Stins MF, Sullivan DJ: Plasmodium falciparum-infected erythrocytes induce NF-kappaB regulated inflammatory pathways in human cerebral endothelium. Blood. 2009, 114: 4243-4252. 10.1182/blood-2009-06-226415.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.