Genetic causes and gene–nutrient interactions in mammalian zinc deficiencies: Acrodermatitis enteropathica and transient neonatal zinc deficiency as examples

Abstract

Discovering genetic causes of zinc deficiency has been a remarkable scientific journey. It started with the description of a rare skin disease, its treatment with various agents, the successful therapy with zinc, and the identification of mutations in a zinc transporter causing the disease. The journey continues with defining the molecular and cellular pathways that lead to the symptoms caused by zinc deficiency. Remarkably, at least two zinc transporters from separate protein families are now known to be involved in the genetics of zinc deficiency. One is ZIP4, which is involved in intestinal zinc uptake. Its mutations can cause acrodermatitis enteropathica (AE) with autosomal recessive inheritance. The other one is ZnT2, the transporter responsible for supplying human milk with zinc. Mutations in this transporter cause transient neonatal zinc deficiency (TNZD) with symptoms similar to AE but with autosomal dominant inheritance. The two diseases can be distinguished in affected infants. AE is fatal if zinc is not supplied to the infant after weaning, whereas TNZD is a genetic defect of the mother limiting the supply of zinc in the milk, and therefore the infant usually will obtain enough zinc once weaned. Although these diseases are relatively rare, the full functional consequences of the numerous mutations in ZIP4 and ZnT2 and their interactions with dietary zinc are not known. In particular, it remains unexplored whether some mutations cause milder disease phenotypes or increase the risk for other diseases if dietary zinc requirements are not met or exceeded. Thus, it is not known whether widespread zinc deficiency in human populations is based primarily on a nutritional deficiency or determined by genetic factors as well. This consideration becomes even more significant with regard to mutations in the other 22 human zinc transporters, where associations with a range of diseases, including diabetes, heart disease, and mental illnesses have been observed. Therefore, clinical tests for genetic disorders of zinc metabolism need to be developed.

Introduction

The origins of research into a rare and fatal skin disease date back over 100 years to an era when knowledge about zinc in human nutrition did not exist. Grover Wende described dermatitis-like skin lesions around the distal parts of extremities in early childhood as atypical epidermolysis bullosa [1]. The disease was often fatal. It was later referred to as generalized moniliasis [2], [3], [4], [5]. In 1935, Thore Edvard Brandt discussed four cases at the Nordic Dermatological Congress in Copenhagen. All the attending dermatologists concurred that the individuals were suffering from atypical acrodermatitis continua [6], [7]. Although Brandt did not discover the cause of the disease or its cure, his observations were particularly important. He noticed that the symptoms began appearing only when patients were being weaned off breast milk. The unusual onset of symptoms and the fact that restoration of breast milk feeding relieved the symptoms led him to believe that some unidentified component in the milk kept these infants healthy and alive. He also concluded that the gastrointestinal tract required further scrutiny due to the large number of gastrointestinal disturbances that his patients experienced [6], [8].

After the 1935 Copenhagen Congress, Niels Danbolt admitted two patients with acrodermatitis continua to his department (Dermatology, University Hospital, Oslo, Norway) for biochemical investigations [9]. The characteristic clinical presentation of his cases was: skin eruptions consisting of vesiculo-pustulo-bullous plaques located in the region of the natural openings of the body and on the distal areas of the extremities; gastrointestinal disturbances, such as diarrhoea and abdominal discomfort, with periods of aggravations and abatements, and alopecia of the scalp, eyebrows and eyelashes. Danbolt treated his patients with assorted vitamins, thyroid hormones and different diets, none of which were successful. His patients died later. He was not able to determine the cause of the disease either, but he recognized that the symptoms did not fit the clinical picture of either epidermolysis bullosa, moniliasis or acrodermatitis continua. The cases were different from epidermolysis bullosa with respect to alopecia, absence of mechanical trauma, which seemed to play a role in epidermolysis bullosa, and the gastrointestinal dysfunctions [10]. Unlike in moniliasis, yeasts were not always present in the skin eruptions, the viscera lacked the typical monilial lesions, progression was occasional rather than steady, total alopecia rather than partial defluvium was observed, and the disease occurred in association with cachexia [8]. Although the types and locations of skin eruptions could be mistaken with those of acrodermatitis continua, the universal presence of gastrointestinal disturbances and alopecia in Danbolt's cases made the disease distinct. Furthermore, acrodermatitis continua was prevalent in middle-aged women [11], whereas the cases admitted were commonly children with a familial tendency for the disease. Considering all these factors and with specific reference to the gastrointestinal disturbances, Danbolt named the condition acrodermatitis enteropathica (AE) [9].

Based on the opinion of Danbolt and Closs [9] that AE was an acral dermatitis with concomitant intestinal lesions, Dillaha et al. [8] treated a patient with a number of intestinal “disinfectants”. 8-Hydroxyquinoline derivatives such as diiodohydroxyquinoline (Diodoquin) (Fig. 1A) were the drugs of choice at the time.

These agents had a positive effect on the skin lesions of the child after just a short period of treatment. The child slowly became symptom-free but relapsed when the treatment was discontinued. Following this initial success, other halogenated derivatives of 8-hydroxyquinoline were tested in several dermatology departments and their efficacy was confirmed [7]. The drugs tested included 5-chloro-7-iodo-8-hydroxyquinoline (Clioquinol, CQ; Fig. 1B), 2-methyl-5,7-dichloro-8-hydroxyquinoline, and 5,7-dibromo-8-hydroxyquinoline [12]. Diiodohydroxyquinoline had been in use as an anti-amoebial drug to eradicate Entamoeba histolytica from the gastrointestinal tract. Since infection with this parasitic amoeba causes inflammation of the lower gastrointestinal tract, it was thought that AE is an amoebal disease. However, the parasitic organism was absent from the patient's stool and the symptoms of the patients did not fit those seen in amoebiasis. It was therefore assumed that Diodoquin works through another, unknown mechanism when taken daily to reverse all symptoms of AE and thus reducing the mortality of the formerly fatal disease [8].

Unfortunately, the long-term treatment with 8-hydroxyquinoline preparations has serious side effects as it leads to optic nerve atrophy and blindness [13], [14]. Therefore, an alternative treatment was desperately sought.

In London (Great Ormand Street Hospital), Edmund John Moynahan [15] examined a 2-year old girl with AE, who also suffered from lactose and other dietary intolerances. Danbolt found that restoration of breast feeding causes prompt remission of symptoms, but cow's milk or other proteinaceous food does not. Hence, Moynahan opined that oligopeptidase, an intestinal enzyme that cleaves most proteins with the exception of proteins in human breast milk, was either absent or defective in AE [16]. He treated his patient with a lactose-deficient synthetic diet together with sugar-free Diodoquin with little lasting benefit as all her skin lesions, bowel syndromes and alopecia reappeared several times until she was 25 months old. From the treatment of phenylketonuria, other aminoacidopathies, and sugar intolerances [16], [17], [18], [19] it was known that synthetic diets lack many growth factors and can cause skin lesions that are refractory to the treatment with standard multi-vitamin preparations. The skin eruptions were not cured even when trebling the vitamin supplements to improve the child's vitamin A, E and pyridoxine status. It was at this point that Moynahan made the decision to thoroughly analyze the child's diet. Only then a striking deficiency in zinc intake, only 0.81 mg (daily requirement: 3.75 mg), in the diet was noticed and confirmed by a low plasma zinc level of 32 μg/100 mL(dL) (normal range for her age: 70–170 μg/100 mL). When Moynahan supplemented her synthetic diet with zinc (35 mg daily) along with Diodoquin, the skin lesions healed, hair started to grow and the plasma zinc levels returned to normal [15]. He also found that in order to avoid relapse the oral zinc supplements are not to be discontinued. He then tested his zinc deficiency hypothesis and treated nine additional AE patients with zinc sulphate (150 mg daily, divided into several doses) and omitted Diodoquin altogether. The affected children eventually became symptom-free, and those that were prepubescent resumed growth. Unlike the treatment with quinolines, the zinc supplements did not appear to have any side effects [7]. With these observations and treatments, Moynahan established zinc deficiency as the cause of AE [16], which K. Michael Hambidge confirmed almost contemporaneously [20], [21].

Moynahan then tried to salvage his “defective enzyme” hypothesis. He postulated that ‘the oligopeptide itself chelates the zinc in the ingested diet and yields an insoluble complex, thereby reducing the availability of this essential trace metal for the metabolic needs of the patient.’ He also assumed that the beneficial effects of Diodoquin are due to binding to the oligopeptide in such a way that the latter does not bind zinc [16]. While deserving credit for finding an effective and safe cure for AE, this hypothesis turned out to be wrong. Hydroxyquinolines enhance zinc absorption by chelating zinc and by facilitating its translocation across lipid membranes (Fig. 1C) [12], [22]. In the early treatment of the child, Diodoquin treatment was ineffective as the synthetic diet did not contain enough zinc [15].

In order to understand the chronology of discoveries, one needs to realize that work on zinc as an essential micronutrient preceded the discovery of Moynahan. While zinc was already known to be essential for plants and animals for several decades, it was Ananda Prasad who discovered nutritional zinc deficiency in humans in 1963 [23]. His patients in Iran, and later in Egypt, exhibited mental lethargy, rough skin, retarded growth, anaemia, hepatosplenomegaly, testicular atrophy, lack of development of primary and secondary sexual characteristics, including lack of facial, axillary or pubic hair. Prasad demonstrated remission of these symptoms when the patients were treated with zinc. The symptomatology and circumstances are different from those in children with AE (Table 1) [9], [15], [17]. The differences are attributed to different degrees of zinc deficiency: moderate zinc deficiency in patients in Iran and Egypt and severe zinc deficiency in the AE patients [24].

While the notion that AE is a familial disorder of zinc deficiency was now accepted, the molecular basis of the disease remained unknown. When ⁶⁵Zn absorption in AE patients, normal individuals and heterozygotes, i.e. parents that are carriers, was compared, the defect in AE was found to be diminished intestinal zinc absorption rather than increased excretion of zinc into faeces, urine or sweat [26]. Ultrastructural examination of the biopsies demonstrated molecularly undefined inclusions in Paneth cells [27]. AE patients accumulated significantly less ⁶⁵Zn than healthy subjects in the jejunal mucosa [28], and duodenal secretions had low zinc-binding capacity [29]. These investigations suggested that AE is the result of a defect in either uptake of zinc by enterocytes or in a transport-facilitating ligand. With the mechanism of zinc absorption still unknown in the 1980s, there were many speculations about how zinc-binding ligands, facilitators and proteins in the intestine and breast milk might be involved in the intestinal uptake of zinc and in the pathogenesis of AE. Prostaglandins [30], [31], citrate [32], picolinate [33] or metallothionein [34] were all discussed as facilitators of zinc uptake. However, the exact roles of zinc absorption or transport and their altered state in AE remained unclear. Amphotericin B (Fig. 2) also ameliorates the symptoms of AE. It increases the membrane permeability for zinc by forming pores through which zinc can enter cells [35], [36]. Though its use for treating AE was limited due to its side effects [37], its effectiveness supported the hypothesis that zinc uptake is defective in AE.

In the 1990s, groundbreaking discoveries of various specific metal transporters in eukaryotic organisms were made. The control of cellular zinc homoeostasis began to be revealed gradually, beginning with the discovery of the first efflux zinc transporter, ZnT1, from rat kidney and the first influx zinc transporter, Zrt-1 (Zinc regulated transporter) from yeast [38], [39]. In the same year, the iron uptake protein, Irt-1 (Iron regulated transporter) was identified in the plant Arabidopsis thaliana as part of the growing family of metal transporters in fungi, nematodes, plants and humans [40]. These proteins (Zrt and Irt) were then grouped into the ZIP (Zrt-Irt-like protein) family of proteins [41], [42]. In humans, there are 10 ZnT proteins and 14 ZIP proteins [43], [44].

Insights into the biochemical basis of the zinc transport defect in AE patients progressed significantly when a difference in protein expression between normal and AE human fibroblasts was noticed [45], [46]. A protein with a molecular mass between 49.6 and 49.9 kDa and a pI value of 5.1 was absent in AE fibroblasts. Homozygosity mapping then localized the AE gene to chromosomal region 8q24.3 [47]. After discovering this susceptibility locus, the gene responsible for AE was identified independently [48], [49]. It was predicted that the gene encodes a transmembrane, histidine-rich protein with significant similarity to members of the ZIP family. The gene was identified as the fourth member of the solute-linked carrier family 39A (SLC39A), i.e. ZIP4 [41], [50]. The protein is highly expressed in the duodenum, jejunum, colon and kidney. It is expressed in matured enterocytes of the intestinal villi and located at the apical surface [48], [49]. This discovery, incidentally exactly 100 years after Wende's report, closed the first chapter of AE research (Table 2). But, as often in science, it posed more questions than it answered. The second chapter now began with discoveries into the multiple mutations causing AE and the molecular functions of the transporter and the biochemical pathways leading to zinc deficiency.